link
stringlengths 41
45
| date
stringlengths 9
9
| paper
dict | reviews
listlengths 1
6
| version
int64 1
5
| main
stringlengths 38
42
|
|---|---|---|---|---|---|
https://f1000research.com/articles/9-1209/v1
|
08 Oct 20
|
{
"type": "Research Article",
"title": "Road traffic injuries in Nepal during COVID-19 lockdown",
"authors": [
"Bhagabati Sedain",
"Puspa Raj Pant",
"Puspa Raj Pant"
],
"abstract": "Background: As the world is busy addressing COVID-19, road traffic injuries, another major cause of deaths is continuously killing people on the roads. In Nepal, there were frequent media reports of occurrences of road crashes, injuries, and deaths despite nationwide lockdown. This paper aims to describe the situation of road traffic crashes and casualties during the period of complete lockdown. Methods: This study used secondary data from two sources: Nepal Police and media reports between 24 March and 14 June 2020 (because the government lifted the nationwide lockdown from 15 June 2020). Available details of crashes, deaths, and injuries for this period were extracted from media reports and the summary data that was obtained from the Police. Narrative comparison is done between the data for the same period from both the sources, where possible. Results: Nepal Police recorded 1,801 incidents of road crashes during the 82 days of the COVID-19 lockdown with 256 deaths (on average 3.1 deaths daily) and 1,824 injuries (on average 22.2 injuries daily). Motorcycles comprised over 21% of all vehicles involved in crashes. Ambulances and other vehicles for essential services were also found to be involved in crashes. Speeding itself was the cause for almost a quarter of the incidents during the lockdown. Conclusions: Even when the movement restrictions were imposed in Nepal, the number of road crashes was not substantially reduced. Media reports were mainly found to be reporting the crashes where deaths occurred, but police records also included nonfatal injuries. The incidence of crashes in this period shows that it is important to work for road safety to save lives from road traffic crashes in Nepal.",
"keywords": [
"COVID-19",
"Lockdown",
"Road Traffic Crashes",
"Injuries",
"Deaths"
],
"content": "Background\n\nThe world experienced a series of unprecedented events since December 2019 after the detection of the novel coronavirus disease 2019 (COVID-19) (Asian Development Bank, 2020). The World Health Organisation (WHO) declared it a worldwide pandemic on 11 March 2020 (World Health Organization, 2020). During this period, social-distancing and lockdown were implemented throughout the world. As of 14 June 2020, the spread of COVID-19 has reached all countries and territories around the globe with 282,733 deaths (Worldometer, 2020)\n\nThe concept of restrained movement and physical distancing is believed to support the breaking of the chain of infection (World Health Organization, 2014) and slowing the spread of the virus by limiting contact with infected people and contaminated surfaces. In many countries, everyone but essential workers have been instructed to stay at home and work from home. Consequently, transportation through all means has reduced in an unprecedented manner. There are also reports of improvement in air quality (Wang et al., 2020) and reduced bed occupancy for road crash trauma in emergency departments (Zhu et al., 2020), which might have enabled health service systems to prepare and cope with a sudden rise in the number of COVID-19 hospitalisation. However, keeping people at home was not an easy job; governments had to impose notices with strict provisions – including fines and potential imprison if their decisions were violated.\n\nNepal has also joined the global practice for the prevention of the spread of COVID-19 and declared a ban on long-distance public travels from 22 March 2020 through the Prime Minister's statement to the nation. \"All international flights coming to Nepal have been suspended effective from March 22 until 31. Effective from March 22, long-distance passenger vehicles will be suspended throughout Nepal for some time. Crowded places like cinema halls have been shut down for the time being.\" Prime Minister KP Sharma Oli, 20 March 2020 (Embassy of Nepal, 2020).\n\nWithin the window of the partial lockdown, an estimated 1.5 million residents left the capital Kathmandu for different parts of the country. Similarly, about half a million migrant workers from India also returned to their homes in the wake of the government’s decision to lock down the country (Pokhrel & Awale, 2020). This resulted in a sudden rise in the use of motorised vehicles during the 21st, 22nd, and 23rd of March. Meanwhile, the second case of COVID-19 was detected on 23 March. Only then did the Government decided to impose countrywide complete lockdown, from 24 March 2020 (Budhathoki, 2020). Hence, the country's efforts and resources converged towards the prevention of coronavirus transmission.\n\nHowever, the government authorised a special pass-permit to use private vehicles and motorcycles in case of an emergency. Only vehicles required for essential services, i.e. ambulances, police, fire service, milk-tankers, water-tankers, and food deliveries, were allowed on the road without the pass. Due to these activities, a sudden decline in vehicular movement was observed in Nepal. Subsequently, a large reduction in the number of crashes and casualty was expected during this lockdown. Unfortunately, there were frequent media reports of road crashes occurrences, injuries, and deaths despite of nationwide lockdown.\n\nThis paper aims to describe the situation of road traffic crashes and the subsequent casualties during the period of COVID-19 lockdown using secondary sources of data. Similarly, it also aims to relate relevant lessons learned from prevention measures for coronavirus to harness prevention from road crashes in the post-lockdown period.\n\n\nMethods\n\nThis study utilised two secondary data sources, i.e. media reports, and published or unpublished police records. Data collection was done in two ways: 1) a daily online search of media reports for vehicle crash incidents on Google, it was done using search terms in Nepali language in order to capture most of the reports across the country. The search terms were on “deaths or injuries” due to “road crash” “car crash” “motorcycle crash” “vehicle crash” “pedestrian hit by” “bicyclist hit by” “ambulance crash” “tractor crash” “truck crashes” or “crashes or collisions occurred during lockdown;” and 2) through contacting the police to obtain road crash records. From both these data sources, only limited number of variables were available. The location of crash, vehicles involved in crashes and their counterparts (animal, people, or objects etc.), resultant number of deaths & injuries, and the age and gender of victims were extracted from media reports into an Excel spreadsheet. The total number crashes, deaths and injuries occurring in districts and provinces were taken from police records. This paper includes the road crashes information for 82 days of the lockdown (24 March to 14 June 2020) from media reports and police records. The exact location and types of vehicles involved in fatal crashes were not available from the Police data, therefore the exact detail of the vehicles and the location of crashes were extracted from the media reporting. With the available information of location, Palika level – (local government unit) for fatal crashes, the cases were nationally mapped. Similarly, comparisons have been made between the data for the same period for both the sources, where needed.\n\n\nResults\n\nAltogether, there were 1,801 incidents of road crashes recorded by the traffic police in 82 days (24 March to 14 June 2020) of lockdown from all seven provinces of Nepal, which included 2,602 vehicles (96% motorized) that claimed 256 lives and led to a further 1,824 injuries (among which 32% were severely injured). However, the media mostly reported fatal crashes, as 200 deaths and 322 injuries were extracted through media reports for the same period. The number of deaths and injuries reported by local media and taken from police records are given in the Underlying data (Sedain & Pant, 2020).\n\nIn this lockdown, no vehicles were allowed to operate without a government-issued pass for essential services. Police records show that in 82 days of full lockdown, an average of 3.1 people died and 22.2 people were injured daily as a result of road crashes. The media reporting of fatal road crashes was 21.8% less than the police record, and very few injuries and vehicle crashes were reported (Table 1).\n\nSource: Nepal Police Province 1, 2, Bagmati, Gandaki, 5, Karnali and Sudurpaschim headquarters, Metropolitan Traffic Police Office and media reports of road crashes reported\n\nKathmandu Valley comprises of three districts, namely Kathmandu, Bhaktapur, and Lalitpur. Nepal Police has not recorded the road traffic deaths separately for these three districts and the records of the crashes were presented for Kathmandu Valley as a whole. Therefore, by including the three districts of Kathmandu Valley, Table 2 displays road traffic deaths from the 12 districts. In the lockdown period, 12 districts comprised more than half (53.4%) of the total deaths in Nepal. The largest number of people were killed in Kathmandu Valley’s roads, followed by Kailali. Furthermore, Province 5 has the highest proportion (20.3%) of road traffic deaths, followed by Bagmati province (13.6%).\n\nSource: Nepal Police Province 1, 2, Bagmati, 5 and Sudurpaschim headquarters record for road crashes incidents.\n\n*Kathmandu Valley comprise three districts (Kathmandu, Bhaktapur and Lalitpur)\n\nThe information on fatal crashes by location extracted from media reports has been visualized (Figure 1) to show the crash-prone areas of Nepal. The visualization has shown that fatal crashes were concentrated more in the middle and lower region of country. Regarding provinces, the fatal crashes were higher in various locations of Bagmati Province and Province 5. The visualization additionally demonstrates that a large number of fatal crashes have occurred in local units in the junction of national highways of different local government units (or Palikas).\n\nSource: Locations of road crashes and fatalities from media reports. Map reproduced with the permission of the Survey Department of Nepal (2020) (Nepal Government Survey Department, 2020).\n\nTable 3 shows the type of vehicles that instigated crashes and their counterparts involved in it. Motorcycles were the most common vehicles involved in fatal crashes, as usual. Among the total vehicles involved in the crashes, more than one-fifth (22.1%) were motorcycles followed by jeeps, tractors, and trucks. Along with other vehicles, 20 ambulances were found to have instigated road crashes which either hit other vehicles, people animals or roadside objects. The majority of the vehicles involved in the crashes were, reportedly went out of driver's control (52.2%) due to speeding. Similarly, 37 pedestrians were hit by the vehicles in these 82 days.\n\nSource: Media reporting of road crashes for the lockdown period (24 March to 14 June, 2020).\n\n\nDiscussion\n\nSince testing for COVID-19 cases commenced in Nepal, 5,760 positive cases (as of 14 June 2020) and 19 deaths have been identified (Worldometer, 2020). On the contrary, 256 deaths and 1,824 injuries from road crashes were recorded between 24 March to 14 June 2020.\n\nRoad traffic injuries are the leading cause of death for the people aged 5–29 years worldwide (World Health Organization, 2018). It is also the leading cause of death and disabilities among people aged between 15 and 49 years in Nepal (Pant et al., 2020). Regardless of the figures, which may vary from source to source, we aim to highlight road safety measures and their importance for the essential-service vehicles during adversity.\n\nThe total burden of road traffic injuries in Nepal is calculated to be approximately 123 million USD, and 90% of this amount comprised of indirect costs (Banstola et al., 2020). Two-wheeled motorized vehicles (motorcycles and scooters) were most frequently involved in crashes and are found to be putting the largest burden on the economy directly and indirectly (Sapkota et al., 2016). Tractors and jeeps were the second-most frequently involved vehicles in road crashes, which is shown by both the police and media records. An incident of injury tends to become a matter of interest to the media even in an adverse situation. Therefore, not all incidents of road crashes are covered by media. From our data, it is also apparent that the fatal cases are consistently reported in police records and media reports but cases of injuries are much less reported by the media.\n\nFrom police records, an average of 154 incidents of road crashes takes place weekly, killing 22 people and causing 156 injuries during the period of lockdown. In the normal situation 7.6 people die in the road crashes (Nepal Police, 2019) and in this lockdown with the minimal transport mobility about 40% (3.1) people die per day in Nepal. Similarly, the ratio of deaths and injuries - a number that has surged from 1 death per 22.8 injuries during the non-lockdown period (Karkee & Lee, 2016), to 1 death per 7.1 injuries in lockdown. Perhaps the injured individuals were involved in more severe crashes during lockdown due to people's tendency to maintain higher speeds on the road. Drivers want to drive their vehicles at high speeds for different reasons (Gabany et al., 1997), and when the roads are empty, speeding might become obvious if there are no measures in place for speed control.\n\nThe casualty data indicate that the burden of road crashes remains high in the lockdown period, a discovery that is different from a popular belief that causality or crashes have decreased substantially. In the absence of evidence-based practice of road safety, people incorrectly assume that reduced vehicular movement automatically reduces the risks of a crash. Given the small number of vehicles in operation, the problem is rather big. The number of deaths and injuries during the lockdown in Nepal would account for an entire year's road deaths for countries and territories such as Fiji, Suriname, Estonia, Montenegro, Saint Lucia, Cyprus and a further 17 countries (World Health Organization, 2018).\n\nThese crash and casualty figures worryingly indicating the magnitude of the problem when regular transportation will eventually resume in Nepal. In rural areas, the use of tractors on unsafe roads increases the risks of crashes. Further, our findings also indicate the lack of a safety culture among the operators of the essential services (including ambulances and the vehicles used by law enforcers). The current focus of the government is to improve roads, but free roads encourage drivers to speed, which is dangerous in terms of road crashes. Therefore, a system of speed monitoring must also be integrated.\n\n\nConclusion\n\nRoadways are the major means of transport in Nepal. In this lockdown, large number of people had to make journeys to their homes by roads; many of those were exposed to the risks of unsafe roads transportation which led to their deaths and injuries. Therefore, this lockdown has reinforced how important the management of safer mobility issue is in Nepal. Interestingly, some of the preventative measures that have been proven effective to decelerate the spread of coronavirus apply in the context of road safety as these measures can teach us something for the road safety epidemic as well (Job, 2020). Therefore, for better road safety, unnecessary travels must be avoided and a safe distance should be maintained between the vehicles on the move. Likewise, the use of helmets, seat-belts, and child restraints are similar to the use of PPE, whereas the regular testing of vehicles, like COVID testing, is a must. In cities, traffic congestion was eased during the lockdown which consequently resulted higher speed, increasing the chances and impact of crashes. Therefore, awareness of safety and taking into account road and weather conditions when deciding to take a journey would help to keep people safe on the roads. The Government of Nepal has mobilised unprecedented amount of resources in terms of human resources, budget and materials to address COVID-19 which has kept the rates of infection and deaths at minimum. If similar efforts and investments are done to address the problem of road traffic injuries, it would be possible to reverse the trend of ever-increasing burden of road injuries.\n\n\nData availability\n\nFigshare: Road Traffic Injuries in Nepal during COVID-19 Lockdown_ Media reporting and Police record (24 March to 14 June, 2020).csv. https://doi.org/10.6084/m9.figshare.12958373.v3 (Sedain & Pant, 2020).\n\nThis project contains the following underlying data:\n\nRoad Traffic Injuries in Nepal during COVID-19 Lockdown_Media reporting (24 March to 14 June, 2020).csv. (Road traffic injuries and deaths reported by local media.)\n\nRoad Traffic Injuries in Nepal during COVID-19 Lockdown_Police records (24 March to 14 June, 2020).xlsx.csv. (Road traffic injuries and deaths taken form police records.)\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Author contributions\n\n\n\nMs. Bhagabati Sedain conceptualised the study and prepared the first draft of the article with suggestions and critical comments from Dr. Puspa Raj Pant. Both the authors approved the final manuscript.\n\n\nAcknowledgments\n\nWe would like to thank Mr. Santosh Sapkota, Metro Traffic FM, Nepal Police Human Resource and Administration Department for their valuable support. Similarly, I would like to thank Mr. Abhasha Joshi of Department of Survey for his technical support in plotting the locations of road crashes on the map of Nepal.\n\n\nReferences\n\nAsian Development Bank: The Economic Impact of the COVID-19 Outbreak on Developing Asia. 2020; 9(128). Publisher Full Text\n\nBanstola A, Kigozi J, Barton P, et al.: Economic burden of road traffic injuries in Nepal. Int J Environ Res Public Health. 2020; 17(12): 4571. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEmbassy of Nepal, B: Address to the nation by right honourable prime minister Mr. KP Sharma Oli on control and prevention of coronavirus. 2020; Retrieved March 26, 2020. Reference Source\n\nGabany GS, Plummer P, Grigg P: Why drivers speed: The speeding perception inventory. J Safety Res.. 1997; 28(1): 29–35. Publisher Full Text\n\nJob S: Can COVID-19 teach us something for the road safety epidemic? 2020. Reference Source\n\nKarkee R, Lee AH: Epidemiology of road traffic injuries in Nepal, 2001-2013: Systematic review and secondary data analysis. BMJ Open. 2016; 6(4): e010757. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBudhathoki A: COVID-19 Imperils Nepal’s high economic ambitions. THE DIPLOMAT. 2020. Reference Source\n\nNepal Government Survey Department: Nepal’s (political and administrative)map, Shapefile (GIS Data). 2020 ; Retrieved June 15, 2020. Reference Source\n\nNepal Police: Police mirror 2019. 2019; Kathmandu. Reference Source\n\nPant PR, Banstola A, Bhatta S, et al.: Burden of injuries in Nepal, 1990 – 2017 : findings from the Global Burden of Disease Study 2017. Inj Prev. 2020; 26(Supp 1): i57–i66. PubMed Abstract | Publisher Full Text\n\nPokhrel M, Awale S: Returnees may be taking coronavirus to rural Nepal. Nepali Times. 2020. Reference Source\n\nSapkota D, Bista B, Adhikari SR: Economic costs associated with motorbike accidents in Kathmandu, Nepal. Front Public Health. 2016; 4: 273. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSedain B, Pant PR: Road Traffic Injuries in Nepal during COVID-19 Lockdown_ Media reporting and Police record (24 March to 14 June, 2020).csv. figshare. Dataset. 2020. http://www.doi.org/10.6084/m9.figshare.12958373.v3\n\nZhu W, Yang J, Xu L, et al.: A plunge in the number of traumatic traffic injuries in an emergency center in Anhui province, China. Am J Emerg Med. 2020; S0735-6757(20)30169-8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang Y, Yuan Y, Wang Q, et al.: Changes in air quality related to the control of coronavirus in China : Implications for traffic and industrial emissions. Sci Total Environ. 2020; 731: 139133. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorld Health Organization: Infection prevention and control of epidemic- and pandemic-prone acute respiratory infections in health care. 2020; Geneva. Reference Source\n\nWorld Health Organization: Global status report on road safety 2018. 2018; Geneva. Reference Source\n\nWorld Health Organization: Coronavirus disease 2019 (COVID-19) Situation Report – 72. 2018; Geneva. Reference Source\n\nWorld Health Organization: Infection prevention and control of epidemic- and pandemic-prone acute respiratory infections in health care. 2014; Geneva. Reference Source\n\nWorldometer: Coronavirus worldwide graphs. 2020; Retrieved June 15, 2020. Reference Source"
}
|
[
{
"id": "72710",
"date": "29 Oct 2020",
"name": "Kulanthayan KC Mani",
"expertise": [
"Reviewer Expertise Injury prevention and safety promotion"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAbstract – For the methods sections, usage of secondary data from Police is fine and acceptable, but not from the media reports. Media reports does not cover all crashes and there could also be duplication of a single case is being reported by various media. Also it’s the reporter and editor view. No one to verify it. Not recommended for scientific article unless no other options.\n\nMethods – is it possible to do away with the media reports and completely rely on the police data (secondary data)? What kind of information we can get from police? If that is sufficient, then go ahead with police data only.\n\nResults. Table 1. The paragraph tend to indicate as though the article is trying to compare two reporting system police vs media which is not the case as I believe you are using media to complement and close the gap for info not available from police source. Intention good, however the mechanism to address it is questionable since its unverified and its individually done by reporters.\n\nTable 2. Purpose of it? My understanding is based on the title, there could be two options of story flow: 1. Before and during covid: RTC and RTI status. Option 2: study RTC and RTI during covid (there must be a purpose and benefit of doing one). Till this stage not clear yet.\n\nFigure 1. Purpose of it? Does it tell us during covid the mapping of RTC and RTI are different compared to non-covid time. Therefore this are the areas to be targeted and focused by the government towards addressing the problem. Is this what you intend to?\n\nTable 3. Purpose of it? Any different on its distribution / patterns from normal day which warrants attention? How best can we use this info towards our next move in terms of strategies or policy change with an aim to reduce RTC and RTI during covid.\n\nDid we learn anything during covid which we can put to use after this which can give us better results in terms of managing RTC and RTI – example physical distancing, reducing exposure, lower ridership in vehicle etc. Effect of covid could be shift in travel mode to private vehicles or single rider vehicles to avoid crowd. Vehicle speed increases as more space with lower vehicle. There could be reductions in RTC, but RTI Fatal may not drop much.\n\nDiscussion para 3. USD123 million is for which year estimates? Interestingly 90% is based on indirect cost. Compared to other country studies, what are their range of % for indirect cost? Is it as high as this 90%?\n\nWhat do the data indicate? RTC and RTI during covid should be compared during normal day within Nepal also. That is the right comparison and not against other countries like Fiji, Estonia, Cyprus etc.\n\nInteresting to note on crashes involving ambulances during covid in comparison to normal day. Are they at more risk? If yes, what need to be done next? Increase in tractor usage does it adds to more crash? Any police data to show increase? Generally tractor as low speed vehicle pose less harm. Its only risk is when it shares the roadway with mixed traffic and the other traffic speed is very high and huge difference with the tractor speed.\n\nConclusion – the content in conclusion is very much different compared to the text inside the article. Many aspects not touched in the article is being concluded here example of using PPE, safety helmet, seat belts and CRS. Prefer conclusion strictly based on this study findings.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": [
{
"c_id": "6145",
"date": "14 Dec 2020",
"name": "Bhagabati Sedain",
"role": "Author Response",
"response": "1.Abstract – For the methods sections, usage of secondary data from Police is fine and acceptable, but not from the media reports. Media reports does not cover all crashes and there could also be duplication of a single case is being reported by various media. Also, it’s the reporter and editor view. No one to verify it. Not recommended for scientific article unless no other options. Author response: Thank you for the concerns. We were aware that media reports do not cover all crashes, particularly those with non-fatal outcomes. During the period of lockdown and minimum vehicular movement, Nepali media covered 78% of all fatal incidents compared to police records (Table 1). We extracted detailed information for the individual crashes from media reports. A typical Nepali media report of a road crash, it reports – location of the crash; vehicle type and registration number; gender and age of victims (with names); outcomes of the crash. The details provided in the report obviously helped to see any duplication of the records. Such information was not possible to get from police records; the police source only provides compiled reports of incidents. We have revised the last sentence under \"Methods\" in the abstract. 2.Methods – is it possible to do away with the media reports and completely rely on the police data (secondary data)? What kind of information we can get from police? If that is sufficient, then go ahead with police data only. Author response: Using police data only will limit the information and we will not be able to make Figure 1 and Table 3, which is derived from media report data. We have amended the last sentence of the section Methods as – \"In this study, we have presented the results from the two data sources to provide the most possible details complemented by one-another.\" 3.Results. Table 1. The paragraph tends to indicate as though the article is trying to compare two reporting system police vs media which is not the case as I believe you are using media to complement and close the gap for info not available from police source. Intention good, however the mechanism to address it is questionable since its unverified and its individually done by reporters. Author response: Thank you for the comment. We have revised our statement and removed the term 'compare' throughout the text because these two data sources complement one another. However, we have discussed the differences in the variables obtained from the two sources. 4.Table 2. Purpose of it? My understanding is based on the title, there could be two options of story flow: 1. Before and during covid: RTC and RTI status. Option 2: study RTC and RTI during covid (there must be a purpose and benefit of doing one). Till this stage not clear yet. Author response: Thank you for the suggestion. We described the data as \"option 2\" in your comment; the purpose was to show the districts that were the most affected. We have now revised the title of table 2 \"Table 2. Districts with the highest number of road traffic deaths in Nepal during national level COVID-19 lockdown.\" We have replaced \"Province 5\" with its official name, Lumbini which is given after we submitted the manuscript. 5.Figure 1. Purpose of it? Does it tell us during covid the mapping of RTC and RTI are different compared to non-covid time? Therefore, this are the areas to be targeted and focused by the government towards addressing the problem. Is this what you intend to? Author response: The map (figure 1) shows the locations of road crashes that occurred during the lockdown. Therefore, it was not possible to compare the numbers before and during the lockdown. This information is derived from news reports, as the police data used in this study does not tell the location of crashes. We believe this map is a novel idea to show the locations of fatal road crashes and this would help relevant authorities to focus interventions. 6.Table 3. Purpose of it? Any different on its distribution / patterns from normal day which warrants attention? How best can we use this info towards our next move in terms of strategies or policy change with an aim to reduce RTC and RTI during covid. Author response: 1) the purpose of table 3 was to convey to the readers what type of vehicles were involved in road crashes; and the counterpart in case it was a collision. This information was extracted from media reports because this information was not available from police data used in this paper. We have made slight changes to the title of Table 3 as \"Distribution of road crashes during national level covid-19 lockdown according to the types of vehicles involved and their counterparts.\" The key message here is – Motorcyclists, pedestrians and cyclists are the most at-risk road users from motorcyclists and the vehicles at essential services i.e. jeep/cars, tractors, trucks, and ambulances. Therefore, the enforcement of road safety laws must be of priority always. 7.Did we learn anything during covid which we can put to use after this which can give us better results in terms of managing RTC and RTI – example physical distancing, reducing exposure, lower ridership in vehicle etc. Effect of covid could be shift in travel mode to private vehicles or single rider vehicles to avoid crowd. Vehicle speed increases as more space with lower vehicle. There could be reductions in RTC, but RTI Fatal may not drop much. Author response: It is beyond the scope of this study, we only studied the period of lockdown where all non-essential travel was prohibited. 8.Discussion para 3. USD123 million is for which year estimates? Interestingly 90% is based on indirect cost. Compared to other country studies, what are their range of % for indirect cost? Is it as high as this 90%? Author response: Thanks for your comment. As mentioned in the reference cited in this paper, higher indirect costs are also seen in other countries. We have revised the sentences with additional reference: \"The total burden of road traffic injuries in Nepal is calculated to be approximately 123 million USD for 2017, and 90% of this amount comprised of indirect costs ( Banstola et al., 2020). The amount of indirect cost of road traffic injuries ranges from 51% in Iran to 90% in Nepal (Rezaei, Arab, Karami, & Akbari Sari, 2014; Banstola et al., 2020). This road crashes deaths toll demonstrated the economic impact of road crashes is bigger in Nepal.\" 9.What do the data indicate? RTC and RTI during covid should be compared during normal day within Nepal also. That is the right comparison and not against other countries like Fiji, Estonia, Cyprus etc. Author response: Thank you very much for the comment. We have revised this section by adding the following text - \"This increased number of deaths and injuries during the lockdown in Nepal can be related to higher speed due to lower traffic volume and limited law enforcement. High speed means higher impact if there is a crash. It has also reported elsewhere that speed law violations and failing to stop due to high speeding were increased during lockdown (Inada et al, 2020).\" 10.Interesting to note on crashes involving ambulances during covid in comparison to normal day. Are they at more risk? If yes, what need to be done next? Increase in tractor usage does it adds to more crash? Any police data to show increase? Generally, tractor as low speed vehicles pose less harm. Its only risk is when it shares the roadway with mixed traffic and the other traffic speed is very high and huge difference with the tractor speed. Author response: Thank you very much for the comment. a) Road crashes involving Ambulances are something that is not explored yet. Classification and standardization of ambulances in Nepal are yet evolving. Ambulances are a very important part of the post-crash response (pillar 5) aspect of road safety. Ambulances in Police records are not separately recorded and collectively included with other four-wheeled light vehicles (under jeep cars). Therefore, we cannot say whether there was an increase in ambulance movement or crashes during the COVID-19 lockdown. b) the use of tractors was increased for transportation of relief-aid materials (food, quarantine construction materials, etc) during the COVID-19 lockdown. It is true that the mix of vehicles with different speeds caused crashes; also, tractors from rural areas and hilly tracks were also reported. 11.Conclusion – the content in conclusion is very much different compared to the text inside the article. Many aspects not touched in the article is being concluded here example of using PPE, safety helmet, seat belts and CRS. Prefer conclusion strictly based on this study findings. Author response: Thank you very much for the comment and suggestions. We have now revised the conclusion section. Roadways are the major means of transport in Nepal. Just before this lockdown, a large number of people had to make journeys to their homes (mainly going out of Kathmandu and coming to Nepal from bordering India) by roads; many of those were exposed to the risks of unsafe roads transportation which led to their deaths and injuries. Our study found that deaths on Nepali roads was not stopped during lock down. Comparing the pattern of road crashes during the same period last year, lock down witnessed almost half of the number of incidents (1,801 vs 3,480) and the number of vehicles involved in crashes (2,602 vs 5,560). When comparing the statistics with the situation with the three months before the lock down, it was observed that the percentages of tractors, trucks/tankers and cyclists was higher (published police records). Therefore, this lockdown has reinforced how important the management of safer mobility issue is in Nepal. Interestingly, some of the preventative measures that have been proven effective to decelerate the spread of coronavirus apply in the context of road safety as these measures can teach us something for the road safety epidemic as well (Job, 2020). Therefore, for better road safety, unnecessary travels must be avoided and a safe distance should be maintained between the vehicles on the move. Likewise, the use of helmets, seat-belts, and child restraints are similar to the use of PPE, whereas the regular testing of vehicles, like COVID testing, is a must. In cities, traffic congestion was eased during the lockdown which consequently resulted higher speed, increasing the chances and impact of crashes. Therefore, awareness of safety and taking into account road and weather conditions when deciding to take a journey would help to keep people safe on the roads. Therefore, this lockdown has reinforced how important the management of safer mobility issue is in Nepal. Interestingly, some of the preventative measures that have been proven effective to decelerate the spread of coronavirus apply in the context of road safety as these measures can teach us something for the road safety epidemic as well (Job, 2020). The Government of Nepal has mobilised unprecedented amount of resources in terms of human resources, budget and materials to address COVID-19 which has kept the rates of infection and deaths at minimum. If similar efforts and investments are done to address the problem of road traffic injuries, it would be possible to reverse the trend of ever-increasing burden of road injuries."
}
]
},
{
"id": "72705",
"date": "10 Nov 2020",
"name": "Felix Wilhelm Siebert",
"expertise": [
"Reviewer Expertise Traffic psychology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study presents secondary source road traffic injury and fatality data from Nepal for the time frame of the recent COVID-19 related lockdown between March and June 2020. I understood the study as follows: The authors used Nepal Traffic Police data to extract numbers of fatalities and injuries (distinguished between heavy and non-heavy injuries) for the timeframe between March and June 2020. In addition, news articles were analyzed and reports on injuries and fatalities were collected for the same timeframe. Both datasets are then compared, and the authors find that not all fatal crashes are reported on, while even less of a percentage of crashes that “only” result in injuries are reported in the news. The authors list potential factors related to accident causes and hypothesize that speeding might play a larger role in crashes during the lockdown, as the road system is less crowded. The article is very easy to read and all data is presented in a concise and clear way. I have two major comments before going through the manuscript in detail.\nComparison to pre-lockdown data Before reading the manuscript, I expected that the authors would compare injury and fatality data during the lockdown to data collected outside of the lockdown period, e.g. to the same days in 2019. This would give an indication if the lockdown has led to a general increase in fatalities and injuries (this would be expected). The authors report non-lockdown data in the discussion section of the article (p.5 “In the normal situation 7.6 people die in the road crashes […]“) but I would have expected this comparison sooner, i.e. in the results section. In addition, the comparison of the ratio of injuries to deaths should be presented sooner than in the discussion. The ratio named for the non-lockdown period (“Similarly, the ratio of deaths and injuries - a number that has surged from 1 death per 22.8 injuries during the non-lockdown period (Karkee & Lee, 2016), to 1 death per 7.1 injuries in lockdown.”) is old, as the data used in the Karkee & Lee (2016) paper is from 2013. Furthermore, it’s unclear if the correct data from the Karkee and Lee paper is used to arrive at the death/injury ratio. It appears that the ratio of 1 death per 22.8 injuries is calculated from Table 1 in the cited article, but erroneously only injury and fatality data for Kathmandu is used. Calculating the death/injury ratio from the correct column (listing the “total” number for the last year of available data) results in a death/injury ratio of 6.6. Hence, I advise that the authors either make it more clear which data is used for the calculation of the ratio, or (preferably) calculate the more current ratio for the whole year 2019 and for the exact same time frame as the 2020 lockdown, using 2019 data.\nTimely distribution of fatalities and injuries within the lockdown time frame The authors do a great job in describing outside factors that have influenced traffic density during the lockdown, i.e. especially before the lockdown there seems to have been higher traffic as people left for their home villages / home countries. While reading, I was wondering if this potential uneven distribution of traffic within the observed time frame would be visible in the distribution of injuries and fatalities, e.g. more injuries and fatalities in the beginning of the lockdown than in the middle/end. This could be visualized easily by plotting the percentage of registered injuries and fatalities over the lockdown time frame. This could also be presented alongside the same data for the same time frame in 2019.\nOther comments:\nThe title omits the analysis of fatalities during the lockdown, the authors should consider revising the title “Road traffic injuries and fatalities during […]”\n\nI understand the urge to contrast numbers of road related fatalities with COVID-19 numbers, but they are very different things, mainly as they have completely different potential for exponential increases. Hence, I would suggest the authors revise the first sentence of the abstract to make road safety and COVID less contrastive.\n\nThe results section of the abstract should present comparative data to the same time frame from a non-lockdown period.\n\nThe conclusion part of the abstract is not supported by the results part of the abstract. The sentence “Even when the movement restrictions were imposed in Nepal, the number of road crashes was not substantially reduced.” needs to be supported by additional analyses in the manuscript, which will need to be mentioned in the results part of the abstract.\n\n“Within the window of the partial lockdown” please list clear time frames (i.e. dates).\n\n“Meanwhile, the second case of COVID-19 was detected on 23 March.” when was the first case detected? The authors should consider adding a timeline of COVID-related events in Nepal (e.g. First case, partial lockdown, lockdown, opening up, analyzed time frame).\n\nDo the authors have any information on the number of special “pass-permits” handed out during the lockdown? This could be related to the number of registered vehicles in Nepal, to get an idea of how much the traffic was reduced through the lockdown (hypothetically).\n\nIn the discussion section, I would suggest the authors do not directly contrast COVID-19 and road related fatalities (same argument as before). It might be enough to change “On the contrary […]” to a different wording.\n\nThe second and third section in the discussion (“Road traffic injuries are the leading cause of death for […]”) are quite general and do not directly relate to the results found in the study. I think they are better suited to frame road safety challenges (global and in Nepal), and would be better placed in the introduction.\n\nFor the sentence “Similarly, the ratio of deaths and injuries - a number that has surged from 1 death per 22.8 injuries during the non-lockdown period (Karkee & Lee, 2016), to 1 death per 7.1 injuries in lockdown.” Please revise for grammar. (and please also see my first major comment above).\n\nThe section “What do the data indicate? The casualty data indicate that the burden of road crashes remains high in the lockdown period […]” is not well supported by the results section of the paper (see my first major comment).\n\nIn the conclusion, the authors write “Roadways are the major means of transport in Nepal. In this lockdown, large number of people had to make journeys to their homes by roads […]”. But the authors earlier state, that the main movement of people happened during partial lockdown, i.e. before the time frame analyzed in this study. One could assume that during the early days of lockdown, some people used their vehicles illegally (without the special pass-permit), but this would need to be stated more clearly. The authors should clear this up (see also my second major comment).\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "6146",
"date": "14 Dec 2020",
"name": "Bhagabati Sedain",
"role": "Author Response",
"response": "Comparison to pre-lockdown data Before reading the manuscript, I expected that the authors would compare injury and fatality data during the lockdown to data collected outside of the lockdown period, e.g. to the same days in 2019. This would give an indication if the lockdown has led to a general increase in fatalities and injuries (this would be expected). The authors report non-lockdown data in the discussion section of the article (p.5 “In the normal situation 7.6 people die in the road crashes […]“) but I would have expected this comparison sooner, i.e. in the results section. In addition, the comparison of the ratio of injuries to deaths should be presented sooner than in the discussion. The ratio named for the non-lockdown period (“Similarly, the ratio of deaths and injuries - a number that has surged from 1 death per 22.8 injuries during the non-lockdown period (Karkee & Lee, 2016), to 1 death per 7.1 injuries in lockdown.”) is old, as the data used in the Karkee & Lee (2016) paper is from 2013. Furthermore, it’s unclear if the correct data from the Karkee and Lee paper is used to arrive at the death/injury ratio. It appears that the ratio of 1 death per 22.8 injuries is calculated from Table 1 in the cited article, but erroneously only injury and fatality data for Kathmandu is used. Calculating the death/injury ratio from the correct column (listing the “total” number for the last year of available data) results in a death/injury ratio of 6.6. Hence, I advise that the authors either make it more clear which data is used for the calculation of the ratio, or (preferably) calculate the more current ratio for the whole year 2019 and for the exact same time frame as the 2020 lockdown, using 2019 data. Authors' response: Thanks for cogent observations. We have addressed them as below: compare injury and fatality data during the lockdown to data collected outside of the lockdown period, e.g. to the same days in 2019. used them accordingly. We have now replaced the terms relating to 'comparison' throughout the text. The ratio of deaths and injuries for lockdown and non-lockdown period Authors' response: Thank you very much for spotting the error. We appreciate your advice for using more recent figures. We compared the figures with those during the same period last year (i.e. 3 months of lockdown compared with the same 3 months last year). we have revised this statement as: (“Similarly, the ratio of deaths and injuries has surged (in Kathmandu) from 1 death per 46.3 injuries during the non-lockdown period (Police Records), to 1 death per 20.6 injuries during the lockdown.” A new sentence is added \"Nationally, the number of casualty (death plus injuries) per 100 involved vehicles were slightly higher (79.9) during lockdown compared to the same period last year (76.5).\" Timely distribution of fatalities and injuries within the lockdown time frame The authors do a great job in describing outside factors that have influenced traffic density during the lockdown, i.e. especially before the lockdown there seems to have been higher traffic as people left for their home villages / home countries. While reading, I was wondering if this potential uneven distribution of traffic within the observed time frame would be visible in the distribution of injuries and fatalities, e.g. more injuries and fatalities in the beginning of the lockdown than in the middle/end. This could be visualized easily by plotting the percentage of registered injuries and fatalities over the lockdown time frame. This could also be presented alongside the same data for the same time frame in 2019. Other comments: The title omits the analysis of fatalities during the lockdown, the authors should consider revising the title “Road traffic injuries and fatalities during […]” Authors' Response: Thank you very much for your suggestion. We think the title is ok. I understand the urge to contrast numbers of road related fatalities with COVID-19 numbers, but they are very different things, mainly as they have completely different potential for exponential increases. Hence, I would suggest the authors revise the first sentence of the abstract to make road safety and COVID less contrastive. Authors' response: Thanks for your suggestion but we think this statement is reasonable and we would like to keep it as is. The results section of the abstract should present comparative data to the same time frame from a non-lockdown period. Authors' response: We agree that adding a figure with the information during non-lock down would enhance the text. Figure 3 has been added to the discussion section and cited in the fourth paragraph (p5). \"Comparing the police data for lockdown and the same period last year shows a considerable reduction in the number of incidents, involved vehicles and casualties (Figure 3). Given the lower number of vehicles allowed to operate the figures are high.\" The conclusion part of the abstract is not supported by the results part of the abstract. The sentence “Even when the movement restrictions were imposed in Nepal, the number of road crashes was not substantially reduced.” needs to be supported by additional analyses in the manuscript, which will need to be mentioned in the results part of the abstract. Authors Response: We have revised the conclusion, it was also suggested by reviewer 1 (please see the conclusion section in the text). “Within the window of the partial lockdown” please list clear time frames (i.e. dates). Authors Response: 21 march to 23 march “Meanwhile, the second case of COVID-19 was detected on 23 March.” when was the first case detected? The authors should consider adding a timeline of COVID-related events in Nepal (e.g. First case, partial lockdown, lockdown, opening up, analyzed time frame). Authors' response: The first case was confirmed on 13th January and we have added a figure (Figure 1) in the introduction Do the authors have any information on the number of special “pass-permits” handed out during the lockdown? This could be related to the number of registered vehicles in Nepal, to get an idea of how much the traffic was reduced through the lockdown (hypothetically). Authors' response: unfortunately, no information available in this regard In the discussion section, I would suggest the authors do not directly contrast COVID-19 and road related fatalities (same argument as before). It might be enough to change “On the contrary […]” to a different wording. Authors' response: we have changed \"on the contrary\", as \"meanwhile\" The second and third section in the discussion (“Road traffic injuries are the leading cause of death for […]”) are quite general and do not directly relate to the results found in the study. I think they are better suited to frame road safety challenges (global and in Nepal), and would be better placed in the introduction. Authors' response: Thanks for the suggestion, but we think these two statements are setting the context for discussion. For the sentence “Similarly, the ratio of deaths and injuries - a number that has surged from 1 death per 22.8 injuries during the non-lockdown period (Karkee & Lee, 2016), to 1 death per 7.1 injuries in lockdown.” Please revise for grammar. (and please also see my first major comment above). Authors' response: we have responded to it above The section “What do the data indicate? The casualty data indicate that the burden of road crashes remains high in the lockdown period […]” is not well supported by the results section of the paper (see my first major comment). Authors' response: Thanks for the suggestion, we have now revised the sentence: \"The casualty data indicate that the burden of road crashes remains high in the lockdown period despite of very much restricted vehicular movement\" In the conclusion, the authors write “Roadways are the major means of transport in Nepal. In this lockdown, large number of people had to make journeys to their homes by roads […]”. But the authors earlier state, that the main movement of people happened during the partial lockdown, i.e. before the time frame analyzed in this study. One could assume that during the early days of lockdown, some people used their vehicles illegally (without the special pass-permit), but this would need to be stated more clearly. The authors should clear this up (see also my second major comment). Authors' response: Thanks for the suggestion, we have now revised the sentence as “Roadways are the major means of transport in Nepal. Just before this lockdown, a large number of people had to make journeys to their homes (mainly going out of Kathmandu and coming to Nepal from bordering India) by roads immediately after the announcement of the lockdown”."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1209
|
https://f1000research.com/articles/9-1027/v1
|
24 Aug 20
|
{
"type": "Brief Report",
"title": "Journal editors: How do their editing incomes compare?",
"authors": [
"Janice C. L. Lee",
"Jennifer Watt",
"Diane Kelsall",
"Sharon E. Straus",
"Janice C. L. Lee",
"Jennifer Watt",
"Diane Kelsall"
],
"abstract": "Background: The work of journal editors is essential to producing high-quality literature, and editing can be a very rewarding career; however, the profession may not be immune to gender pay gaps found in many professions and industries, including academia and clinical medicine. Our study aimed to quantify remuneration for journal editors from core clinical journals, determine if a gender pay gap exists, and assess if there are remuneration differences across publishing models and journal characteristics. Methods: We completed an online survey of journal editors with substantial editing roles including section editors and editors-in-chief, identified from the Abridged Index Medicus “Core Clinical” journals in MEDLINE. We analyzed information on demographics, editing income, and journal characteristics using a multivariable partial proportional odds model for ordinal logistic regression. Results: There were 166 survey respondents (response rate of 9%), which represented editors from 69 of 111 journals (62%). A total of 140 fully completed surveys were analyzed (95 males and 45 females); 50 (36%) editors did not receive remuneration for editorial work. No gender pay gap and no difference in remuneration between editors who worked in subscription-based publishing vs. open access journals were detected. Editors who were not primarily health care providers were more likely to have higher editing incomes (adjusted odds ratio [OR] 2.96, 95% confidence interval [CI] 1.18-7.46). Editors who worked more than 10 hours per week editing earned more than those who worked 10 hours or less per week (adjusted OR 16.7, 95%CI 7.02-39.76). Conclusions: We were unable to detect a gender pay gap and a difference in remuneration between editors who worked in subscription-based publishing and those in open access journals. More than one third of editors surveyed from core clinical journals did not get remunerated for their editing work.",
"keywords": [
"Journal",
"Medical Journalism",
"Peer review",
"Publishing",
"Income"
],
"content": "Introduction\n\nThe number of academic journals continues to grow each year. In 2018, there were 5399 clinical journals tracked by Journal Citation Reports in comparison to only 3681 10 years prior1. A rise in open-access journals is also evident; while in 2008 there were just 249 open-access journal titles in Web of Science, that number ballooned to 1431 in 20181. Growth in the journal industry comes with more opportunities to become a journal editor. Editors are often scientists, researchers, administrators, and clinicians who have expertise in a particular field. Editorial work can be a very satisfying part-time job or a full-time career. However, there is currently very limited literature that evaluates how journal editors are compensated for their work. An international email survey of 88 editors of nursing journals found that their mean annual salary was $12,749 USD (ranging from $0 to $56,000) for a mean of 13.4 hours worked per week2. A total of 8% of survey respondents (7 of the 88 editors) did not receive any monetary compensation and only 31% of participants felt that their compensation was adequate2. It is concerning that the critical job of editors to uphold the integrity of academic literature can be low-paid or voluntary.\n\nVarious factors may influence remuneration for journal editors; of interest in the recent news are gender pay gaps and different journal publishing platforms. Substantial evidence of gender inequity in academia exists, including disparities in compensation3–5. For example, a 2012 US survey found that male physician researchers had higher average salaries ($13,399 USD; p=0.001) than females after adjusting for specialty, academic rank, and research productivity6. Salary deficits can be up to 12-28% for female physicians7,8 when compared to male counterparts.\n\nIn April 2018, the United Kingdom led the way for national pay transparency and equity by mandating annual reporting of gender pay gap for any organization comprising more than 250 employees9. France followed suit with mandatory reporting, while Germany and Iceland enacted pay equity and transparency laws. With this mandatory data reporting, The Lancet reported in 2018 that major publishing companies in the UK had gender pay gaps of 13-40% favouring men over women10. Specific data for journal editors are not available but some of these data do highlight the potential gaps.\n\nThere is substantial heterogeneity among journals, such as publishing platform, scope, publication frequency, and Journal Impact Factor11. Though open access publishing comes at a large cost to authors, its popularity is supported by many benefits that may include faster publishing times and the ability to reach bigger audiences compared to subscription-based publishing. It is unknown whether a journal editor’s remuneration is affected by these journal variables.\n\nThe objectives of this study were to quantify remuneration for journal editors from core clinical journals, determine if a gender pay gap exists, and assess if there are remuneration differences across publishing models (e.g., subscription-based or open-access) and other journal characteristics (e.g., publication frequency and Journal Impact Factor).\n\n\nMethods\n\nWe completed an international online survey of full-time and part-time journal editors identified from the Abridged Index Medicus “Core Clinical” journals in MEDLINE, which represented 111 peer-reviewed core medical journals. Our target population of journal editors were those with substantial editing roles including editors-in-chief, deputy editors, executive editors, senior editors, associate editors, and editors of a specialty section. Editors were identified through each journal’s webpage in December 2018. We excluded roles classified as statistical editors, assistant editors, international editors and editorial board members. If the editors of a section or specialty section numbered greater than 200 for a single journal, these individuals were excluded based on the assumption that they may not have a substantial editorial role. Publicly available emails were found via the English-language search engine, Google. Major sources included journal web pages, academic institutional web pages, and corresponding author on recently published articles. Reporting of this online survey was guided by the CHERRIES reporting guideline12,13.\n\nThe online survey was conducted using Qualtrics CoreXM Survey Tool software14, but an open access alternative such as SurveyMonkey15 could be used for replication of methods. We developed the survey to capture demographic data, editing remuneration in USD, editing experience, and journal characteristics. A blank copy of the survey is available as Extended data13. An online pilot test was sent out to our knowledge user team of three journal editors to identify poorly constructed questions, and to assess face validity before distribution. The survey included 10 questions on three pages, and adaptive questioning was used. Respondents were able to review their answers before submission. The following variables were collected in our survey and included in the final multivariable model: sex (male vs. female), gender identity, publishing model (subscription-based, vs. open or hybrid [open access option or open access for developing countries]), primary role (health care provider vs. other), academic rank (any professorship vs. none), 2017 Journal Impact Factor, frequency of publication (monthly/bimonthly vs. weekly/biweekly), editing role (section/specialty/associate editor vs. editor-in-chief/executive/senior/deputy editor/other), years in editing (>10 vs. ≤10 years), hours/week in editing role (>10 vs. ≤10 hours), and years worked for current journal (>5 vs. ≤5 years) .\n\nEmail invitations with the survey link were distributed via Qualtrics in February 2019. This was a voluntary survey, and no incentives were offered. We employed established methods to enhance survey completion rates with reminder emails at week 2 and week 416.\n\nOnly completed surveys were analyzed. Dichotomous baseline characteristics for male and female journal editors were presented as frequencies and percentages and compared with the chi-square test or Fisher’s exact test when expected sample sizes were 5 or less. Journal Impact Factor of the journal where male and female editors worked was a non-normally distributed continuous variable presented as a median with interquartile range. The Journal Impact Factor of each group was compared with the Wilcoxon rank sum test.\n\nOur outcome of interest was journal editor salary, which was modeled as an ordinal variable with three categories: ≤$10,000 per year, $10,001 to $50,000 per year, and ≥$50,000 per year. We derived adjusted odds ratios (OR) and 95% confidence intervals (CI) for our outcome of interest from a multivariable partial proportional odds model for ordinal logistic regression in SAS version 9.4 (SAS Institute, Cary, North Carolina). The variables for editor sex and academic appointment did not satisfy the proportional odds assumption; therefore, these variables were assumed to have nonproportional odds in the final multivariable partial proportional odds model. Two-sided p-values were reported and p-values <0.05 were considered statistically significant.\n\nInstitutional review board approval was obtained through the Unity Health Toronto Research Ethics Board in Toronto, Canada. The survey landing page included study information and consent to participate was implied by survey completion and submission.\n\n\nResults\n\nA total of 2165 editors were identified in December 2018; of these, survey invitations were successfully sent to 1844 with publicly available email addresses. A total of 193 surveys were started, and 166 surveys were submitted (overall response rate of 9%). A de-identified version of the dataset is available as Underlying data13. We received survey responses from journal editors at 69 of 111 journals (62%). Only the 140 fully completed surveys (defined as those without any responses of “prefer not to answer”) were included in our analyses (Figure 1). This was composed of 95 male and 45 female editors. All respondents identified as cis gender. A total of 50 (36%) editors did not receive remuneration for editorial work; 111 (79%) and 29 (21%) editors worked for a subscription-based journal versus an open/hybrid journal, respectively. A total of 90% of editors held an academic position. The median 2017 Journal Impact Factor was 4.9 (Interquartile range [IQR] 3.5-6.6 for males and 3.5-7.5 for females). There was a larger proportion of female survey respondents who were fulltime journal editors compared to male counterparts (10 females [22%] vs. 4 males [4%] p=<0.01). Details on baseline characteristics are provided in Table 1.\n\nUSD = United States dollar. IQR = Interquartile range. * chi-square test. **Fisher’s exact test. ***Mann-Whitney U test.\n\nIn univariate analyses (Table 2), journal editors received more remuneration if the journal was open/hybrid access rather than subscription-based (OR 3.06, 95% CI 1.39-6.75), if the editor was not primarily a health care provider (OR 3.67, 95% CI 1.7-7.91), if the issues were weekly/biweekly rather than monthly/bimonthly (OR 2.07, 95% CI 1-4.28), if the editors held senior editing positions (OR 4.5, 95% CI 2.16-9.59), if the editors spent more than 10 hours per week editing (OR 16.7, 95% CI 7.02-39.76), and if the editors had worked for more than 5 years at the journal (OR 2.28, 95% CI 1.13-4.62).\n\nUSD = United States dollar. IQR = Interquartile range. CI = Confidence interval. OR = odds ratio. *Adjusted OR for all characteristics in the table.\n\nIn multivariable analysis (Table 2), there was no gender pay gap detected or remuneration differences identified between publishing models. We found that editors who were not primarily health care providers were more likely to have higher editing incomes (adjusted OR 2.96, 95%CI 1.18-7.46). Editors who worked more than 10 hours per week editing earned more than those who worked 10 hours or less per week (adjusted OR 16.7, 95%CI 7.02-39.76).\n\n\nDiscussion\n\nIn addition to a recent study by The Lancet highlighting the presence of a gender pay gap of employees at major academic publishing companies10, numerous studies in the past have highlighted similar inequities across academia3–8. In our study of part-time and full-time journal editors from a focused subset of core clinical journals, we did not detect a gender pay gap. We found that annual remuneration for editing was higher for editors working more than 10 hours per week and if their primary occupation was not a health care provider. Full-time journal editing positions were disproportionately more likely to be held by a woman. While we hypothesized that the different revenue structure of subscription-based versus open access journals may translate to a difference in editor’s remuneration, we did not find it to be true after adjusting for predictor variables. A journal’s Impact Factor did not affect an editor’s remuneration. Although 36% of editors surveyed reported no direct earnings from their editorial work, there can be other benefits including subsidies for scientific meeting registration fees and travel costs.\n\nOur study had several limitations. First, our results had limited generalizability due to a low response rate. However, our participants did represent 69% of the core clinical journals. Moreover, our sample size of 140 was comparable to an international survey of 148 scientific editors from biomedical journals to evaluate journal editing core competencies17. We were unfortunately not powered to compare subgroups of editing roles and journal characteristics. Second, given our approach to identifying email addresses for journal editors, our survey was biased towards editors who had either an English-language academic profile on an institutional website or if they were a corresponding author of a recently published article. Third, we were unable to make direct comparisons between for-profit and non-profit journals, because several companies publish journals on behalf of non-profit organizations. This distinction may factor into a journal editor’s desire to edit for a journal at a given remuneration rate. Lastly, there was substantial heterogeneity in editing roles and the lack of standardization in editing titles made it difficult to draw conclusions that were generalizable to the population. Future studies can include qualitative interviews to gauge the various roles and responsibilities of editors, and remuneration practices.\n\n\nConclusion\n\nWe conducted an international survey of journal editors from core clinical journals to understand how remuneration varied across editor’s demographics, professional experience, and journal characteristics. We did not detect a gender pay gap or a difference in remuneration between editors who worked in subscription-based publishing vs. open access journals. More than a third of editors surveyed were not remunerated for their work.\n\n\nData availability\n\nHarvard Dataverse: Journal editors: How do their editing incomes compare? https://doi.org/10.7910/DVN/AHMB8G13.\n\nFile ‘Journal editors deidentified data.tab’ contains a de-identified version of the dataset generated in this study.\n\nDue to the nature of this research, data provided is a limited de-identified dataset without potentially identifying information, i.e. clinical specialty (if applicable) and journal characteristics.\n\nIndividual(s) wishing for access to the full de-identified dataset requires written support from the principal investigator (Dr. Sharon Straus, the corresponding author) to be added to the study as study personnel(s), and subsequently needs approval from the Unity Health Toronto Research Ethics Board in Toronto, Canada.\n\nHarvard Dataverse: Journal editors: How do their editing incomes compare? https://doi.org/10.7910/DVN/AHMB8G13.\n\nFile ‘Journal Editors Supplementary 2 - Survey.docx’ contains a copy of the survey used in this study.\n\nHarvard Dataverse: CHERRIES checklist for ‘Journal editors: How do their editing incomes compare?’. https://doi.org/10.7910/DVN/AHMB8G/XPMGJL13.\n\nData are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).",
"appendix": "References\n\nInCites Journal Citation Reports. Clarivate Analytics. 2019. Reference Source\n\nFreda MC, Kearney MH: A first look at nurse editors' compensation. Nurs Econ. 2007; 25(6): 371–275. PubMed Abstract\n\nMascarenhas A, Moore JE, Tricco AC, et al.: Perceptions and experiences of a gender gap at a Canadian research institute and potential strategies to mitigate this gap: a sequential mixed-methods study. CMAJ Open. 2017; 5(1): E144–51. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaulus JK, Switkowski KM, Allison GM, et al.: Where is the leak in the pipeline? Investigating gender differences in academic promotion at an academic medical centre. Perspect Med Educ. 2016; 5(2): 125–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReed DA, Enders F, Lindor R, et al.: Gender differences in academic productivity and leadership appointments of physicians throughout academic careers. Acad Med. 2011; 86(1): 43–7. PubMed Abstract | Publisher Full Text\n\nJagsi R, Griffith KA, Stewart A, et al.: Gender differences in the salaries of physician researchers. JAMA. 2012; 307(22): 2410–7. PubMed Abstract | Publisher Full Text\n\nAsgari MM, Carr PL, Bates CK: Closing the Gender Wage Gap and Achieving Professional Equity in Medicine. JAMA. 2019; 321(17): 1665–1666. PubMed Abstract | Publisher Full Text\n\nHenderson MT, Fijalkowski N, Wang SK, et al.: Gender differences in compensation in academic medicine: The results from four neurological specialties within the University of California Healthcare System. Scientometrics. 2014; 100: 297–306. Publisher Full Text\n\nGender pay gap reporting: overview. 2017. Reference Source\n\nThe Lancet: Closing the gender pay gap: when and how? Lancet. 2018; 391(10129): 1455. PubMed Abstract | Publisher Full Text\n\n2017 Journal Impact Factor. Clarivate Analytics. 2018.\n\nEysenbach G: Improving the quality of web surveys: The Checklist for Reporting Results of Internet E-Surveys (CHERRIES). J Med Internet Res. 2004; 6(3): e34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee J, Watt J, Kelsall D, et al.: Journal editors: How do their editing incomes compare? [data set]. Harvard dataverse V1; 2020 [accessed Aug 8, 2020]. http://www.doi.org/10.7910/DVN/AHMB8G\n\nCoreXM of Qualtrics. Provo, UT, USA. 2019. Reference Source\n\nSurveyMonkey Inc. San Mateo, California, USA. 2020. Reference Source\n\nDillman DA, Smyth JD, Christian LM: Internet, phone, mail, and mixed-mode surveys: The tailored design method. 4th ed. Hoboken, NJ, US: John Wiley & Sons Inc. 2014. Reference Source\n\nGalipeau J, Cobey K, Barbour V, et al.: An international survey and modified Delphi process revealed editors’ perceptions, training needs, and ratings of competency-related statements for the development of core competencies for scientific editors of biomedical journals [version 1; peer review: 2 approved] . F1000Res. 2017; 6: 1634. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "72508",
"date": "27 Oct 2020",
"name": "Kelly Cobey",
"expertise": [
"Reviewer Expertise Journalology",
"publication science",
"open science"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study aimed to:\nquantify remuneration for journal editors from core clinical journals\n\ndetermine if a gender pay gap exists\n\nassess if there are remuneration differences across publishing models and journal characteristics.\nThe authors addressed these aims through conducting an online survey of journal editors identified from the Abridged Index Medicus “Core Clinical” journals in MEDLINE. The paper is reported clearly and acknowledges relevant limitations. I describe a few suggestions for changes below:\nThe authors specify:\n“The following variables were collected in our survey and included in the final multivariable model: sex (male vs. female), gender identity, publishing model (subscription-based, vs. open or hybrid [open access option or open access for developing countries]), primary role (health care provider vs. other), academic rank (any professorship vs. none), 2017 Journal Impact Factor, frequency of publication (monthly/ bimonthly vs. weekly/biweekly), editing role (section/specialty/ associate editor vs. editor-in-chief/executive/senior/deputy editor/other), years in editing (>10 vs. ≤10 years), hours/week in editing role (>10 vs. ≤10 hours), and years worked for current journal (>5 vs. ≤5 years) .” When I review the survey provided by the authors, I don’t see any items asking about the 2017 journal impact factor or journal frequency of publication. Perhaps I missed something? If not, this leads me to think that perhaps the authors extracted this information online based on participants responses indicating which journal they are an editor at, or that the incorrect survey has been uploaded. This should be clarified in the paper.\nFurther, the authors indicate the survey captured the frequency at which the included editors’ journals publish using the options: monthly/ bimonthly vs. weekly/biweekly. This is unlikely to apply uniformly across journals. For illustration, I don’t know how the F1000 journal editor would accurately respond to this item.\n\nDid the authors record bounce-backs among editors who they e-mailed their invitation to? In comparable work I have conducted this has been a fairly significant issue, especially when editor information is found via Google rather than through the journal website. If so, what was the N and could this be reflected in the study recruitment description and figure? The response rate may be somewhat higher if this is considered.\n\nMinor comments:\nGiven that this is not a sub-specialty journal, the authors could consider briefly addressing what the role and responsibilities of journal editors are in the introduction.\n\nThe authors note: “The online survey was conducted using Qualtrics CoreXM Survey Tool software14, but an open access alternative such as SurveyMonkey15 could be used for replication of methods. ; I don’t think the free SurveyMonkey would be appropriate for replicating the survey described, I understand it allows for only 10 questions (each sub question ‘counts’) and 100 responses.\n\nThe authors could provide rational for the journal sampling approach.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6262",
"date": "18 Jan 2021",
"name": "Janice CL Lee",
"role": "Author Response",
"response": "Re: Extracting journal information--Yes, revised to state that journal information was extracted online based on journal selection in the survey. Re: Journal frequency—Journal frequency was extracted from journals and rounded up/down to fit the categories. Re: Bounce-back emails—Yes, revised to clarify that there were 104 bounce-backs. The response rate was calculated based on the number of successful emails sent. Re: Minor comment on role of journal editors—Thank you, this has been added to the introduction. Re: Minor comment on free survey software—Thank you, this has been removed. To our knowledge, there are no easily accessible free survey software for survey replication. Re: Minor comment on the rational for journal sampling—Revised to state that we chose the core clinical journal for their high clinical impact on medicine."
}
]
},
{
"id": "75952",
"date": "11 Dec 2020",
"name": "Carl Singleton",
"expertise": [
"Reviewer Expertise Labour Economics",
"Applied Econometrics",
"including published works on gender in the workplace and the gender pay gap."
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study aimed to quantify gender differences in pay for editing services within academia, specifically for “Core Clinical” journals. It also aimed to test for the presence of a gender pay gap, conditional on other observable (surveyed) characteristics of editors and their journals.\nOverall, the article is well-written, the methods are clearly explained, and some of the limitations are fairly outlined.\nCobey (2020) has already provided a thoughtful review of this article, and I would echo those comments.\nIn addition, I would like to make the following further suggestions:\nIn the introduction and the limitations, the authors may wish to dwell on the other reasons that editors choose to do what they do, besides direct financial gain (e.g., enjoyment, prestige, compensation from their main employer in the form of relief from teaching or other activities). These motivations are all potentially endogenous factors (omitted variables) in the regression model. While in theory they may not be correlated with gender in the population of editors, the sample size here is small and the response to the survey probably non-random.\n\nThe literature cited for the gender pay gap is in academia is limited and very specific, focused on medical sciences. But there are good examples and literature reviews of gender pay gaps in academia more widely. One recent example I am aware of is Mumford & Sechel (2020)1 in the British Journal of Industrial Relations. While that study focuses on the gender pay gap among economists, it also provides some prominent examples of the wider literature on the gender gaps in academia, including in Medicine and Science (e.g., Connolly and Holdcroft, 20092). The authors could look to reference more of this previous work, particularly since this journal is not field-specific.\n\nThe authors state that “Only completed surveys were analyzed”. Elsewhere in the article it is mentioned that responses where editors “preferred not to say” were discarded. Having looked at the survey (Lee et al., 2020), there are some questions asked that are not being used in the analysis. Therefore, the article could be clearer on what information is being thrown away. For instance, for the unadjusted odds ratios, I am not sure why an editor preferring not to say what the frequency of their publication is should exclude them from the sample when estimating the raw gender pay gap. Overall, it is important that the authors explain why they don’t consider a larger sample for their statistics just because of some non-response to some survey questions.\n\nLike the previous comment, I am also concerned about the amount of information lost in the chosen modelling strategy. The survey collects information on much narrower bands of pay (and hours). While I appreciate that in this field odds models are common and often preferred, in my own field, as an applied economist/econometrician, we would prefer not to lose all the information contained in the pay data collected. For example, I would suggest also estimating a censored linear least squares regression model (i.e., tobit, censored because of the zeroes), where the mid-point of each and every range in the survey is imputed as the pay value, which becomes the dependent variable (normally in natural logs). Similarly, combining the narrow pay bands and narrower hours bands in the survey could allow the authors to look at a measure of hourly pay, which is what matters most economically, being the compensation for a unit of labour input (see Mumford and Smith, 20073, for a relevant example of extracting more information from ordinal pay and hours data in this suggested way). Results following this alternative empirical strategy could be presented as a complement to those already shown in the article.\n\nIn the current ordinal logistic regression model approach, I would be curious what interacting gender with some the other covariates shows. For example, the estimation sample shows a large difference between the male and female in the proportion of editors whose main occupation is Health Care Provider. This variable looks as though it could be worth interacting with gender. This could be added as a third column of results in Table 2.\nMinor comment:\nThe Introduction claims that the UK led the wag on gender pay gap reporting legislation. Denmark and Austria have legislation requiring firms to make their pay gaps available to their workers, which I believe substantially pre-date the UK policy (2006 in Denmark; see Bennedsen et al., 20194).\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "6263",
"date": "18 Jan 2021",
"name": "Janice CL Lee",
"role": "Author Response",
"response": "Re: #1) Thank you. Revised to address this limitation in both the introduction and limitations sections. Re: #2) Thank you. This was added to the limitations section. Mumford & Sechels 2020 and Connolly and Hodcroft 2009 were cited. Re: #3) Thank you. This was clarified in the manuscript. Of the 166 surveys, 140 were analyzed, excluding surveys that had incomplete responses. Responses of “prefer not to answer” were analyzed but often had small cell counts (n<5), and therefore the data were not reported to preserve privacy of participants. Re: #4) We did not perform the tobit regression, since we did not have exact salaries and cannot make an assumption regarding the mid-point values for the last category of salaries greater than $150,000 USD. In analyzing the zero/nonzero pay by male/female to explore censoring from the zeros, we found the following: chi-squared 1.35 (p=0.25) 2x2 table: Zero salary, male: 37 Zero salary, female: 13 Nonzero salary, male: 58 Nonzero salary, female: 32 This analysis has not been included in the article. Re: #5) We presented the data based on male/female as primary exposure and did not examine for possible interaction of gender with covariates. We are unable to return to the data at this time to update the analysis to explore this possibility. Re: Minor #1) Thank you. This has been revised to add Denmark and Austria’s reporting legislation."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1027
|
https://f1000research.com/articles/8-2138/v1
|
23 Dec 19
|
{
"type": "Research Article",
"title": "Benchmarking of long-read assemblers for prokaryote whole genome sequencing",
"authors": [
"Ryan R. Wick",
"Kathryn E. Holt",
"Kathryn E. Holt"
],
"abstract": "Background: Data sets from long-read sequencing platforms (Oxford Nanopore Technologies and Pacific Biosciences) allow for most prokaryote genomes to be completely assembled – one contig per chromosome or plasmid. However, the high per-read error rate of long-read sequencing necessitates different approaches to assembly than those used for short-read sequencing. Multiple assembly tools (assemblers) exist, which use a variety of algorithms for long-read assembly. Methods: We used 500 simulated read sets and 120 real read sets to assess the performance of six long-read assemblers (Canu, Flye, Miniasm/Minipolish, Raven, Redbean and Shasta) across a wide variety of genomes and read parameters. Assemblies were assessed on their structural accuracy/completeness, sequence identity, contig circularisation and computational resources used. Results: Canu v1.9 produced moderately reliable assemblies but had the longest runtimes of all assemblers tested. Flye v2.6 was more reliable and did particularly well with plasmid assembly. Miniasm/Minipolish v0.3 was the only assembler which consistently produced clean contig circularisation. Raven v0.0.5 was the most reliable for chromosome assembly, though it did not perform well on small plasmids and had circularisation issues. Redbean v2.5 and Shasta v0.3.0 were computationally efficient but more likely to produce incomplete assemblies. Conclusions: Of the assemblers tested, Flye, Miniasm/Minipolish and Raven performed best overall. However, no single tool performed well on all metrics, highlighting the need for continued development on long-read assembly algorithms.",
"keywords": [
"Assembly",
"long-read sequencing",
"Oxford Nanopore Technologies",
"Pacific Biosciences",
"microbial genomics",
"benchmarking"
],
"content": "Introduction\n\nGenome assembly is the computational process of using shotgun whole-genome sequencing data (reads) to reconstruct an organism’s true genomic sequence to the greatest extent possible1. Software tools which carry out assembly (assemblers) take sequencing reads as input and produce reconstructed contiguous pieces of the genome (contigs) as output.\n\nIf a genome contains repetitive sequences (repeats) which are longer than the sequencing reads, then the underlying genome cannot be fully reconstructed without additional information; i.e. if no read spans a repeat in the genome, then that repeat cannot be resolved, limiting contig length2. Short-read sequencing platforms (e.g. those made by Illumina) produce reads hundreds of bases in length and tend to result in shorter contigs. In contrast, long-read platforms from Oxford Nanopore Technologies (ONT) and Pacific Biosciences (PacBio) can generate reads tens of thousands of bases in length which span more repeats and thus result in longer contigs3.\n\nProkaryote genomes are simpler than eukaryote genomes in a few aspects relevant to assembly. First, they are smaller, most being less than 10 Mbp in size4. Second, they contain less repetitive content and their longest repeat sequences are often less than 10 kbp in length5. Third, prokaryote genomes are haploid and thus avoid assembly-related complications from diploidy/polyploidy6. These facts make prokaryote genome assembly a more tractable problem than eukaryote genome assembly, and in most cases a long-read set of sufficient depth should contain enough information to generate a complete assembly – each replicon in the genome being fully assembled into a single contig7. Prokaryote genomes also have two other features relevant to assembly: they may contain plasmids that differ from the chromosome in copy number and therefore read depth, and most prokaryote replicons are circular with no defined start/end point.\n\nIn this study, we examine the performance of various long-read assemblers in the context of prokaryote whole genomes. We assessed each tool on its ability to generate complete assemblies using both simulated and real read sets. We also investigated prokaryote-specific aspects of assembly, such as performance on plasmids and the circularisation of contigs.\n\n\nMethods\n\nSimulated read sets (read sequences generated in silico from reference genomes) offer some advantages over real read sets when assessing assemblers. They allow for a confident ground truth – i.e. the true underlying genome is known with certainty. They allow for large sample sizes, in practice limited only by computational resources. Also, a variety of genomes and read set parameters can be used to examine assembler performance over a wide range of scenarios. For this study, we simulated 500 read sets to test the assemblers, each using different parameters and a different prokaryote genome.\n\nTo select reference genomes for the simulated read sets, we first downloaded all bacterial and archaeal RefSeq genomes using ncbi-genome-download v0.2.10 (14333 genomes at the time of download)8. We then performed some quality control steps: excluding genomes with a >10 Mbp chromosome, a <500 kbp chromosome, any >300 kbp plasmid, any plasmid >25% of the chromosome size or more than 9 plasmids (Extended data, Figure S1)9. We then ran Assembly Dereplicator v0.1.0 with a threshold of 0.1, resulting in 3153 unique genomes10.\n\nTo produce a final set of 500 genomes with 500 plasmids, we randomly selected 250 genomes from those containing plasmids, repeating this selection until the genomes contained exactly 500 plasmids. We then added 250 genomes randomly selected from those without plasmids. Any ambiguous bases in the assemblies were replaced with ‘A’ to ensure that sequences contained only the four canonical DNA bases.\n\nWe then used Badread v0.1.5 to generate one read set for each input genome11. The parameters for each set (controlling read depth, length, identity and errors) were randomly chosen to ensure a large amount of variability (Extended data, Figure S2)9. Note that not all of these read sets were sufficient to reconstruct the original genome (due to low depth or short read length), so even an ideal assembler would be incapable of completing an assembly for all 500 test sets.\n\nFor genomes containing plasmids, the read depth of plasmids relative to the chromosome was also set randomly, with limits based on the plasmid size (Extended data, Figure S3)9. Large plasmids were simulated at depths close to that of the chromosome while small plasmids spanned a wider range of depth. This was done to model the observed pattern that small plasmids often have a high per-cell copy number (i.e. may be high read depth) but can be biased against in library preparations (i.e. may be low read depth)12. All replicons (chromosomes and plasmids) were treated as circular sequences in Badread, so the simulated read sets do not test assembler performance on linear sequences.\n\nDespite the advantages of simulated read sets, they can be unrealistic because read simulation tools (such as Badread) may not accurately model all relevant features: error profiles, read lengths, quality scores, etc. Real read sets are therefore also valuable when assessing assemblers. The challenge with real read sets is obtaining a ground truth genome against which assemblies can be checked. Since many reference genome sequences are produced using long-read assemblies, there is the risk of circular reasoning – if we use an assembly as our ground truth reference, our results will be biased in favour of whichever assembler produced the reference.\n\nTo avoid this issue, we used the datasets produced in a recent study comparing ONT and PacBio data which also included Illumina reads for each isolate13. For each of the 20 bacterial isolates in that study, we conducted two hybrid assemblies using Unicycler v0.4.7: Illumina+ONT and Illumina+PacBio14. Unicycler works by first generating an assembly graph using the Illumina reads, then using long-read alignments to scaffold the graph’s contigs into a completed genome – a distinct approach from any of the long-read assemblers tested in this study. We ran the assemblies using Unicycler’s --no_miniasm option so it skipped its Miniasm-based step which could bias the results in favour of Miniasm/Minipolish. We then excluded any isolate where either hybrid assembly failed to reach completion or where there were structural differences between the two assemblies as determined by a Minimap2 alignment15. This left six isolates for inclusion.\n\nThe ONT and PacBio read sets for these isolates were quite deep (156× to 535×) so to increase the number of assembly tests, we produced ten random read subsets of each, ranging from 40× to 100× read depth. This resulted in 120 total read sets for testing the assemblers (6 genomes × 2 platforms × 10 read subsets). The Illumina+ONT hybrid assembly was used as ground truth for each isolate.\n\nAll real and simulated read sets16 and reference genomes17 are available as Underlying data.\n\nWe assembled each of the read sets using the current versions of six long-read assemblers: Canu v1.9, Flye v2.6, Miniasm/Minipolish v0.3, Raven v0.0.5, Redbean v2.5 and Shasta v0.3.0. Default parameters were used except where stated, and exact commands for each tool are given in the Extended data, Figure S49. Assemblers that only work on PacBio reads (i.e. not on ONT reads) were excluded (HGAP18, FALCON19, HINGE20 and Dazzler21), as were hybrid assemblers which also require short read input (Unicycler14 and MaSuRCA22).\n\nCanu has the longest history of all the assemblers tested, with its first release dating back to 2015. It performs assembly by first correcting reads, then trimming reads (removing adapters and breaking chimeras) and finally assembling reads into contigs23. Its assembly strategy uses a modified version of the string graph algorithm24, sometimes referred to as the overlap-layout-consensus (OLC) approach.\n\nFlye takes a different approach to assembly: first combining reads into error-prone disjointigs, then collapsing repetitive sequences to make a repeat graph and finally resolving the graph’s repeats to make the final contigs25. Of particular note to prokaryote assemblies, Flye has options for recovery of small plasmids (--plasmids) and uneven depth of coverage (--meta), both of which we used in this analysis.\n\nMiniasm builds a string graph from a set of read overlaps – i.e. it performs only the layout step of OLC. It does not perform read overlapping which must be done separately with Minimap2, and it does not have a consensus step, so its assembly error rates are comparable to raw read error rates. A separate polishing tool such as Racon is therefore required to achieve high sequence identity26. For this study, we developed a tool called Minipolish to simplify this process by conducting Racon polishing (two rounds by default) on a Miniasm assembly graph. To ensure clean circularisation of prokaryote replicons, circular contigs are ‘rotated’ (have their starting position adjusted) between rounds. Minipolish also comes with a script (miniasm_and_minipolish.sh) which carries out all assembly steps (Minimap2 overlapping, Miniasm assembly and Minipolish consensus) in a single command, and subsequent references to ‘Miniasm/Minipolish’ refer to this entire pipeline.\n\nRaven (previously known as Ra) is another tool which takes an OLC approach to assembly27. Its overlapping step shares algorithms with Minimap2, and its consensus step is based on Racon, making it similar to Miniasm/Minipolish. It differs in its layout step which includes novel approaches to remove spurious overlaps from the graph, helping to improve assembly contiguity.\n\nRedbean (previously known as Wtdbg2) uses an approach to long-read assembly called a fuzzy Bruijn graph28. This is modelled on the De Bruijn graph concept widely used for short-read assembly29 but modified to work with the inexact sequence matches present in noisy long reads.\n\nShasta is an assembler designed for computational efficiency30. To achieve this, much of its assembly pipeline is performed not directly on read sequences but rather on a reduced representation of marker k-mers. These markers are used to find overlaps and build an assembly graph from which a consensus sequence is derived.\n\nAll assemblies were run on Ubuntu 18.04 instances of Australia’s Nectar Research Cloud which contained 32 vCPUs and 64 GB of RAM (m3.xxlarge flavour). To guard against performance variation caused by vCPU overcommit, the assemblers were limited to 16 threads (half the number of available vCPUs) in their options. Any assembly which exceeded 24 hours of runtime or 64 GB of memory usage was terminated.\n\nOur primary metric of assembly quality was contiguity, defined here as the longest single Minimap2 alignment between the assembly and the reference replicon, relative to the reference replicon length. Contiguity of exactly 100% indicates that the replicon was assembled completely with no missing or extra sequence (Extended data, Figure S5A)9. Contiguity of slightly less than 100% (e.g. 99.9%) indicates that the assembly was complete, but some bases were lost at the start/end of the contig (Extended data, Figure S5B)9. Contiguity of more than 100% (e.g. 101%) indicates that the contig contains duplicated sequence via start-end overlap (Extended data, Figure S5C)9. Much lower contiguity (e.g. 70%) indicates that the assembly was not complete due to fragmentation (Extended data, Figure S5D)9, missing sequence (Extended data, Figure S5E)9 or misassembly (Extended data, Figure S5F)9. Contiguity values were determined by aligning the contigs to a tripled version of the reference replicon, necessary to ensure that contigs can fully align even with start-end overlap and regardless of their starting position relative to that of the linearised reference sequence (Extended data, Figure S6)9.\n\nContiguity values were determined for each replicon in the assemblies – e.g. if a genome contained two plasmids, then the assemblies of that genome have three contiguity values: one for the chromosome and one for each plasmid. A status of ‘fully complete’ was assigned to assemblies where all replicons (the chromosome and any plasmids if present) achieved a contiguity of ≥99%. If an assembly had a chromosome with a contiguity of ≥99% but incomplete plasmids, it was given a status of ‘complete chromosome’. If the chromosome had a contiguity of <99%, the assembly was deemed ‘incomplete’. If the assembly was empty or missing (possibly due to the assembler prematurely terminating with an error), it was given a status of ‘empty’. If the assembly terminated due to exhausting the available RAM, it was given a status of ‘out of memory’. Computational metrics were also observed for each assembly: time to complete and maximum RAM usage.\n\n\nResults and discussion\n\nFigure 1 and Figure 2 summarise the assembly results for the simulated and real read sets, respectively. Full tabulated results can be found in the Extended data9. The assemblies, times and terminal outputs generated by each assembler are available as Underlying data31.\n\n(A) Proportion of each possible assembly outcome. (B) Relative contiguity of the chromosome for each assembly, showing cleanliness of circularisation. (C) Sequence identity of each assembly’s longest alignment to the chromosome. (D) Total time taken (wall time) for each assembly. (E) Maximum RAM usage for each assembly. ‘Miniasm+’ here refers to the entire Miniasm/Minipolish assembly pipeline.\n\n(A) Proportion of each possible assembly outcome. (B) Relative contiguity of the chromosome for each assembly, showing cleanliness of circularisation. (C) Sequence identity of each assembly’s longest alignment to the chromosome. (D) Total time taken (wall time) for each assembly. (E) Maximum RAM usage for each assembly. ‘Miniasm+’ here refers to the entire Miniasm/Minipolish assembly pipeline.\n\nFigure 1A/Figure 2A shows the proportion of read sets with each assembly status. For the real read sets, a higher proportion of completed assemblies indicates a more reliable assembler – one which is likely to make a completed assembly given a typical set of input reads. For the simulated read sets, a higher proportion of completed assemblies indicates a more robust assembler – one which is able to tolerate a wide range of input read parameters. Extended data, Figure S79 plots assembly contiguity against specific read set parameters to give a more detailed assessment of robustness. Plasmid assembly status, plotted with plasmid length and read depth, is shown in Extended data, Figure S8 and Figure S99 for the simulated and real read sets, respectively.\n\nFigure 1B/Figure 2B shows the chromosome contiguity values for each assembly, focusing on the range near 100%. These plots show how well assemblers can circularise contigs – i.e. whether sequence is duplicated or missing at the contig start/end (Extended data, Figure S5)9. The closer contiguity is to 100% the better, with exactly 100% indicating perfect circularisation. Plasmid contiguity values are shown in Extended data, Figure S109.\n\nAssembly identity (consensus identity) is a measure of the base-level accuracy of an assembled contig relative to the reference sequence (how few substitution and small indel errors are present) and is shown in Figure 1C/Figure 2C. The identity of assembled sequences is almost always higher than the identity of individual reads because errors can be ‘averaged out’ using read depth, producing more accurate consensus base calls. However, systematic read errors (e.g. mistakes in homopolymer length) can make perfect sequence identity difficult to achieve, regardless of assembly strategy32.\n\nAssembler resource usage is shown in terms of total runtime (Figure 1D/Figure 2D) and the maximum RAM usage during assembly (Figure 1E/Figure 2E).\n\nWhen considering only the chromosome, Raven was the most reliable assembler, closely followed by Flye – both were able to complete the chromosome in over three-quarters of the real read sets (Figure 2A). If plasmids are also considered, then Flye was the most reliable assembler. Miniasm/Minipolish and Canu were moderately reliable, completing over half of the real read set chromosomes. Redbean and Shasta were the least reliable and completed less than half of the chromosomes.\n\nFlye, Miniasm/Minipolish and Raven were the most robust assemblers, able to complete over half of the assemblies attempted with the simulated read sets (Figure 1A). Flye and Redbean performed best in cases of low read depth, able to complete assemblies down to ~10× depth (Extended data, Figure S7A)9. Raven performed the best with low-identity read sets (Extended data, Figure S7B)9. The assemblers performed similarly with regards to read length, except for Shasta which required longer reads (Extended data, Figure S7C)9. The assemblers were similarly unaffected by random reads, junk reads, chimeric reads or adapter sequences (Extended data, Figure S7D–F)9. Read glitches (local breaks in continuity) were well-tolerated by the assemblers except for Redbean and Shasta (Extended data, Figure S7G)9.\n\nIn our real read tests, Canu achieved high sequence identity on PacBio reads, Miniasm/Minipolish and Raven did well on ONT reads, and Flye did well on both platforms (Figure 2C). For each assembler, real PacBio reads resulted in higher identities than real ONT reads. For the simulated reads (which contain artificial error profiles), results were more erratic, with Canu, Miniasm/Minipolish and Raven performing best (Figure 1C).\n\nThe nature of read errors depends on the sequencing platform and basecalling software used, so these results may not hold true for all read sets. Post-assembly polishing tools (including Racon26, Nanopolish7, Medaka33 and Arrow34) are routinely used to improve the accuracy of long-read assemblies35, and identity can be further increased by polishing with Illumina reads where available (e.g. with Pilon36). Therefore, the sequence identity produced by the assembler itself is potentially unimportant for many users.\n\nCanu was the slowest assembler tested on both real (Figure 2D) and simulated (Figure 1D) read sets, sometimes taking hours to complete. Its runtime was correlated with read accuracy and read set size, with low-accuracy and large read sets being more likely to result in a long runtime.\n\nFlye was typically faster than Canu, taking less than 15 minutes for the real read sets and usually less than an hour for the simulated read sets. It sometimes took multiple hours to assemble simulated read sets, and this was correlated with the amount of junk (low-complexity) reads, suggesting that removal of such reads via pre-assembly QC may be beneficial. Flye had the highest RAM usage of the tested assemblers and occasionally hit our 64 GB limit for simulated read sets. Its RAM usage was correlated with read N50 and read set size, with long and large read sets being more likely to result in high RAM usage.\n\nMiniasm/Minipolish, Raven and Redbean were comparable in performance, typically completing assemblies in less than 10 minutes and with less than 16 GB of RAM. While not tested in this study, Racon (which is used in Minipolish) and Raven can be run with GPU acceleration to further improve speed performance. Shasta was the fastest assembler and had the lowest memory usage.\n\nOf all assemblers tested, Miniasm/Minipolish was the only one to regularly achieve exact circularisation (contiguity=100%), due to Minipolish’s polishing pipeline (Figure 1B/Figure 2B). Flye often excluded a small amount of sequence (tens of bases) from the start/end of circular contigs (contiguity <100%), and Raven typically excluded moderate amounts of sequence (hundreds of bases). Canu’s contiguities usually exceeded 100%, indicating a large amount (thousands of bases) of start/end overlap. The amount of overlap in a Canu assembly was correlated with the read N50 length (Extended data, Figure S7C)9. Redbean and Shasta were both erratic in their circularisation, often producing some sequence duplication (contiguity >100%) but occasionally dropping sequence (contiguity <100%).\n\nIn addition to cleanly circularising contig sequences, it is valuable for a prokaryote genome assembler to clearly distinguish between circular and linear contigs. This can provide users with a clue as to whether or not the genome was assembled to completion. Flye, Miniasm/Minipolish and Shasta produce graph files of their final assembly which can indicate circularity. Canu indicates circularity via the ‘suggestCircular’ text in its contig headers. Raven and Redbean do not signal to users whether a contig is circular.\n\nCanu and Flye were the two assemblers most able to assemble plasmids at a broad range of size and depth (Extended data, Figures S8, S9)9. Miniasm/Minipolish also performed well, though it failed to assemble plasmids if they were very small or had a very high read depth. Raven was able to assemble most large plasmids but not small plasmids. Redbean and Shasta were least successful at plasmid assembly.\n\nCircularisation of plasmids followed the same pattern as for chromosomes, with only Miniasm/Minipolish consistently achieving clean circularisation (Extended data, Figure S10)9. For smaller plasmids, start/end overlap could sometimes result in contiguities of ∼200% – i.e. the plasmid sequence was duplicated in a single contig. This was most common with Canu, though it occurred with other assemblers as well.\n\nAll assemblers tested were relatively easy to use, either running with a single command (Canu, Flye, Raven and Shasta) or providing a convenience script to bundle the commands together (Miniasm/Minipolish and Redbean). All were able to take long reads in FASTQ format as input, with the exception of Shasta which required reads to first be converted to FASTA format (Extended data, Figure S4)9. We encountered no difficulty installing any of the tools by following the instructions provided.\n\nSome of the assemblers needed a predicted genome size as input (Canu, Flye and Redbean) while others (Miniasm/Minipolish, Raven and Shasta) did not. This requirement could be a nuisance when assembling unknown isolates, as it may be hard to specify a genome size before the species is known.\n\nWhile we ran our assemblies using default and/or recommended commands (Extended data, Figure S4)9, some of the assemblers have parameters which can be used to alter their behaviour. Raven was the least configurable assembler tested, with few options available to users. Flye offers some parameters, including overlap and coverage thresholds. Miniasm/Minipolish, Redbean and Shasta all offer more options, and Canu is the most configurable with hundreds of adjustable parameters. Many of the available parameters are arcane (e.g. Miniasm’s ‘max and min overlap drop ratio’ or Shasta’s ‘pruneIterationCount’), and only experienced power users are likely to adjust them – most will likely stick with default settings or only adjust easier-to-understand options. However, the presence of low-level parameters provides an opportunity to experiment and gain greater control over assemblies and are therefore appreciated even when unlikely to be used.\n\nAnother aspect worth noting is whether an assembler produces useful files other than its final assembly. Canu stands out in this respect, as it creates corrected and trimmed reads in its pipeline which have low error rates and are mostly free of adapters and chimeric sequences. Canu can therefore be considered not just an assembler but also a long-read correction tool suitable for use in other analyses.\n\nCanu v1.9 was the slowest assembler and not the most reliable or robust. Its strength is in its configurability, so power users who are willing to learn Canu’s nuances may find that they can tune it to fit their needs. However, it is probably not the best choice for users wanting a quick and simple prokaryote genome assembly.\n\nFlye v2.6 was an overall strong performer in our tests: reliable, robust and good with plasmids. However, it requires a genome size parameter, tended to delete some sequence (usually on the order of tens of bases) when circularising contigs and could be excessive in its RAM usage when assembling simulated read sets.\n\nMiniasm/Minipolish v0.3 was not the most reliable assembler but was fairly robust to read set parameters. Its main strength is that it was the only assembler to consistently achieve perfect contig circularisation (as this is a specific goal of its polishing step). Also, it does not require a genome size parameter to run, which makes it easier to run than Canu, Flye or Redbean for unknown genomes.\n\nRaven v0.0.5 was the most reliable and robust assembler for chromosome assembly. However, it suffered from worse circularisation problems than Flye (often deleting hundreds of bases) and wasn’t good with small plasmids. Like Miniasm/Minipolish, it does not require a genome size parameter.\n\nRedbean v2.5 assemblies tended to have glitches in the sequence which caused breaks in contiguity, making it perform poorly in both reliability and robustness. This, combined with its erratic circularisation performance and requirement to specify genome size, make it a less-than ideal choice for long-read prokaryote read sets.\n\nShasta v0.3.0 was the fastest assembler tested and used the least RAM, but it had the worst reliability and robustness. It is therefore more suited to assembly of large genomes in resource-limited settings (the use case for which it was designed) than it is for prokaryote genome assembly.\n\n\nConclusions\n\nEach of the different assemblers has pros and cons, and while no single assembler emerged as an ideal choice for prokaryote genome long-read assembly, the overall best performers were Flye, Miniasm/Minipolish and Raven. Flye was very reliable, especially for plasmid assembly, and was the best performing assembler at low read depths. Miniasm/Minipolish was the only assembler to reliably achieve clean contig circularisation. Raven was the most reliable for chromosome assembly and the most tolerant of low-identity read sets.\n\nFor users looking to achieve an optimal assembly, we recommend trying multiple different tools and comparing the results. This will provide the opportunity for validation – confidence in an assembly is greater when it is in agreement with other independent assemblies. It also offers a chance to detect and repair circularisation issues, as different assemblers are likely to give different contig start/end positions for a circular replicon.\n\nAn ideal prokaryotic long-read assembler would reliably complete assemblies, be robust against read set problems, be easy to use, have low computational requirements, cleanly circularise contigs and assemble plasmids of any size. The importance of long-read assembly will continue to grow as long-read sequencing becomes more commonplace in microbial genomics, and so development of assemblers towards this ideal is crucial.\n\n\nData availability\n\nFigshare: Read sets. https://doi.org/10.26180/5df6f5d06cf0416.\n\nThese files contain the input read sets (both simulated and real) for assembly.\n\nFigshare: Reference genomes. https://doi.org/10.26180/5df6e99ff3eed17.\n\nThis file contains the reference genomes against which the long-read assemblies were compared. For the simulated read sets, these genomes were the source sequence from which the reads were generated.\n\nFigshare: Assemblies. https://doi.org/10.26180/5df6e2864a65831.\n\nThese files contain assemblies (in FASTA format), times and terminal outputs for each of the assemblers.\n\nZenodo: Long-read-assembler-comparison. https://doi.org/10.5281/zenodo.27024429.\n\nThis project contains the following extended data:\n\nResults (tables of results data, (including information on eachreference genome, read set parameters and metrics foreach assembly).\n\nScripts (scripts used to generate plots).\n\nFigure S1. Distributions of chromosome sizes (A), plasmid sizes (B) and per-genome plasmid counts (C) for the reference genomes used to make the simulated read sets.\n\nFigure S2. Badread parameter histograms for the simulated read sets. (A) Mean read depths were sampled from a uniform distribution ranging from 5× to 200×. (B) mean read lengths were sampled from a uniform distribution ranging from 100 to 20000 bp. C: read length standard deviations were sampled from a uniform distribution ranging from 100 to twice that set’s mean length (up to 40000 bp). D: mean read identities were sampled from a uniform distribution ranging from 80% to 99%. (E) Max read identities were sampled from a uniform distribution ranging from that set’s mean identity plus 1% to 100%. (F) Read identity standard deviations were sampled from a uniform distribution ranging from 1% to the max identity minus the mean identity. (G, H and I) Junk, random and chimera rates were all sampled from an exponential distribution with a mean of 2%. (J) Glitch sizes/skips were sampled from a uniform distribution ranging from 0 to 100. (K) Glitch rates for each set were calculated from the size/skip according to this formula: 100000/1.6986s/10. (L) Adapter lengths were sampled from an exponential distribution with a mean of 50.\n\nFigure S3. Top: the target simulated depth of each replicon relative to the chromosome. The smaller the plasmid, the wider the range of possible depths. Bottom: the absolute read set of each replicon after read simulation.\n\nFigure S4. Commands used for each of the six assemblers tested.\n\nFigure S5. Possible states for the assembly of a circular replicon. Reference sequences are shown in the inner circles in black and aligned contig sequences are shown in the outer circles in colour (red at the contig start to violet at the contig end). (A) Complete assembly with perfect circularisation. (B) Complete assembly but with missing bases leading to a gapped circularisation. (C) Complete assembly but with duplicated bases leading to overlapping circularisation. (D) Incomplete assembly due to fragmentation (multiple contigs per replicon). (E) Incomplete assembly due to missing sequence. (F) Incomplete assembly due to misassembly (noncontiguous sequence in the contig).\n\nFigure S6. Reference triplication for assembly assessment. (A) Due to the ambiguous starting position of a circular replicon, a completely-assembled contig will typically not align to the reference in a single unbroken alignment. (B) Doubling the reference sequence will allow for a single alignment, regardless of starting position. (C) However, if the contig contains start/end overlap (i.e. contiguity >100%) then even a doubled reference may not be sufficient to achieve a single alignment, depending on the starting position. (D) A tripled reference allows for an unbroken alignment, regardless of starting position, even in cases of >100% contiguity.\n\nFigure S7. Contiguity of the simulated read set assemblies plotted against Badread parameters for each of the tested assemblers. These plots show how well the assemblers tolerate different problems in the read sets. (A) Mean read depth (higher is better). (B) Max read identity (higher is better). (C) N50 read length (higher is better). (D) The sum of random read rate and junk read rate (lower is better). (E) Chimeric read rate (lower is better). (F) Adapter sequence length (lower is better). (G) Glitch size/skip (lower is better).\n\nFigure S8. Plasmid completion for the simulated read set assemblies for each of the tested assemblers, plotted with plasmid length and read depth. Solid dots indicate completely assembled plasmids (contiguity ≥99%) while open dots indicate incomplete plasmids (contiguity <99%). Percentages in the plot titles give the proportion of plasmids which were completely assembled.\n\nFigure S9. Plasmid completion for the real read set assemblies for each of the tested assemblers, plotted with plasmid length and read depth. Solid dots indicate completely assembled plasmids (contiguity ≥99%) while open dots indicate incomplete plasmids (contiguity <99%). Percentages in the plot titles give the proportion of plasmids which were completely assembled.\n\nFigure S10. The relative contiguity of the plasmids for each real read set assembly (A) and simulated read set assembly (B).\n\nExtended data are also available on GitHub.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Acknowledgements\n\nThis research was supported by use of the Nectar Research Cloud, a collaborative Australian research platform supported by the National Collaborative Research Infrastructure Strategy (NCRIS).\n\n\nReferences\n\nMyers EW: A history of DNA sequence assembly. IT - Information Technology. 2016; 58(3): 126–132. Publisher Full Text\n\nGurevich A, Saveliev V, Vyahhi N, et al.: QUAST: quality assessment tool for genome assemblies. Bioinformatics. 2013; 29(8): 1072–1075. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoodwin S, McPherson JD, McCombie WR: Coming of age: ten years of next-generation sequencing technologies. Nat Rev Genet. 2016; 17(6): 333–351. PubMed Abstract | Publisher Full Text\n\nLand M, Hauser L, Jun SR, et al.: Insights from 20 years of bacterial genome sequencing. Funct Integr Genomics. 2015; 15(2): 141–161. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHaubold B, Wiehe T: How repetitive are genomes? BMC Bioinformatics. 2006; 7: 541. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKyriakidou M, Tai HH, Anglin NL, et al.: Current Strategies of Polyploid Plant Genome Sequence Assembly. Front Plant Sci. 2018; 9: 1660. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLoman NJ, Quick J, Simpson JT: A complete bacterial genome assembled de novo using only nanopore sequencing data. Nat Methods. 2015; 12(8): 733–735. PubMed Abstract | Publisher Full Text\n\nBlin K: Ncbi genome downloading scripts. 2019. Reference Source\n\nWick R: rrwick/Long-read-assembler-comparison: Add supplementary figures. 2019. http://www.doi.org/10.5281/zenodo.3581590\n\nWick RR, Holt KE: rrwick/Assembly-Dereplicator: Assembly Dereplicator v0.1.0. 2019. Publisher Full Text\n\nWick RR: Badread: simulation of error-prone long reads. J Open Source Softw. 2019; 4(36): 1316. Publisher Full Text\n\nWick RR, Judd LM, Gorrie CL, et al.: Completing bacterial genome assemblies with multiplex MinION sequencing. Microb Genom. 2017; 3(10): e000132. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDe Maio N, Shaw LP, Hubbard A, et al.: Comparison of long-read sequencing technologies in the hybrid assembly of complex bacterial genomes. Microb Genom. 2019; 5(9): e000294. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWick RR, Judd LM, Gorrie CL, et al.: Unicycler: Resolving bacterial genome assemblies from short and long sequencing reads. PLoS Comput Biol. 2017; 13(6): e1005595. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi H: Minimap2: pairwise alignment for nucleotide sequences. Bioinformatics. 2018; 34(18): 3094–3100. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWick R: Read sets. 2019. http://www.doi.org/10.26180/5df6f5d06cf04\n\nWick R: Reference genomes. 2019. http://www.doi.org/10.26180/5df6e99ff3eed\n\nChin CS, Alexander DH, Marks P, et al.: Nonhybrid, finished microbial genome assemblies from long-read SMRT sequencing data. Nat Methods. 2013; 10(6): 563–569. PubMed Abstract | Publisher Full Text\n\nChin CS, Peluso P, Sedlazeck FJ, et al.: Phased diploid genome assembly with single-molecule real-time sequencing. Nat Methods. 2016; 13(12): 1050–1054. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKamath GM, Shomorony I, Xia F, et al.: HINGE: long-read assembly achieves optimal repeat resolution. Genome Res. 2017; 27(5): 747–756. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMyers EW: Efficient local alignment discovery amongst noisy long reads. Lecture Notes in Computer Science. LNBI, 2014; 8701: 52–67. Publisher Full Text\n\nZimin AV, Marçais G, Puiu D, et al.: The MaSuRCA genome assembler. Bioinformatics. 2013; 29(21): 2669–2677. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKoren S, Walenz BP, Berlin K, et al.: Canu: scalable and accurate long-read assembly via adaptive k-mer weighting and repeat separation. Genome Res. 2017; 27(5): 722–736. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMyers EW: The fragment assembly string graph. Bioinformatics. 2005; 21 Suppl 2: ii79–85. PubMed Abstract | Publisher Full Text\n\nKolmogorov M, Yuan J, Lin Y, et al.: Assembly of long, error-prone reads using repeat graphs. Nat Biotechnol. 2019; 37(5): 540–546. PubMed Abstract | Publisher Full Text\n\nVaser R, Sović I, Nagarajan N, et al.: Fast and accurate de novo genome assembly from long uncorrected reads. Genome Res. 2017; 27(5): 737–746. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVaser R, Šikić M: Yet another de novo genome assembler. bioRxiv. 2019. Publisher Full Text\n\nRuan J, Li H: Fast and accurate long-read assembly with wtdbg2. Nat Methods. 2019. PubMed Abstract | Publisher Full Text\n\nZerbino DR, Birney E: Velvet: algorithms for de novo short read assembly using de Bruijn graphs. Genome Res. 2008; 18(5): 821–829. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShafin K, Pesout T, Lorig-Roach R, et al.: Efficient de novo assembly of eleven human genomes using PromethION sequencing and a novel nanopore toolkit. bioRxiv. 2019. Publisher Full Text\n\nWick R: Assemblies. 2019. http://www.doi.org/10.26180/5df6e2864a658\n\nWick RR, Judd LM, Holt KE: Performance of neural network basecalling tools for Oxford Nanopore sequencing. Genome Biol. 2019; 20(1): 129. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWright CJ: Medaka. 2019. Reference Source\n\nAlexander DH: GenomicConsensus. 2019. Reference Source\n\nWick RR, Judd LM, Holt KE: August 2019 consensus accuracy update. 2019. Reference Source\n\nWalker BJ, Abeel T, Shea T, et al.: Pilon: an integrated tool for comprehensive microbial variant detection and genome assembly improvement. PLoS One. 2014; 9(11): e112963. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "58115",
"date": "09 Jan 2020",
"name": "Aleksey V. Zimin",
"expertise": [
"Reviewer Expertise Genomics",
"computational biology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe report is clear and concise, easy to read, and the authors' conclusions are well supported by their experimental results. The authors are to be commended for their unusual attention to reproducibility, and for making all data easily available.\nWe just have a couple of minor suggestions:\nReliability vs. robustness: the authors summarized their findings using the terms \"reliability\" for performance on real data sets, and \"robustness\" on simulated data sets. These terms might be a bit misleading to some readers. Reliability can be defined as consistent performance with good results, and robustness (in contrast) might be the ability to perform well under adverse conditions. The real data sets do vary in quality and coverage, although not as much as the simulated data. But it seems that both reliability and robustness can be evaluated on both types of data. If they want to use the term \"robustness,\" perhaps they could also plot the number of successful assemblies (or contiguity) vs the read error rate for each assembler. In this respect, a high error rate might be considered an adverse condition.\n\nFigure 1 is excellent, and provides a really nice summary of the performance on simulated data. However, only 1 of the programs, Flye, failed due to running out of memory, which was limited to 64 GB of RAM. Flye was otherwise one of the best performers. RAM is fairly inexpensive today, and it's not hard to find a server with >64 GB. The Figure doesn't show how much more memory Flye would need, and it would be really helpful to know that. Would 128GB allow it to complete in all cases? We suggest they run those failed assemblies on a larger-memory server and report what was needed. Another consideration here, though, is that depending on overcommit ratio and swap parameters, processes may be killed or slowed down long before they reach the 64GB physical memory limit. The impact of swap space on performance is an unknown here as well. For a clean evaluation, they should be sure (and maybe they did this, we can't tell) that swap was disabled and that the overcommit ratio was set to 97% to allow a process to use essentially all avaliable RAM. (There's more information about memory overcommit settings here) If swapping came into play on any of these jobs, then it would drastically increase runtime.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "5425",
"date": "22 Apr 2020",
"name": "Ryan Wick",
"role": "Author Response",
"response": "We thank the reviewer for their feedback, and changes to the article will be incorporated in its next version (along with updated results for newer assemblers/versions). Regarding point number 1: Supplementary figure S7 (available here) plots assembly contiguity against many different parameters used to generate the simulated reads, including maximum read identity. This gives a more detailed look at assembler ‘robustness’ towards a number of adverse conditions. Also, in the main text where the terms ‘reliability’ and ‘robustness’ are introduced, we have clarified that the simulated read sets contain adverse conditions which are not present in the real read sets. Regarding point number 2: We have created a new virtual machine on the Nectar Research Cloud with 128 GB of RAM (the most available in that service) and all new results (including those for Flye v2.7) were run on this VM. This has prevented assemblies from failing due to lack of memory. Since the larger VM allowed all assemblies to complete, we have opted to not alter the Linux memory settings and instead use the defaults. We checked memory statistics (as reported by /usr/bin/env time -v) and saw that major page fault counts were low (usually zero, sometimes in the tens and occasionally a few hundred for Canu), so we don’t believe that memory swapping has significantly impacted performance."
}
]
},
{
"id": "58113",
"date": "16 Jan 2020",
"name": "Robert Vaser",
"expertise": [
"Reviewer Expertise Sequence alignment",
"de novo assembly",
"algorithms",
"machine learning"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present a benchmark regarding prokaryotic organisms for several state-of-the-art long-read assemblers. The comparison includes both third generation sequencing technologies with real and simulated data, assessing various assembly traits with the conclusion that no assembler is perfect. The manuscript is well written, the figures look neat and all the data is freely available online.\nMinor comments:\nGenerating the assembly with a hybrid approach which is different from all benchmarked assemblers is a good approach, but is there a possibility to analyse in details datasets which have reference genomes assembled with Sanger sequencing (such as CFT073 and MGH78578 datasets used in De Maio N, Shaw LP, Hubbard A, et al.1)?\n\nAs minipolish is a new pipeline introduced in this paper, I would suggest describing it a bit more in detail.\n\nRa assembler has been published as a conference proceedings here.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "5426",
"date": "22 Apr 2020",
"name": "Ryan Wick",
"role": "Author Response",
"response": "We thank the reviewer for their feedback, and changes to the article will be incorporated in its next version (along with updated results for newer assemblers/versions). Regarding point number 1: We were reluctant to use Sanger-finished genomes as references for this study due to the dynamic nature of bacterial genomes. I.e. when a strain is sequenced multiple times from separate colonies and DNA extractions, there can be discrepancies between the underlying genomes. We encountered this problem when benchmarking Unicycler using public datasets for the E. coli K-12 MG1655 genome (10.1371/journal.pcbi.1005595). In that case, an insertion sequence had shifted in the genome relative to the Sanger-finished reference, causing false positive misassemblies. Scenarios such as this would be detrimental in our current study where even a single such discrepancy could seriously impact the contiguity metric we used (which requires zero misassemblies to achieve a contiguity of 100%). Instead, we opted to produce our own reference sequences (as described in the article) using De Maio et al’s single DNA extraction per isolate. Regarding point number 2: Further information on the Minipolish process is available on its GitHub page. We have now created a DOI for this repository to make a permanent digital record (10.5281/zenodo.3752203) and added it to the article’s references. Regarding point number 3: We have updated the article’s reference for Ra to the provided conference proceedings."
}
]
},
{
"id": "58301",
"date": "22 Jan 2020",
"name": "Mikhail Kolmogorov",
"expertise": [
"Reviewer Expertise Bioinformatics",
"genomics"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article presents the benchmarking of the current popular long-read assemblers (Canu, Flye, Miniasm/Minipolish, Raven, Redbean and Shasta) on various prokaryotic genomes. Wick & Holt have simulated 500 long-read datasets to reflect various genomic features (such as repeat length and complexity) as well as different sequencing parameters (depth, read length, sequencing artifacts etc). In addition, the authors test the assemblers on 160 real PacBio and Oxford Nanopore datasets. For each benchmarked algorithm, Wick & Holt summarize the important assembly metrics, such as contiguity or base-level accuracy (measured against the corresponding references), as well as overall user experience.\nThe manuscript is well-written, and the study design is sound. The presented benchmarks will be a valuable resource for the long-read genomics community, both for developers and users. Importantly, the authors have made all data sets and benchmarking pipelines freely available. I only have the following minor suggestions:\nIn my view, the evaluation pipeline designed by the authors could be highlighted more in the main text. E.g. how can a developer test a different assembler using the described benchmarks? Is it quick to reproduce? What would be the resource requirements?\n\nIt would be useful to compare the pros and cons of this work with the other assembly evaluation methods (such as QUAST) in a short discussion.\n\nOn Figure 2, triangles and circles are somewhat difficult to distinguish. Is there a way to better visually separate PacBio and ONT data points (maybe color tones or background pattern)?\n\nFor the sake of completeness, it is worth mentioning the minimap2 alignment identity threshold that is used for contiguity evaluation.\n\nDOI links to read sets and generated assemblies seem to have an unneeded space that break the URLs.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "5427",
"date": "22 Apr 2020",
"name": "Ryan Wick",
"role": "Author Response",
"response": "We thank the reviewer for their feedback, and changes to the article will be incorporated in its next version (along with updated results for newer assemblers/versions).Regarding point number 1:We have refined the script used to assess assemblies to make it more generalisable and usable: command line help text and usage information at the top of the script. We have also added a mention of the script and where it can be found to the main text of the paper: ‘The script for conducting this analysis (assess_assembly.py) is available in Extended data.’Regarding point number 2:We have added a brief comparison between our evaluation metric (contiguity) and QUAST to the main text: ‘This provides a simpler picture of assembly quality than is created by QUAST (which quantifies misassemblies and other metrics such as NG50) but is appropriate for cases where complete assembly is likely.’Regarding point number 3:We have changed the triangles for PacBio data points to X shapes, which are easier to distinguish from the circles used for ONT data points.Regarding point number 4:We have added the exact minimap2 options used to the main text of the article: ‘To encourage longer alignments, Minimap2 was run with the asm20 preset and chain elongation and banding thresholds of 10 kbp.’Regarding point number 5:We have removed the space to fix the links for these URLs."
}
]
},
{
"id": "58116",
"date": "30 Jan 2020",
"name": "Olin Silander",
"expertise": [
"Reviewer Expertise Microbial genomics and evolution",
"transcription",
"metagenomics"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors compare six long read genome assemblers using simulated and real data (PacBio and Nanopore). They find that there is no single best method, and that each offers distinct advantages and disadvantages. I enjoyed reading this paper. It was well written and clearly presented. As I understand, the authors plan to continually update the benchmarking is a fantastic step forward and considerably improves the utility of such a publication. This should be noted more explicitly in the manuscript.\nMajor comments:\nP.3 “Real Read Sets”. Could the authors note which fraction of the PacBio reads were CCS / HiFi reads?\n\np.4 para.1: We then excluded any isolate where either hybrid assembly failed to reach completion or where there were structural differences between the two assemblies as determined by a Minimap2 alignment. I wonder if this biases the genomes that were used such that they were easier to assemble than the genomes that were left out. I do not have a big problem with this, but it could be mentioned. It would also be good to provide slightly more detail on what precisely “structural differences between the two assemblies” means - e.g. does this include large indels (size range), inversions, etc.\n\nP.5 para.4: Figure 1B/Figure 2B shows the chromosome contiguity values for each assembly. There are some interesting patterns in 1B and 2B. First is the large number of Shasta assemblies have precisely 100.005% contiguity (looks to be mostly ONT assemblies). I am also surprised by the sort of bimodality in 1C/2C flye assemblies (and somewhat the miniasm assemblies). I would expect an even spread, but instead it looks like some assemblies have similar to 99% identity, whereas others have ~ 2-fold lower error rate (99.5% identity, my guesstimate). Is there an explanation for either of these patterns?\n\nP.5 Discussion of Identity. The authors could note the level generally achieved by polishing, which for ONT I think is around 99.98% (I am sure the authors are more aware than I am).\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "5428",
"date": "22 Apr 2020",
"name": "Ryan Wick",
"role": "Author Response",
"response": "We thank the reviewer for their feedback, and changes to the article will be incorporated in its next version (along with updated results for newer assemblers/versions). Regarding point number 1: None of the PacBio read sets were CCS – all were CLR. We have clarified this in the main text of the paper, noting that they are CLR reads when first introduced. Regarding point number 2: We have clarified both of these points in the text. The relevant section now reads: ‘We then excluded any isolate where either hybrid assembly failed to reach completion or where there were >50 nucleotide differences between the two assemblies as determined by a Minimap2 alignment. I.e. the Illumina+ONT and Illumina+PacBio hybrid assemblies needed to be in near-perfect agreement with each other. This left six isolates for inclusion. The above process may have biased these isolates in favour of easier-to-assemble genomes, as more complex genomes would be more likely to encounter inconsistencies between the two Unicycler assemblies.’ Regarding point number 3: These are indeed interesting patterns, but I can only speculate as to what the explanations are. Shasta is prone to producing ~10-15 bp of overlap in its assemblies. This may be related to the fact that Shasta operates on a reduced representation of the read sequences that is based on 10-mers. The bimodality of the Flye ONT assembly identity distribution is not as pronounced for the newer version of Flye (v2.7) but it is still there. The identity is relatively consistent within each genome (e.g. two read sets for a given genome tend to yield similar assembly identity), so I would speculate that the cause has something to do with the genome itself. E.g. perhaps the lower identity genomes have some type of DNA modification motif that is more likely to cause errors in the consensus sequence. Regarding point number 4: We have added to the text to elaborate on polished assembly identity: ‘Platform-specific post-assembly polishing tools (including Nanopolish, Medaka and Arrow) are routinely used to improve the accuracy of long-read assemblies, and these can often achieve assembly identities of >99.9% for ONT read sets and >99.999% for PacBio read sets (i.e. better than any of the assemblers were able to achieve on their own).’"
}
]
}
] | 1
|
https://f1000research.com/articles/8-2138
|
https://f1000research.com/articles/9-1050/v1
|
26 Aug 20
|
{
"type": "Research Article",
"title": "Detection of antifungal drug-resistant and ERG11 gene mutations among clinical isolates of Candida species isolated from Khartoum, Sudan.",
"authors": [
"Ahmed Osman Mohamed",
"Malik Suliman Mohamed",
"Mohamed Abdelrahman Hussain",
"Ibrahim Fatahalrahman Ahmed",
"Malik Suliman Mohamed",
"Mohamed Abdelrahman Hussain",
"Ibrahim Fatahalrahman Ahmed"
],
"abstract": "Background: Candida species are one of the most important opportunistic fungal pathogens that cause both superficial and systemic infections, especially in immunocompromised individuals. Considering the sharp increase in the rate of Candida infections, and resistance to commonly used antifungal agents in the last decades; this study was conducted to determine the rate of resistance among clinical isolates of Candida species, and to characterize some of the resistant genes among resistant isolates collected in Khartoum. Methods: This is a cross-sectional laboratory-based study included 100 pre-screened Candida species isolates from Khartoum state hospitals. Chromogenic media was used for Candida isolation and/or identification. The standard disc diffusion method was performed to investigate the susceptibility to fluconazole, itraconazole, and amphotericin. Following genomic DNA extraction, the entire ERG11 gene was amplified from some C. albicans resistant isolates, sequenced, and further analyzed. Results: Out of 100 clinical isolates collected, 51% were C. albicans, followed by C. glabrata (31%), C. krusie (8%), C. tropicals (5%), and C. dupliniens (5%). Resistance rate was 23% for fluconazole, 4% for itraconazole, while there were no amphotericin resistant isolates detected. C. albicans ERG11 gene sequence reveals 15 different mutations. Among these, three (D116E, E266D, and V488I) were missense mutations; however, these substitutions do not contribute to fluconazole resistance. Conclusion: C. albicans was found to be the most common species. Resistance against fluconazole was observed most frequently; however, mutations in ERG11 are unlikely to be the reason behind fluconazole resistance among these isolates.",
"keywords": [
"Candida species",
"fluconazole resistance",
"ERG11"
],
"content": "Introduction\n\nThe genus Candida is a dimorphic opportunistic fungal pathogen that colonizes the vagina, gastrointestinal and mucosal oral cavity of immunocompetent individuals. In contrast, critically ill and/or immunocompromised patients frequently develop Candida infection that range from superficial to systemic infections1. Candida comprises over 150 species, of which 17 are prevalent and known to infect humans; these include Candida albicans, Candida glabrata, Candida parapsilosis, Candida tropicalis, and Candida krusei2,3.\n\nSince the 1980s, there has been a steady increase in the incidence and prevalence of serious secondary systemic fungal infections4. The risk factors for developing systemic or superficial Candida infections include intensive care unit admissions, HIV infection, organ transplantation, cancer and anticancer drugs, diabetes mellitus and other demographic factors such as age and sex5–9.\n\nVarious antifungals are used for treatment of Candida infections, among them, azoles which showed good activity and are relatively safe, however, resistance to this group is occurring more frequently10. On the other hand, resistance of Candida species against other antifungals such as Polyenes, Echinocandins, and Allylamines has not been reported extensively11.\n\nThere are different molecular mechanisms through which eukaryotic cells may develop drug resistance such as alteration in efflux pump, alteration in intracellular drug processing i.e. modification or degradation, alterations in the target enzyme and/ or other enzymes. Among these mechanisms, alteration of the targeted enzyme and alteration of efflux pump are the most common in Candida species12.\n\nThe azoles bind to and inhibit the activity of 14 α-demethylase, a key enzyme in the fungal ergosterol synthesis pathway, which belongs to the cytochrome P450 family10. One of the known potential resistance mechanism of azoles is alteration in the ergosterol syntheses pathway. Candida species can develop resistance by mutation/s in the gene (ERG11) which codes for the enzyme 14α-demethylase13. The point mutation in the ERG11 gene can result in an amino acid substitution which in turn produces a conformational change in the enzyme and decreases the affinity to azoles, however, susceptibilities are not affected equally by different mutations as the presence of some mutations such as Y132H, R467K, and I471T confirm resistance, on the other hand, mutations such as E266D does not affect the resistance14. In addition, drug resistance might develop through overexpression of the ERG11 gene through increased mRNA level which might increase the concentration of the enzyme in comparison to sensitive isolates15.\n\nEfflux pumps are the basic mechanism in most eukaryotic cells by which unwanted toxic materials are forced out of the cell. Two types of efflux pumps have been identified: ATP binding cassette (ABC) transporters and major facilitator superfamily (MFS)16. ABC transporters are pumps that are actively associated with the efflux of potentially toxic molecules to the cell, and they are primarily hydrophobic and lipophilic17. Many different pumps are known to belong to the different families, such as Candida resistance (CDR1&2) genes that are related to the PDR5 family, and known to be associated with resistance to antifungals15. The MFS pumps work by antiproton power i.e. proton pumped inside the cell and hydrophobic and/or lipophilic material pumped outside the cell. This is coded by multidrug resistance gene (MDR1) which found to be overexpressed in fluconazole resistant isolates, however, the gene was not overexpressed in other azoles, ketoconazole and itraconazole, resistant strains18. Some authors have tried to link overexpression of CDR1, 2 and MDR1, concluding that deletions of these genes will result in more susceptible isolate than each gene alone12.\n\nDue to the scarcity of reports about the rate of drug resistance and resistant genes among Candida species in the study area, this study was conducted to screen the susceptibility of different Candida species towards commonly used antifungals, and to identify the role of ERG11 gene mutation/s in the development of fluconazole resistance. To this end, we collected and identified isolates of Candida species, selected the most resistant isolates, amplified and sequenced the conserved domain of the ERG11 gene and detected the impact of the mutation(s) on the enzyme 14α-demethylase’s structure and function using in silico tools19,20.\n\n\nMethods\n\nThis study was reviewed and approved by the Research ethical committee, Faculty of Medicine & Health Science, International University of Africa (IUA) (6-2017). Oral informed consent was obtained from the participating patients, when their hospital laboratory result was positive for Candida species. Oral consent was obtained over written consent (where it was recorded), since the majority of the patients included in this study were illiterate (in case of minors, consent was obtained from parents or guardians). The structure of the consent was approved by the Research Ethics Committee.\n\nThis was a cross-sectional laboratory-based study using clinical isolates of Candida species. Clinical isolates were collected from different Khartoum state hospitals in a period between September 2017 to September 2018, all clinical isolates primarily identified as Candida species regardless of age, gender, and site of isolations, were included. Samples and demographic data were obtained directly from the patients within each hospital after consent was obtained by the principle investigators.\n\nSample size was calculated using the following formula on cross-sectional studies21:\n\nn = Z2 * P (1-P) / d2\n\nWhere, n = desired sample size, Z = critical value and a standard value for the corresponding level of confidence (1.96), P = expected frequency of resistance obtained from previous studies (7%)22,23, d = margin of error (0.05).\n\nn = 1.962 * 0.07 (1-0.07) / (0.05)2 = 100 samples.\n\nA total of 100 clinical isolates of Candida species were collected from Khartoum state hospitals, Sudan. The isolates were pre-identified at each hospital’s laboratory using conventional methods such as wet mount, gram stain, germ tube, and growth on Sabouraud Dextrose Agar media (SDA). Immediately after collection, the isolates were grown into SDA (M063, HIMEDIA, Mumbai, India) at 32°C for 24–48 hours. From the grown culture, colonies were picked and streaked over a slant of SDA in screw-cap tubes, each slant was filled with sterile liquid glycerol and tightly closed and stored in 4 C° until recovered.\n\nChromogenic media Hi-Chrome Candida differential agar media supplemented with chloramphenicol 0.5g/L (M1297A, HIMEDIA, Mumbai, India) was used to differentiate between Candida species based on colonies’ color and morphology. A subculture from the stock culture was allowed to grow on SDA for 24 hour, subsequently one well isolated colony from the grown culture was picked out and streaked over the prepared Hi-Chrome media and incubated at 32°C for 24–48 hours, the result was interpreted as per manufacture instructions (C. albicans—light green, C. glabrata—cream to white, Candida krusei—pale, fuzzy and C. tropicalis—blue to purple, C. dupliensis—pale green)24.\n\nSensitivity testing to all isolated Candida species was carried out as recommended by the Clinical Laboratories and Standard Institute (CLSI)25. A modified Mueller Hinton Agar media (M173, HIMEDIA, Mumbai, India) supplemented with 2% glucose and methyl blue 5µg/mL was used. Using overnight culture on SDA, the inoculum was prepared by suspending 4 well-isolated colonies in 5mL sterile saline, inoculum size was adjusted by matching the turbidity with standard McFarland which was prepared by adding 0.5 mL BaCl2 (0.048 mol/L) to 99.5 mL H2SO4. Within 15 minutes after adjusting the turbidity and by using sterile cotton swab, the microorganisms were streaked from the center of the petri dish to the top, each time the plate was rotated 60° to ensure that the agar surface is at least double streaked. Within 15 minutes after streaking, three discs, namely fluconazole 25µg, itraconazole 10µg and amphotericin 10µg (SD232, SD221, SD111. HIMEDIA, Mumbai, India) were applied to each inoculated petri dish, gently pressed into the agar using sterile forceps, incubated at 32C° for 24–48 hours, zone dimeter around each disc were measured using calipers and result was interpreted as per CLSI25.\n\nGenomic DNA was extracted using guanidine chloride method, briefly, DNA extraction was carried out using 48 hours grown culture on SDA media, three to five colonies were washed with 5 mL phosphate buffer saline (PBS) three times, then 2 mL white cell lysis buffer and 20 μL of proteinase K (10 mg/mL; iNtRON Inc, Korea) were added to the pellet in a Falcon tube, vortexed and incubated at 37°C overnight. Then 1 mL from guanidine chloride (7M; iNtRON Inc, Korea) and 350 μL of ammonium acetate (7M; Loba Chemie, India) were added. The tubes were vortexed and incubated at 65°C in a water bath for 2 hours. After that 2 mL pre chilled chloroform (sd Fine-Chem limited, India) was added, and centrifuged at 6000 RPM for 20 minutes and the supernatant was transferred into a new Falcon tube and completed to 10 mL volume with pre chilled absolute ethanol (Carlo Erba, France) and incubated overnight at -20°C for completion of DNA precipitation. After incubation the tubes were centrifuged at 6000 RPM for 20 minutes, then the ethanol was poured off and the same step was repeated with 70% ethanol. After that the tubes were left to air dry. Finally, DNA was suspended in 80 μL TE buffer (iNtRON Inc, Korea) and incubated at 4 °C until used26, as a template for PCR. The ERG11 gene from C. albicans was amplified using previously published primers14,27: forward: (5′-CAAGAAGATCATAACTCAAT-3′) and reverse (3′-AGAACACTGAATCGAAAG-5′) (Macrogen Inc. Korea). All PCRs were carried out in final volume of 20 µL containing Maxime PCR PreMix Kit i-Tag (2.5U i-TagTM DNA polymerase, 2.5mM each dNTPs, 1x reaction buffer and 1x Gel loading buffer), 1µL each forward and reverse primer (10pmol final concentration), 2.5µL genomic DNA, and the volume was completed with distilled water. The PCR was carried out in G-storm thermocycler with the following conditions: initial denaturation at 94°C for 4 min; 35 cycle of denaturation at 95°C for 30 s, primer annealing at 55°C for 30 s, and extension at 70°C for one min; followed by final extension step at 72°C for 10 min. The final product was visualized by loading 3µL in 0.8% agarose gel electrophoresis for 45 minutes under the voltage 100 V. Distinct bands were observed under UV light and photographed. The ERG11 gene from the most C. albicans resistant isolates (isolate 10, 13, and 14) and one sensitive isolate (24 randomly selected, double blinded by independent researcher) were selected for sequencing (Sanger sequencing in BGI, Shenzhen, China). All sequences were deposited in GenBank and accession numbers MT081007, MT081008, MT081009, MT081010 for isolate 10, 13, 14, and 24, respectively, were obtained. Sequences were aligned based on fluconazole susceptible strain SC5314 (GenBank accession number X13296) using BioEdit software version 7.2.5.\n\n\nResults\n\nOut of 100 samples collected, 80 were from females and 20 from males with a mean age of 40. 66 isolates were from urine samples, 17 from sputum, 12 were high from vaginal swabs, and 5 from other sites. Overall, the most common species was C. albicans (n=51), while the most prevalent Non-albicans species was C. glabrata (n=31), followed by C. krusei (n=8). Table 1 provides a detailed description on the frequency of each species with respect to the site of isolation28.\n\nFluconazole was the least effective agent followed by itraconazole. Itraconazole resistance was observed among Non-C. albicans (NCA) species, high frequency of intermediate susceptibility dose-dependent (ISDD) was observed for itraconazole among all Candida species , while there was no amphotericin resistant isolate detected.\n\nFluconazole resistance was observed in 23% of C. albicans samples, only 2 isolates were ISDD, and the remaining isolates (72.5%) were sensitive, there were no itraconazole and amphotericin resistant C. albicans isolates. However, 31% were categorized as ISDD to itraconazole. Among NCA species, 19.4% of C. glabrata were fluconazole resistant, as well as all C. krusei isolates (8). Azole cross resistance was observed among 2 C. glabrata and 2 C. krusei isolates. One C. dupliensis was resistant to fluconazole while there were no C. tropicalis resistant isolates. Complete AST results are shown in Table 228.\n\nF: Fluconazole, I: Itraconazole, A: Amphotericin.\n\nThe complete ERG11 gene coding region (1587 bp) from C. albicans resistant isolates (12) and one sensitive isolate (control) were amplified as shown in Figure 129. Sequence analysis revealed 15 different mutations, 12 of which were silent, and 3 were non-synonyms Table 3. T495A and G1609A were observed only in resistant isolates (isolate 10 and 14 respectively), while A945C was observed in sensitive and resistant isolates (isolate 13, 14, and 24). All mutations were previously reported (Table 330).\n\nFrom left to right lanes: L, 100 bp DNA ladder, 1 to 12 lanes are the amplified Candida albicans ERG11’s gene (1587 bp).\n\nAsp= Aspartic acid, Glu= Glutamic acid, Val= Valine, Ile= Isoleucine\n\nAll sequences were aligned based on C. albicans reference strain, with X13296 GenBank accession number.\n\n\nDiscussion\n\nIn the current study, 100 clinical isolates of different Candida species were collected, identified, and their susceptibilities to fluconazole, itraconazole, and amphotericin were determined. The ERG11 gene was amplified from some resistant isolates to investigate the impact of different mutations in the enzyme activity and hence drug susceptibility.\n\nNearly half of the samples collected were C. albicans (51%), while C. glabrata (31%) and C. krusie (8%) being the most prevalent of the other species identified (Table 128). C. albicans remains to be the most common species (51%), a similar finding was obtained in Sudan, in 2008, in a study that aimed to characterize vaginal candidiasis among pregnant woman indicating that the prevalence was 81%31. More recently in 2018 a study conducted on cancer patients at the Isotope and Radiation Centar in Sudan, concluded that the prevalence of C. albicans was 59%32, however, these numbers indicate that the prevalence of non-Candida albicans species are increasing: 19% in 2008, 41% in 2018, and 49% in this study. This result indicates the necessity of culturing any suspected Candida infections at species level for proper management.\n\nAntimicrobial sensitivity testing reveals that fluconazole was the least effective agent, followed by itraconazole, while there were no amphotericin resistant isolates (Table 228,29). In this regard, several studies report nearly the same degree of fluconazole susceptibility against C. albicans (23%)33,34. Azoles cross-resistance was observed in 2 C. glabrata and 2 C. krusie isolates (Table 228,29). It has been observed that drug resistant fungal pathogens are increasing, and reduced susceptibility to azoles, especially fluconazole, along with azole cross resistance was detected3. According to these findings we believe that there is urgent need for AST, especially when physicians intend to prescribe fluconazole as it is the least effective agents, or in such settings that non- C. albicans species are suspected as they possess relatively high rates of resistance.\n\nThe ERG11 gene was sequenced from some C. albicans resistant isolates and one sensitive isolate for the purpose of examining the impact of different mutations (if present) on fluconazole resistance. The three detected mutations (T495A, A945C, and G1609A, which precipitate D116E, E266D, and V488I aa substitutions respectively) have been described previously in both sensitive and resistant isolates, and strongly suggesting that they are not contributed directly to resistance.\n\nIn the present study, E266D aa substitution which was described by some authors as the most common polymorphism in the ERG11 gene has been detected in both sensitive (isolate 24) and resistant (isolate 13, 14) isolates (Table 330), so our finding completely agree with previous data which concluded that this mutation alone has no role in resistance27,35.\n\nIn our analysis, D116E and V488I aa substitutions were detected only in resistant isolates (isolate 10 and 14 respectively, Table 330); the same results have been described previously27,36; however, the detection of these mutation in fluconazole susceptible isolates indicates that they lack a vital role in the development of resistance14. In isolate 14, E266D and V488I were found together, a similar finding was obtained previously36. The impact of E266D occurring simultaneously with other mutations such as K143R, F145L, and G464S have been well characterized using site directed mutagenesis35, unlike the coexistence of E266D and V488I which is needed to be more clarified.\n\nOne of the limitation in this study is that we are unable to detect some regions and therefore some mutations at the beginning and/or end of the ERG11 gene because of their low quality (common problem in Sanger sequencing for sequences more than 1000 bp)37. We have tried to solve this problem by using either forward or reverse sequencing reads, however, it was very difficult to double check some of these mutations for further confirmations. Our recommendation in this regard is to consider different sequencing techniques that are able to detect the entire region (1587bp) reliably.\n\nAccording to our results, ERG11 gene mutations in C. albicans was not the main causes of resistance, our future recommendations lay on considering alternative resistance mechanisms, more especially, studying the expression level of CDR1, CDR2, MDR1, and ERG11 genes which expected to give a complete view of the resistance processes.\n\n\nConclusion\n\nNearly half of the identified isolates were C. albicans, and the most prevalent non- C. albicans was C. glabrata. Among all antifungals tested, fluconazole was the least effective, while all isolates were sensitive to amphotericin. The detected missense mutations were not directly associated with fluconazole resistance; however, resistance among these isolates might be due to other mechanisms such as efflux pump gene overexpression.\n\n\nData availability\n\nAll sequences were deposited in GenBank under accession numbers MT081007, MT081008, MT081009, and MT081010 for isolate 10, 13, 14, and 24, respectively.\n\nFigshare: demographic, identification, sensitivity test data. https://doi.org/10.6084/m9.figshare.12449615.v128\n\nThis project contains the following underlying data:\n\n- sample collection sheet for F1000.xlsx (A spreadsheet contain data regarding patients age, gender and site of isolation, species assay results for each sample coupled with susceptibility to fluconazole, itraconazole and amphotericin).\n\nFigshare: sequence data for ERG11 gene. https://doi.org/10.6084/m9.figshare.12600128.v130\n\nThis project contains the following underlying data:\n\n- 10__[19070622] F__D01_1907004923G.ab1 (Raw sequence data for opening in Finish TV for isolate 10, forward sequencing).\n\n- 10__[19070624]R__A02_1907004923G.ab1 (Raw sequence data for opening in Finish TV for isolate 10, reverse sequencing).\n\n- 13__[19070622]F__E01_1907004924G.ab1 (Raw sequence data for opening in Finish TV for isolate 13, forward sequencing).\n\n- 13__[19070624]R__B02_1907004924G.ab1 (Raw sequence data for opening in Finish TV for isolate 13, reverse sequencing).\n\n- 14__[19070622]F__F01_1907004925G.ab1 (Raw sequence data for opening in Finish TV for isolate 14, forward sequencing).\n\n- 14__[19070624]R__C02_1907004925G.ab1 (Raw sequence data for opening in Finish TV for isolate 14, reverse sequencing).\n\n- 24__[19070622]F__G01_1907004926G.ab1 (Raw sequence data for opening in Finish TV for isolate 24, forward sequencing).\n\n- 24__[19070622]F__G01_1907004926G.ab1 (Raw sequence data for opening in Finish TV for isolate 24, reverse sequencing).\n\nFigshare: candida identification using Hi-chrome media and gel electrophoresis for ERG11 gene. https://doi.org/10.6084/m9.figshare.12775769.v129\n\nThis project contains the following underlying data:\n\n- ERG11 gene from C. albicans.jpg (Raw image for PCR gel for ERG11 gene).\n\n- 131733_2018.3.7.jpg (Raw images for identification of isolates 1-9, inverted plate).\n\n- 131739_2018.3.7.jpg (Raw images for identification of isolates 1-9, upright plate).\n\n- 131718_2018.3.7.jpg (Raw images for identification of isolates 15-22, inverted plate).\n\n- 131724_2018.3.7.jpg (Raw images for identification of isolates 15-22, upright plate).\n\n- 131630_2018.3.7.jpg (Raw images for identification of isolates 27-34, upright plate).\n\n- 131637_2018.3.7.jpg (Raw images for identification of isolates 27-34, inverted plate).\n\n- 131532_2018.3.7.jpg (Raw images for identification of isolates 45-52, inverted plate).\n\n- 131541_2018.3.7.jpg (Raw images for identification of isolates 45-52, upright plate).\n\n- 131555_2018.3.7.jpg (Raw images for identification of isolates 56-63, upright plate).\n\n- 131605_2018.3.7.jpg (Raw images for identification of isolates 56-63, inverted plate).\n\n- 131619_2018.3.7.jpg (Raw images for identification of isolates 64-71, upright plate).\n\n- 131645_2018.3.7.jpg (Raw images for identification of isolates 64-71, inverted plate).\n\n- 131659_2018.3.7.jpg (Raw images for identification of isolates 73-80, inverted plate).\n\n- 131708_2018.3.7.jpg (Raw images for identification of isolates 73-80, upright plate).\n\n- 131508_2018.3.7.jpg (Raw images for identification of isolates 82-86, upright plate).\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "References\n\nSardi J, Scorzoni L, Bernardi T, et al.: Candida species: current epidemiology, pathogenicity, biofilm formation, natural antifungal products and new therapeutic options. J Med Microbiol. 2013; 62(Pt 1): 10–24. PubMed Abstract | Publisher Full Text\n\nLass-Flörl C: The changing face of epidemiology of invasive fungal disease in Europe. Mycoses. 2009; 52(3): 197–205. PubMed Abstract | Publisher Full Text\n\nSzweda P, Gucwa K, Romanowska E, et al.: Mechanisms of azole resistance among clinical isolates of Candida glabrata in Poland. J Med Microbiol. 2015; 64(6): 610–9. PubMed Abstract | Publisher Full Text\n\nYoon HJ, Choi HY, Kim YK, et al.: Prevalence of fungal infections using National Health Insurance data from 2009-2013, South Korea. Epidemiol Health. 2014; 36: e2014017. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSah P, Patel P, Chandrashekar C, et al.: Oral candidal carriage correlates with CD4+ cell count but not with HIV and highly active antiretroviral therapy status. J Investig Clin Dent. 2019; 10(4): e12438. PubMed Abstract | Publisher Full Text\n\nKoehler P, Stecher M, Cornely OA, et al.: Morbidity and mortality of candidaemia in Europe: an epidemiologic meta-analysis. Clin Microbiol Infect. 2019; 25(10): 1200–1212. PubMed Abstract | Publisher Full Text\n\nChouhan S, Kallianpur S, Prabhu KT, et al.: Candidal prevalence in diabetics and its species identification. Int J Appl Basic Med Res. 2019; 9(1): 49–54. PubMed Abstract | Free Full Text\n\nKim YJ, Kim SI, Choi JY, et al.: Invasive fungal infection in liver transplant recipients in a prophylactic era: A multicenter retrospective cohort study in Korea. Medicine (Baltimore). 2019; 98(26): e16179. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPfaller M, Diekema D: Epidemiology of invasive candidiasis: a persistent public health problem. Clin Microbiol Rev. 2007; 20(1): 133–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAllen D, Wilson D, Drew R, et al.: Azole antifungals: 35 years of invasive fungal infection management. Expert Rev Anti Infect Ther. 2015; 13(6): 787–98. PubMed Abstract | Publisher Full Text\n\nBossche HV, Marichal P, Odds FC: Molecular mechanisms of drug resistance in fungi. Trends Microbiol. 1994; 2(10): 393–400. PubMed Abstract | Publisher Full Text\n\nWhite TC, Marr KA, Bowden RA: Clinical, cellular, and molecular factors that contribute to antifungal drug resistance. Clin Microbiol Rev. 1998; 11(2): 382–402. PubMed Abstract | Free Full Text\n\nSpettel K, Barousch W, Makristathis A, et al.: Analysis of antifungal resistance genes in Candida albicans and Candida glabrata using next generation sequencing. PLoS One. 2019; 14(1): e0210397. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYing Y, Zhao Y, Hu X, et al.: In vitro fluconazole susceptibility of 1,903 clinical isolates of Candida albicans and the identification of ERG11 mutations. Microb Drug Resist. 2013; 19(4): 266–73. PubMed Abstract | Publisher Full Text\n\nChen L, Xu Y, Zhou C, et al.: Overexpression of CDR1 and CDR2 genes plays an important role in fluconazole resistance in Candida albicans with G487T and T916C mutations. J Int Med Res. 2010; 38(2): 536–45. PubMed Abstract | Publisher Full Text\n\nJia XM, Ma ZP, Jia Y, et al.: RTA2, a novel gene involved in azole resistance in Candida albicans. Biochem Biophys Res Commun. 2008; 373(4): 631–6. PubMed Abstract | Publisher Full Text\n\nPrasad R, De Wergifosse P, Goffeau A, et al.: Molecular cloning and characterization of a novel gene of Candida albicans, CDR1, conferring multiple resistance to drugs and antifungals. Curr Genet. 1995; 27(4): 320–9. PubMed Abstract | Publisher Full Text\n\nMarger MD, Saier Jr MH: A major superfamily of transmembrane facilitators that catalyse uniport, symport and antiport. Trends Biochem Sci. 1993; 18(1): 13–20. PubMed Abstract | Publisher Full Text\n\nVenselaar H, te Beek TA, Kuipers RK, et al.: Protein structure analysis of mutations causing inheritable diseases. An e-Science approach with life scientist friendly interfaces. BMC Bioinformatics. 2010; 11(1): 548. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCapriotti E, Fariselli P, Casadio R: I-Mutant2.0: predicting stability changes upon mutation from the protein sequence or structure. Nucleic Acids Res. 2005; 33(Web Server issue): W306–W10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPourhoseingholi MA, Vahedi M, Rahimzadeh M: Sample size calculation in medical studies. Gastroenterol Hepatol Bed Bench. 2013; 6(1): 14–7. PubMed Abstract | Free Full Text\n\nMulu A, Kassu A, Anagaw B, et al.: Frequent detection of ‘azole’resistant Candida species among late presenting AIDS patients in northwest Ethiopia. BMC Infect Dis. 2013; 13(1): 82. PubMed Abstract | Publisher Full Text | Free Full Text\n\nToda M, Williams SR, Berkow EL, et al.: Population-based active surveillance for culture-confirmed candidemia–four sites, United States, 2012–2016. MMWR Surveill Summ. 2019; 68(8): 1–15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBaradkar VP, Mathur M, Kumar S: Hichrom candida agar for identification of Candida species. Indian J Pathol Microbiol. 2010; 53(1): 93–5. PubMed Abstract | Publisher Full Text\n\nWayne P: Method for antifungal disk diffusion susceptibility testing of yeasts; approved guideline. M44-A2 edn, Clinical and Laboratory Standards Institute; 2009. Reference Source\n\nAlsadig G, Arbab MA, Aldeaf SA, et al.: Allele frequency of P53 gene Arg72Pro in sudanese meningioma patients and controls. J of SCIENTIFIC & TECHNOLOGY RES. 2014; 3(6): 2277–8616. Reference Source\n\nWang B, Huang LH, Zhao JX, et al.: ERG11 mutations associated with azole resistance in Candida albicans isolates from vulvovaginal candidosis patients. Asian Pac J Trop Biomed. 2015; 5(11): 909–14. Publisher Full Text\n\nKunna A, Osman A: demographic, identification, sensitivity test data. figshare. Dataset. 2020. http://www.doi.org/10.6084/m9.figshare.12449615.v1\n\nKunna A: candida identification using Hi-chrome media and gel electrophoresis for ERG11 gene. figshare. Media. 2020. http://www.doi.org/10.6084/m9.figshare.12775769.v1\n\nKunna A: sequence data for ERG11 gene. figshare. Dataset. 2020. http://www.doi.org/10.6084/m9.figshare.12600128.v1\n\nNemery HM: Phenotypic Characterization of Vaginal Candidiasis in Sudanese Pregnant Women. Afr J Med Sci. 2019; 4(4). Reference Source\n\nNagla MM, El Fadil OE, Muzamil AHM, et al.: Internal transcribed spacer for identification of yeast species isolated from cancer patients at the Isotope and Radiation Center, Khartoum, Sudan: A cross-sectional, case-control study [version 1; peer review: 1 approved, 1 approved with reservations]. F1000Res. 2018; 7: 443. Publisher Full Text\n\nKhadka S, Sherchand JB, Pokhrel BM, et al.: Isolation, speciation and antifungal susceptibility testing of Candida isolates from various clinical specimens at a tertiary care hospital, Nepal. BMC Res Notes. 2017; 10(1): 218. PubMed Abstract | Publisher Full Text | Free Full Text\n\nElFeky DS, Gohar NM, El-Seidi EA, et al.: Species identification and antifungal susceptibility pattern of Candida isolates in cases of vulvovaginal candidiasis. Alexandria Journal of Medicine. 2016; 52(3): 269–77. Publisher Full Text\n\nFlowers SA, Colón B, Whaley SG, et al.: Contribution of clinically derived mutations in ERG11 to azole resistance in Candida albicans. Antimicrob Agents Chemother. 2015; 59(1): 450–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSardari A, Zarrinfar H, Mohammadi R: Detection of ERG11 point mutations in Iranian fluconazole-resistant Candida albicans isolates. Curr Med Mycol. 2019; 5(1): 7–14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang XV, Blades N, Ding J, et al.: Estimation of sequencing error rates in short reads. BMC Bioinformatics. 2012; 13(1): 185. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "70299",
"date": "07 Sep 2020",
"name": "Rasoul Mohammadi",
"expertise": [
"Reviewer Expertise Medical Mycology"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript is a common epidemiological study. It has some major drawbacks including:\nIsolates were identified by phenotypic methods. Genus of Candida contains about 200 species and identification needs to confirm by molecular techniques. Conventional methods are not enough to identify Candida spp.\n\nThe majority of isolates were obtained from urine and sputum (66+17). Candida spp. are normal flora and authors have to confirm the infections caused by these fungi.\n\nIsolation of Candida from urine (candiduria) is not an infection, we have some criteria for urinary tract infections (UTIs), and the authors have to confirm it. Isolation of Candida from sputum is not important because we can isolate Candida spp. from sputum, even in immunocompetent individuals.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6310",
"date": "27 Jan 2021",
"name": "Ahmed Kunna",
"role": "Author Response",
"response": "Dear Prof Rasoul Mohammadi; Thank you for comprehensively reviewing our paper and for your appreciated comments: Regarding the first comments; The primary objective of the study was to isolate the medically important Candida species, characterizing Candida at species level is beyond the scope of the study. The sensitivity and specificity of Chromogenic media in differentiating as well as isolating common species such as (C. albicans, C. glabrata, C. Krusie, and C. tropicals) is well documented in the previous studies (see Evaluation of chromogenic media and seminested PCR in the identification of Candida species ), according to previous publication in Sudan; more than 90% of infections due to Candida species was caused by this species (see Internal transcribed spacer for identification of yeast species isolated from cancer patients at the Isotope and Radiation Center, Khartoum, Sudan: A cross-sectional, case-control study). On the other hand, culturing was the efficient and the easiest way to isolate the mixed isolates within the same sample. With respect to the second comments; Candiduria is considered one of the most controversial issues in patient management. As well as isolation of Candida from oral cavity since this species is a commensal microbe. During the course of sample collections (1 year), we have tried to overcome this problem by carefully selecting our isolates for eligibility since: All of isolates were derived from a patient with some risk factors such as Diabetes mellitus, hospitalised patient, and elderly patients. All isolates were recovered from patients with characteristic sign and symptoms for UTI, while the growth of bacteria was insufficient to prove bacterial infections. In most of the cases urine analysis as well as culture were repeated. The final decision of considering Candida as a sole causative agent was made by a physician, upon that recommendations sample were considered eligible."
}
]
},
{
"id": "73504",
"date": "11 Nov 2020",
"name": "Raja Ahsan Aftab",
"expertise": [
"Reviewer Expertise Clinical Pharmacy",
"infectious disease",
"nephrology",
"treatment outcomes"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study is about anti-fungal sensitivity testing and detection of resistant gene/s. It is important to carry out such kinds of studies to help prescribers and policymakers in making better future plans and guidelines for the treatment of candida infection.\nIt would be better if the authors could clarify why these 3 anti-fungal drugs were selected for the sensitivity testing in the introduction.\n\nThe authors need to add a few more studies and comparisons about the candida resistance during the discussion.\n\nThe authors collected 100 samples from different sites and genders, the exact percentage of resistance/sensitive isolates with regard to gender and site of isolation can be added to the tables or as a text.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6300",
"date": "29 Jan 2021",
"name": "Ahmed Kunna",
"role": "Author Response",
"response": "Dear Prof Raja Ahsan Aftab: Thank you for your valuable comments. In response to your 3 comments we have: Justified the reasons for selecting the 3 antifungals in the Introduction. We have added the percentage of resistant isolates in relation to gender in Result section, and the percentage of each site of isolations as crosstabulation in Table 1. We have comprehensively discussed the pattern of antifungal resistance especially in our region (Africa) in the Discussion section by adding 7 additional citations."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1050
|
https://f1000research.com/articles/9-678/v1
|
06 Jul 20
|
{
"type": "Study Protocol",
"title": "Current practice in systematic reviews including the ‘PICO for each synthesis’ and methods other than meta-analysis: protocol for a cross-sectional study",
"authors": [
"Miranda S. Cumpston",
"Joanne E. McKenzie",
"James Thomas",
"Sue E. Brennan",
"Miranda S. Cumpston",
"Joanne E. McKenzie",
"James Thomas"
],
"abstract": "Introduction: Systematic reviews are used to synthesise research and inform decision making by clinicians, consumers and policy makers. The synthesis component of systematic reviews is often narrowly considered as the use of statistical methods to combine the results of studies, primarily meta-analysis. However, synthesis can be considered more broadly as a process beginning with: (i) defining the groupings of populations, interventions and outcomes to be compared (the ‘PICO for each synthesis’); (ii) examining the characteristics of the available studies; and (iii) applying synthesis methods from among multiple options. To date, there has been limited examination of approaches used in reviews to define and group PICO characteristics and synthesis methods other than meta-analysis. Objectives: To identify and describe current practice in systematic reviews in relation to structuring the PICO for each synthesis and methods for synthesis when meta-analysis is not used. Methods: We will randomly sample 100 systematic reviews of the effects of public health and health systems interventions published in 2018 and indexed in the Health Evidence and Health Systems Evidence databases. Two authors will independently screen studies for eligibility. One author will extract data on approaches to grouping and defining populations, interventions and outcomes, and the rationale for the chosen groups; and the presentation and synthesis methods used (e.g. tabulation, visual displays, statistical synthesis methods such as combining P values, vote counting based on direction of effect). A second author will undertake independent data extraction for a subsample of reviews. Descriptive statistics will be used to summarise the findings. Specifically, we will compare approaches to grouping in reviews that primarily use meta-analysis versus those that do not. Conclusion: This study will provide an understanding of current practice in two important aspects of the synthesis process, enabling future research to test the feasibility and impact of different methodological approaches.",
"keywords": [
"Systematic reviews",
"meta-analysis",
"synthesis",
"subgroup analysis",
"narrative synthesis",
"synthesis without meta-analysis",
"PICO"
],
"content": "Introduction\n\nSystematic reviews provide a method for collating and synthesising research, and are used to inform decision making by clinicians, consumers and policy makers1. The synthesis component of systematic reviews is often narrowly considered as the use of statistical methods to combine the results of studies, primarily meta-analysis, and much of the available guidance focuses on this approach. However, ‘synthesis’ can be considered more broadly as a process, beginning with defining the review questions, planning the groups to be compared, examining the characteristics of the available studies and their data, and applying appropriate synthesis methods from among multiple options (see Figure 1). Decisions made early in the process have important impacts on the information included in the synthesis, and meta-analysis may not always be possible or appropriate.\n\nSteps in evidence synthesis are to plan synthesis, explore data and conduct synthesis. Key issues examined in this study identified in italics. PICO = Population, Intervention, Comparator, Outcome.\n\nIn this study, we plan to examine two intertwined aspects of synthesis that commonly challenge authors of systematic reviews (identified in italics in Figure 1): approaches to planning how studies will be grouped for synthesis within the review (the ‘PICO (Population, Intervention, Comparator, Outcome) for each synthesis’); and the application of methods other than meta-analysis to summarise and synthesise results (hereafter described as ‘other synthesis methods’). There has been limited examination of the range of approaches used to define the PICO for each synthesis and which other synthesis methods are used in current practice. Yet, these are essential aspects of the synthesis in systematic reviews.\n\nRecent guidance published in the Cochrane Handbook for Systematic Reviews of Interventions2–4 has outlined proposed options in these two areas, but further research is required to understand current practice, investigate how review authors approach the PICO for synthesis and other synthesis methods, and assess the feasibility and impact of applying the proposed methods. We now expand on the concept of ‘PICO for each synthesis’ and describe summary and synthesis methods other than meta-analysis.\n\nIn reviews of the effects of interventions, authors commonly use the ‘PICO’ framework to prespecify the populations, interventions, comparators and outcomes that will be used to determine whether studies are eligible for the review5. While this definition of the ‘PICO for the review’ is viewed as a core component of a systematic review, more specific criteria are likely to be needed to define which groups of studies will contribute to each analysis within a review: the ‘PICO for each synthesis’. The PICO for each synthesis can be considered an operationalisation of the review objectives.\n\nThe process for defining the PICO for each synthesis ideally involves identifying characteristics (e.g. of the intervention or population) that may be expected to modify the intervention effect; clearly labelling and defining groups based on these characteristics (these may be based on an existing classification system if available); and planning how these groups will be used in synthesis and reporting. Groups may be analysed together in an overall synthesis, or they may be considered in separate syntheses4. Within an overall analysis, the defined groups may be used to explore any differences in the estimated effects (i.e., to explore statistical heterogeneity through the use of subgroup analysis). An example demonstrating the distinction between the PICO for the review and the PICO for each synthesis is presented in Box 1.\n\n\n\nIn a review of psychosocial interventions for smoking cessation6, the PICO for the review included any psychosocial intervention in pregnant women to help them stop smoking.\n\nOne of the objectives of the review was to examine “the effectiveness of the main psychosocial intervention strategies in supporting women to stop smoking in pregnancy (i.e. counselling, health education, feedback, social support, incentives, exercise)”. In order to meet this objective, a series of syntheses were presented within the review to assess the effects of each intervention strategy. So, for example, the PICO for the first synthesis presented included any counselling intervention for women during pregnancy compared to usual care, measuring the outcome of smoking abstinence in late pregnancy.\n\nAnother objective was to determine whether psychosocial interventions were effective in general. To address this objective, all intervention types were included in a single meta-analysis. Within this analysis, single, multi-component, and tailored interventions were presented as subgroups, to examine whether intervention effects were modified by having multiple or tailored components.\n\nProviding such definition has important advantages. Creating a consistent language to describe different groups or interventions can increase clarity of terminology for readers, as well as allowing authors to compare features between studies and make consistent, transparent decisions about grouping similar studies for inclusion in a synthesis3.\n\nMany systematic reviews examining the effects of health interventions use meta-analysis to combine the results of studies7,8. However, it is estimated that between 35% and 56% of systematic reviews do not use any meta-analysis7,8, and a larger percentage of reviews do not use meta-analysis for at least some outcomes. The reasons for not undertaking meta-analysis vary, but the most commonly reported reason is that the included studies do not report data that is amenable to meta-analysis7,9. For example, studies may report effect estimates without a measure of variance, or only report the direction of effect2.\n\nWhen meta-analysis is not possible, a range of summary and other synthesis methods are available. These methods include structured summaries of results, visual display options (e.g. harvest plots, albatross plots) and alternatives to meta-analysis such as combining P values or vote counting based on the direction of effect2,10. While these other synthesis methods provide more limited information for health care decision making, they may be preferable to textual description of the results in which there is a risk that authors may privilege the results of some studies over others without appropriate justification, possibly introducing bias7.\n\nImportantly, the use of other synthesis methods may alter the nature of the question answered by the review and the type of reasoning used to reach conclusions2,11.\n\nWe are unaware of other studies that have explicitly examined approaches to defining the PICO for each synthesis and planning comparisons. One cross-sectional study collected data on which PICO characteristics (e.g. population) were used to group studies for presentation or analysis within systematic reviews7. However, this study did not capture more detailed information on the basis of these groupings (e.g. was the population grouped by clinical disease characteristics, age or socioeconomic status), nor precisely how these groups were used in the synthesis.\n\nPrevious studies have examined the synthesis methods used in systematic reviews, and have estimated the percentage of reviews with and without meta-analysis8,9,12. One study examined systematic reviews of public health interventions that did not use meta-analysis in further detail7. They captured data on the use and reporting of “narrative” (text-based) synthesis and methods to investigate heterogeneity, but specific details of the synthesis methods used in the reviews were not captured. Another study examined the use of outcome groupings in synthesis and the use of methods other than meta-analysis, but the study was limited to Cochrane systematic reviews published before 201213.\n\nThe objectives of this study are to identify and describe current practice in systematic reviews of public health and health systems interventions in relation to:\n\n1. Approaches to grouping and definition of PICO characteristics for synthesis.\n\n2. Methods of summary and synthesis when meta-analysis is not used.\n\nHere we report the proposed methods for a cross-sectional study of a sample of systematic reviews.\n\n\nMethods\n\nWe will identify a sample of systematic reviews of public health or health systems interventions. We will identify and describe the methods used to define the PICO for each synthesis and the methods used to summarise and synthesise results, including meta-analysis and other methods. Two authors will undertake study selection. One author will undertake data extraction, and a second author will conduct independent data extraction from a subset of studies. Any amendments or additions to this protocol will be reported in resulting publications.\n\nWe will include systematic reviews that meet the following criteria:\n\n1. A study that aims to synthesise the results of primary studies, states eligibility criteria for inclusion of studies, and reports a search strategy to identify potentially eligible studies.\n\n2. Examines quantitative effects of any public health or health systems intervention, including policies, programs and strategies, as well as treatments and elements of care.\n\n3. Includes at least one comparison with at least two studies, where a comparison is defined as examining the effect on an outcome of an intervention compared with a specific alternative.\n\n4. Published in English.\n\nWe will exclude systemic reviews that:\n\n1. Synthesise the results of other systematic reviews, such as overviews of reviews.\n\n2. Answer questions that are not about effectiveness, for example prevalence, association, unplanned environmental exposures, prognosis, diagnosis and research methodology.\n\nOur criterion for deciding that a review is ‘systematic’ is intentionally inclusive compared to available definitions8,14,15. This is because we are explicitly interested in identifying systematic reviews with a range of methods, and not only those meeting a minimum standard of methods or reporting.\n\nOur focus is on systematic reviews of public health and health systems interventions. Reviews in these areas are likely to feature diversity in included populations and settings, as well as intervention complexity16. They are likely to include a range of study designs in addition to randomised trials, which in turn creates diversity in the effect measures used. Systematic reviews of public health and health systems interventions are more likely than other reviews to use synthesis methods other than meta-analysis7,8.\n\nFor reasons of feasibility, we will restrict the number of included reviews to 100. A sample of this size will allow us to estimate the proportion of reviews that use, for example, a particular synthesis or presentation method to within a maximum margin of error of 10%. This assumes a prevalence of 50%, but for a smaller or larger prevalence, the margin of error will be smaller. We anticipate that the proportion of reviews included in our sample that contain no meta-analyses will be approximately 50%7.\n\nRecords of all the systematic reviews published during 2018 will be obtained from two databases of systematic reviews: Health Systems Evidence and Health Evidence (see Table 1). These databases index systematic reviews of public health and health systems interventions, respectively.\n\nDescription of Health Evidence and Health Systems Evidence database content, and limits used to obtain cross-sectional sample of systematic reviews for this study.\n\nSome reviews identified by the search may have final citations outside 2018, for example arising from the difference between the date of online first publication and final publication in an issue of the journal, or the time lag between publication and indexing in a database. In these cases, the reference information will be updated to reflect the final citation, but reviews will not be excluded.\n\nThe records of systematic reviews retrieved from the two databases will initially be stored in Endnote and duplicate records removed. The selection and data extraction processes will then proceed using EPPI-Reviewer17. Reviews will be randomly selected from this larger set using EPPI-Reviewer’s random selection function, and screened for eligibility until our target sample of 100 is met.\n\nRecords will be independently screened by two authors (MC and one of SB or JM) based on the title and abstract, and any clearly ineligible records excluded. The full text of potentially eligible SRs will then be retrieved and assessed independently against the eligibility criteria by one author (MC). A second author (either SB or JM) will assess the full text of a sample of 20% of records. At each stage, we will resolve any disagreements by consensus, and consult a third author if consensus is not possible.\n\nFor each included systematic review, any protocol or registration record referred to in the review will be retrieved. In addition, protocols will be retrieved for any systematic reviews published in the Cochrane or Campbell Libraries, as they are a requirement of publication in these journals.\n\nWe will develop a data extraction form drawing on a previous methodological study that has examined synthesis and presentation methods used in systematic reviews13, as well as frameworks and methods outlined in relevant guidance2–4. The data extraction form will be piloted on a sample of included systematic reviews to identify items that are unclear or missing, and the form and data dictionary will be amended accordingly.\n\nOne author (MC) will extract data from all included reviews, and a second author (either SB or JM) will extract data independently on a sample of 20% of the included reviews (including those with and without meta-analysis). Any uncertainties or discrepancies arising during data extraction will be discussed with three authors (MC, SB, JM) and consensus reached. For any data items in which a high degree of inconsistency is observed, duplicate data extraction will be undertaken for a further random sample of reviews.\n\nWe will limit our data collection to information contained in the published report(s) of the SR, including protocols and registry records, and will not contact authors to obtain additional information.\n\nWe will collect data relating to the review characteristics, PICO characteristics used to group studies for each synthesis, and the synthesis methods used. Examples of data to be collected are presented in Table 2. The complete draft data dictionary is available online as Extended data20. Both explicit methods described in the review and implicit methods observed in textual descriptions, tables and figures will be coded. Both planned and implemented methods will be collected where these differ.\n\nExamples of data items to be collected from sample, including systematic review characteristics, PICO for each synthesis and summary and synthesis methods. PICO = Population, Intervention, Comparator, Outcome.\n\nWe will calculate descriptive summary statistics of features of the reviews (e.g. the synthesis and presentation methods used). For dichotomous or categorical data, we will calculate percentages and frequencies. For continuous or count data, we will calculate the means (with standard deviations) and medians (with interquartile ranges). We will examine whether approaches used to group the PICO for each synthesis are associated with the type of synthesis method by calculating differences in percentages between groups with 95% confidence intervals. Data will be tabulated and summarised in figures. Analyses will be undertaken using STATA21.\n\nThe findings of the research outlined in this protocol will be published. Associated datasets, data collection forms and analyses not included in any publication will be made publicly available via an online repository.\n\nAt submission of this protocol, the search had been conducted and screening of abstracts completed. Full text screening and piloting of the data extraction form was in progress.\n\n\nDiscussion\n\nIn this review, we will examine the methods choices for two intertwined elements of synthesis in systematic reviews. Namely, the approaches used to define and group PICO characteristics, and the types of synthesis methods other than meta-analysis. The results from our review will provide a snapshot of these practices, and highlight where improvements may be required in the application and reporting of the methods. Further, the study will provide a baseline assessment prior to release of recent guidance published in the Cochrane Handbook of Systematic Reviews of Interventions2–4, against which future assessments can be compared.\n\nThere are several strengths to our study. Our sample of systematic reviews is likely to be representative of public health and health systems intervention reviews because the source databases from which we will select our sample, and our inclusion criteria, place no restrictions on the intervention type or other features of the systematic reviews (e.g. the type of included study designs). A further strength is that our data extraction items are based on pre-existing frameworks to classify both the PICO groupings and methods of summary and synthesis. This will ensure that we are capturing specific methods and enhance the consistency of our data extraction.\n\nThere are some possible limitations in our proposed methods. For some items, the sample size may not be large enough to yield precise estimates of the percentage of systematic reviews that use particular methods. In addition, we will not undertake independent full text screening and data extraction of all studies by two authors, leaving some risk that data will be missed or misclassified.\n\nWhen complete, the findings of this study will be published and communicated at conferences, in addition to dissemination through international networks of researchers and authors of methodological guidance in the field of systematic reviews.\n\nAuthors of systematic reviews face challenges in the organisation and analysis of data, including the complexity of grouping studies for comparison, and synthesis methods when meta-analysis is not available. This protocol outlines the methods for a cross-sectional study that aims to examine the approaches used to define and group PICO characteristics, and the types of synthesis methods other than meta-analysis in a sample of systematic reviews of public health and health services interventions.\n\n\nData availability\n\nNo underlying data are associated with this article.\n\nFigshare (Monash University repository, known as Bridges): Draft data dictionary for cross-sectional study of current practice in systematic reviews including the ‘PICO for each synthesis’ and methods other than meta-analysis. https://doi.org/10.26180/5edb178961d6820.\n\nFigshare (Monash University repository, known as Bridges): PRISMA-P reporting checklist for protocol of cross-sectional study of current practice in systematic reviews including the ‘PICO for each synthesis’ and methods other than meta-analysis. https://doi.org/10.26180/5edb35183074f22.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Acknowledgements\n\nWe acknowledge the assistance of Kristin Read, Research Coordinator at Health Evidence™, McMaster University, in providing access to search results from that database.\n\n\nReferences\n\nMcKenzie JE, Beller EM, Forbes AB: Introduction to systematic reviews and meta-analysis. Respirology. 2016; 21(4): 626–37. PubMed Abstract | Publisher Full Text\n\nMcKenzie J, Brennan S: Chapter 12: Synthesizing and presenting findings using other methods. In: Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M, et al., editors. Cochrane Handbook for Systematic Reviews of Interventions. 2nd ed. Chichester (UK): John Wiley & Sons; 2019. Publisher Full Text | Free Full Text\n\nMcKenzie J, Brennan S, Ryan R, et al.: Chapter 9: Summarizing study characteristics and preparing for synthesis. In: Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M, et al., editors. Cochrane Handbook for Systematic Reviews of Interventions. 2nd ed. Chichester (UK): John Wiley & Sons; 2019. Publisher Full Text | Free Full Text\n\nMcKenzie J, Brennan S, Ryan R, et al.: Chapter 3: Defining the criteria for including studies and how they will be grouped for the synthesis. In: Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M, et al., editors. Cochrane Handbook for Systematic Reviews of Interventions. 2nd ed. Chichester (UK): John Wiley & Sons; 2019. Publisher Full Text | Free Full Text\n\nThomas J, Kneale D, McKenzie J, et al.: Chapter 2: Determining the scope of the review and the questions it will address. In: Cochrane Handbook for Systematic Reviews of Interventions. 2nd ed. Chichester (UK): John Wiley & Sons; 2019. Publisher Full Text | Free Full Text\n\nChamberlain C, O'Mara-Eves A, Porter J, et al.: Psychosocial interventions for supporting women to stop smoking in pregnancy. Cochrane Database Syst Rev. 2017; 2(2): CD001055. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCampbell M, Katikireddi SV, Sowden A, et al.: Lack of transparency in reporting narrative synthesis of quantitative data: a methodological assessment of systematic reviews. J Clin Epidemiol. 2019; 105: 1–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPage MJ, Shamseer L, Altman DG, et al.: Epidemiology and Reporting Characteristics of Systematic Reviews of Biomedical Research: A Cross-Sectional Study. PLoS Med. 2016; 13(5): e1002028. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIoannidis JPA, Patsopoulos NA, Rothstein HR: Reasons or excuses for avoiding meta-analysis in forest plots. BMJ. 2008; 336(7658): 1413–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHiggins JPT, López-López JA, Becker BJ, et al.: Synthesising quantitative evidence in systematic reviews of complex health interventions. BMJ Glob Health. 2019; 4(Suppl 1): e000858. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMelendez-Torres GJ, O'Mara-Eves A, Thomas J, et al.: Interpretive analysis of 85 systematic reviews suggests that narrative syntheses and meta-analyses are incommensurate in argumentation. Res Synth Methods. 2017; 8(1): 109–18. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaquette M, Alotaibi AM, Nieuwlaat R, et al.: A meta-epidemiological study of subgroup analyses in cochrane systematic reviews of atrial fibrillation. Syst Rev. 2019; 8(1): 241. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcKenzie J, Brennan S, Page M, et al.: From summary to synthesis: a review of statistical synthesis and presentation methods used in complex reviews [poster]. Better Knowledge for Better Health Un meilleur savoir pour une meilleure santé Abstracts of the 21st Cochrane Colloquium; 2013 19–23 September 2013; Québec City, Canada: John Wiley & Sons. 2013. Reference Source\n\nKrnic Martinic M, Pieper D, Glatt A, et al.: Definition of a systematic review used in overviews of systematic reviews, meta-epidemiological studies and textbooks. BMC Med Res Methodol. 2019; 19(1): 203. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoher D, Liberati A, Tetzlaff J, et al.: Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009; 6(7): e1000097. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAnderson LM, Petticrew M, Chandler J, et al.: Introducing a series of methodological articles on considering complexity in systematic reviews of interventions. J Clin Epidemiol. 2013; 66(11): 1205–8. PubMed Abstract | Publisher Full Text\n\nThomas J, Brunton J, Graziosi S: EPPI-Reviewer Web: software for research synthesis. London: EPPI-Centre Software. Social Science Research Unit, UCL Institute of Education; 2020. Reference Source\n\nHealth Evidence: Health Evidence. McMaster University; 2019. Reference Source\n\nHealth Systems Evidence: About HSE. McMaster University; 2019. Reference Source\n\nCumpston MS, McKenzie JE, Brennan SE: Draft data dictionary for cross-sectional study of current practice in systematic reviews including the ‘PICO for each synthesis’ and methods other than meta-analysis. Bridges: Monash University; 2020. http://www.doi.org/10.26180/5edb178961d68\n\nStataCorp: Stata Statistical Software: Release 15. College Station, TX: StataCorp LLC; 2017.\n\nCumpston MS, McKenzie JE, Brennan SE: PRISMA-P reporting checklist for protocol of cross-sectional study of current practice in systematic reviews including the ‘PICO for each synthesis’ and methods other than meta-analysis. Bridges: Monash University; 2020. http://www.doi.org/10.26180/5edb35183074f"
}
|
[
{
"id": "69547",
"date": "09 Sep 2020",
"name": "Livia Puljak",
"expertise": [
"Reviewer Expertise Research methodology",
"evidence synthesis",
"clinical epidemiology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study proposed is interesting and relevant. The authors should be congratulated for this initiative. My only concern is the sample size. I am aware that the authors have indicated that 100 reviews are proposed to be analyzed due to feasibility reasons, but 100 reviews in this particular context appears like a very small sample to analyze heterogeneity of approaches that could be expected. So my concern is about generalizability of the results with such a small sample. It is unclear how many reviews will be found by the proposed search methods, and what percent of reviews will be analyzed when the sample of reviews is 100.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes",
"responses": [
{
"c_id": "6094",
"date": "05 Nov 2020",
"name": "Miranda Cumpston",
"role": "Author Response",
"response": "We would like to thank Professor Puljak for her review of our study protocol. In the following, we respond to the two issues raised regarding sample size and generalisability. In this cross-sectional study we are using survey sampling methodology to sample the reviews (not aiming to locate and include all reviews meeting our eligibility criteria). Random (or probability) sampling allows valid estimates of population characteristics to be made (e.g. estimating the prevalence of reviews that use vote counting) along with calculation of uncertainty in those estimates [1]. The estimate of uncertainty (margin of error) is unchanged by the population size (total number of reviews), except in the circumstance where the fraction of reviews sampled is greater than approximately 5% of the population of reviews, in which case the margin of error will be appreciably less than we have specified in the protocol. We believe that the sample size of 100 reviews will be generally be sufficient such that the interpretation of the confidence limits will be consistent. For example, if we estimate the percentage of reviews using a particular synthesis method (e.g. vote counting based on the direction of effect) is 15%, the associated 95% confidence limits would be 8% to 22%, and our interpretation of these limits would lead to the same conclusion that the method was not commonly used. As another example, if the outcome was ‘reviewers fully specified the groups to be used in the synthesis’, and 50% of the reviewers were found to fully specify the groups, our interpretation would not differ at the limits of the confidence interval, which in this example would be from 40% to 60%. As noted in the protocol, the maximum margin of error will be 10%, which will occur for an estimated prevalence of 50%. In regard to generalisability (external validity), which we interpret as the extent to which our findings will be applicable to systematic reviews other than those examining public health or health systems interventions, we would be reluctant to generalise to other review types (e.g. prognostic reviews, diagnostic test accuracy reviews). However, importantly, the generalisability of study findings is not dependent on the sample size or statistical power, but on considering whether the reviews included in our study have characteristics that differ from another population of reviews to which the reader might wish to generalise the findings [2]. In reporting the findings of this study, we will ensure that the characteristics of the included studies are clearly summarised to facilitate such judgements by the reader. References: 1. Kish, L., Survey sampling. 1965, New York: John Wiley & Sons, Inc. 2. Kukull, W.A. and M. Ganguli, Generalizability: the trees, the forest, and the low-hanging fruit. Neurology, 2012. 78(23): p. 1886-91."
}
]
},
{
"id": "73015",
"date": "05 Nov 2020",
"name": "Paul Whaley",
"expertise": [
"Reviewer Expertise Systematic review methods in environmental health research."
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGeneral comments The authors present a protocol for a planned systematic review of meta-analysis and \"other\" methods for synthesising evidence in systematic reviews of public health and health systems interventions. It is definitely a novel, interesting, and worthwhile project of potentially very high value - I find myself in full agreement with the authors that \"other\" methods are underutilised and underdeveloped in SRs and a survey of current practices will provide valuable evidence for improvements in this area.\nHowever, I am not convinced that the proposed approach is the finished article. Besides some relatively trivial issues with the clarity of the objectives, I am particularly concerned about the lack of detail on how the authors are going to classify the methods employed in the included SRs. There is no comprehensive code book presented, so it is not possible to judge if the authors are gathering information which is going to optimally feed their intended analysis. The authors also discuss how various elements of the methodology will be piloted - but since this will resolve the issues with the protocol, I would strongly advocate that piloting being conducted now, and the protocol revised on the basis of that. I would definitely like to see a piloted code-book presented as part of the protocol.\nI think this is very promising work, just incomplete. It is also very challenging to communicate some of the issues as written feedback. I hope what I write below makes sense and is fair (or at least is constructive). If I over-explain in places, it is not because I am assuming the authors to be ignorant of their own subject matter but am attempting to provide context for a reader of the published peer-review. Since this review is not anonymous, I am sure the authors can track down my contact details easily enough. I would be more than happy to discuss my comments over the phone if they had any questions.\nSpecific comments\nI am not completely sure if the title clearly and succinctly captures exactly what the review is about. For example, I see why the authors call it a \"cross-sectional study\", but it seems an odd occasion on which to use the term. Would it be plainer to call it something like: \"A survey of current evidence synthesis practices in quantitative systematic reviews of public health and health systems interventions: study protocol\" (I'm sure the authors can improve on this!).\n\nThe abstract is quite difficult to follow - I had to read the paper in order to really grasp what the abstract was about. The authors might want to reconsider how they are summarising their research, maybe focusing on how PICOs structure SRs and syntheses in the introduction. They should clearly state the objective (it's too compressed at the moment). The Methods section gives a lot of relatively trivial information about the extraction strategy, yet only has once sentence for the difficult and interesting part, which is the methodology for comparing approaches (what does it mean by their intent to \"compare approaches\" here? The summary seems too compressed.).\n\nI wonder if the concept of \"synthesis\" is clearly enough introduced, or if it could be more directly introduced and defined at the outset. Currently, the concept is introduced after \"however\" in sentence 3 of para 1, contextualised by the claim of synthesis methods as often being narrowly considered. This seems back-to-front - would it be better to state what \"synthesis\" means, qualify this as being broader than some understandings, and then introduce the role of the PICO in structuring evidence synthesis (whether narrative or quantitative).\n\nI am not sure about the terminology in splitting synthesis into meta-analysis vs. \"other synthesis methods\". There is a rich history of development and analysis of \"other\" methods, which I am sure the authors are aware of but they could perhaps use more fully. Often, the difference is defined as quantitative vs. either non-quantitative methods, narrative synthesis methods, or synthesis without meta-analysis (\"SWIM\"). This feels like contextual information that would be useful to the reader (given the author's assumption that there is a lack of awareness of this in the discipline) and might result in a choice of phrasing which better reflects established conventions.\n\nThis is a more trivial point. The authors state \"Recent guidance published in the Cochrane Handbook for Systematic Reviews of Interventions has outlined proposed options in these two areas\" - it would be helpful if these other areas were briefly indicated.\n\nI am not sure if the discussion of reasons for not conducting meta-analysis is sufficiently nuanced. Certainly, lack of numeric data would challenge meta-analysis, but is a common issue not also excess heterogeneity between study designs? (Maybe this is less important for clinical SRs than it is in my field of environmental health.) Either way, in such cases narrative or SWIM methods come to the fore. They are complex, and I worry they are not sufficiently acknowledged in the rather broad statement that various synthesis methods such as harvest plots, etc. \"may be preferable to textual description of the results in which there is a risk that authors may privilege the results of some studies over others without appropriate justification, possibly introducing bias\". Echoing an earlier comment, I feel that this skates over some quite well-developed narrative or SWIM methodology which seeks to counter these concerns when providing textual summarisation. Introducing more of this theory I think is important for the protocol, coding the included SRs, and interpreting the methods in those SRs. (Note I use the term “narrative” in the sense of synthesis without quantitative methods, not in the sense of a narrative vs. systematic review.)\n\nI wonder if it is worth the authors clarifying up-front that this is a survey of SRs of quantitative data, just to be clear that qualitative research is not the target of investigation. I think I only resolved this at the point of reading the eligibility criteria. Likewise the focus on public health and health systems interventions.\n\nI am not clear as to how are the authors going to code and classify \"other\" methods? The authors have not provided a code book, so the utility and validity of their coding methodology cannot be evaluated. It seems like a potentially complex challenge, particularly if studies do not define their synthesis methods (it may be impossible for them to do so if there is no agreed way of categorising PICO-based synthesis approaches). The code definitions and criteria should be defined as far as possible in advance. I also note that in Table 2 \"Examples of data collection items\" that for summary and synthesis methods that the examples are almost exclusively related to quantitative methods and few examples of SWIM methods are given. This suggests to me that more planning around classifying these \"other\" methods is needed, to anticipate what they are and how they are accommodated in the analysis.\n\nIn terms of providing a code book, as handling editor, I gave the authors of this protocol a similarly hard time as I am giving the present authors. Table A3 may be instructive. https://doi.org/10.1016/j.envint.2020.105826.1\n\nI am also unclear as to how the concept of \"PICO for each synthesis\" is operationalised in the data extraction methods and the synthesis approach the authors will themselves be using. Introducing this concept seems a central concern of the introduction, yet by the methods section, it seems to have faded into the background. How is this concept going to structure the extraction and analysis, such that after conducting this survey we know how authors of SRs are using PICO statements in developing the synthesis components of their SRs?\n\nI am not convinced that having just one person coding such complex data is sufficient. Coding decisions are likely to be difficult and require discussion. The authors seem to acknowledge this by requiring that 20% of extraction and coding be double-checked, but they provide no plan for what to do if that double-checking finds inconsistency. Is 80% agreement enough? Will the second extractor be trained to the same level as the first? If not, what does disagreement between the primary and secondary reviewer even mean? If agreement is low, do they intend to revise the coding criteria or seek to resolve disagreement? Will they double-check everything to ensure consistency across the full set of extracted data? I suspect it would be simpler, if more time-consuming, to train two extractors to an equal degree, code in parallel, then discuss the results to achieve considered consistency.\n\nIn terms of piloting, I would be much more comfortable if this was conducted and reported as part of the protocol development process. Since piloting is part of planning, I am not personally of the view that it is sufficient to state in a protocol that something will be piloted. Doing the piloting now would also help answer quite a few of the questions I have above.\n\nIs the rationale for, and objectives of, the study clearly described? Partly\n\nIs the study design appropriate for the research question? Partly\n\nAre sufficient details of the methods provided to allow replication by others? Partly\n\nAre the datasets clearly presented in a useable and accessible format? Not applicable",
"responses": [
{
"c_id": "6280",
"date": "29 Jan 2021",
"name": "Miranda Cumpston",
"role": "Author Response",
"response": "We thank the reviewer for their constructive feedback and suggestions. We have found the reviewer’s perspective particularly valuable for understanding where more contextual background and methods detail is needed and have revised the text to address the concerns raised. Our response to each comment is presented below. General comments The authors present a protocol for a planned systematic review of meta-analysis and \"other\" methods for synthesising evidence in systematic reviews of public health and health systems interventions. It is definitely a novel, interesting, and worthwhile project of potentially very high value - I find myself in full agreement with the authors that \"other\" methods are underutilised and underdeveloped in SRs and a survey of current practices will provide valuable evidence for improvements in this area. Response: We thank the reviewer for these comments. Specific comments 1. I am not completely sure if the title clearly and succinctly captures exactly what the review is about. For example, I see why the authors call it a \"cross-sectional study\", but it seems an odd occasion on which to use the term. Would it be plainer to call it something like: \"A survey of current evidence synthesis practices in quantitative systematic reviews of public health and health systems interventions: study protocol\" (I'm sure the authors can improve on this!). Response: We used the term ‘cross-sectional’ based on the title of other studies that have also used survey sampling methodology to sample the reviews (e.g. Page, et al.1). However, we agree with the reviewer’s suggestion that ‘survey’ more accurately captures the study design. In addition, we have revised the title to be plainer, while still capturing the specific systematic review practices to be examined. Changes made: The revised title is: “The use of ‘PICO for synthesis’ and methods for synthesis without meta-analysis: protocol for a survey of current practice in systematic reviews of health interventions” 2. The abstract is quite difficult to follow - I had to read the paper in order to really grasp what the abstract was about. The authors might want to reconsider how they are summarising their research, maybe focusing on how PICOs structure SRs and syntheses in the introduction. They should clearly state the objective (it's too compressed at the moment). The Methods section gives a lot of relatively trivial information about the extraction strategy, yet only has once sentence for the difficult and interesting part, which is the methodology for comparing approaches (what does it mean by their intent to \"compare approaches\" here? The summary seems too compressed.). Response: We have revised the Abstract to describe our areas of interest more specifically. We have increased the level of detail in the Objectives and Methods, while acknowledging the constraint of the Abstract word limit. 3. I wonder if the concept of \"synthesis\" is clearly enough introduced, or if it could be more directly introduced and defined at the outset. Currently, the concept is introduced after \"however\" in sentence 3 of para 1, contextualised by the claim of synthesis methods as often being narrowly considered. This seems back-to-front - would it be better to state what \"synthesis\" means, qualify this as being broader than some understandings, and then introduce the role of the PICO in structuring evidence synthesis (whether narrative or quantitative). Response: Reversing the order seemed reasonable, but after making this change, we realised that it did not work well. Synthesis is typically used to refer to the meta-analysis in reviews of health interventions, whereas the broader concept of synthesis that we use in this study is new, so it seemed best to retain an order in which we build from the status quo to the newer concept that is central to our study. 4. I am not sure about the terminology in splitting synthesis into meta-analysis vs. \"other synthesis methods\". There is a rich history of development and analysis of \"other\" methods, which I am sure the authors are aware of but they could perhaps use more fully. Often, the difference is defined as quantitative vs. either non-quantitative methods, narrative synthesis methods, or synthesis without meta-analysis (\"SWIM\"). This feels like contextual information that would be useful to the reader (given the author's assumption that there is a lack of awareness of this in the discipline) and might result in a choice of phrasing which better reflects established conventions. Response: The split between meta-analysis and ‘other methods’ mirrors that in our work in the Cochrane Handbook for Systematic Reviews of Interventions,2 but we now see that ‘other methods’ is too vague in this context. We have revised the heading to be more explanatory (see list of changes below). Terminology is something that we have debated at length with our co-authors as contributors to the Cochrane Handbook2-6 and the SWiM reporting guidance.7 We agree that it is important to follow convention, but in our view, there is a need to increase conceptual clarity in this area through clearer delineation of the steps and methods captured under the umbrella of ‘narrative synthesis’ (and synonyms such as ‘non-quantitative’ and SWiM). Used correctly, these terms are broader than our concept of ‘other synthesis methods’ where we refer to the methods used to present and aggregate quantitative results from two or more studies (for clarification of scope, see response to Comment 8). For this reason, we don’t use ‘narrative synthesis’ (or synonyms) for these methods, but have made revisions to address these conceptual issues. Changes made: In the paragraph starting “Recent guidance …” (Introduction, para 3) we have added and referenced the following “Reporting guidance for ‘synthesis without meta-analysis’ (SWiM) has also been published …”7 At the end of the section on ‘PICO for synthesis’, we have added the following paragraph: “The PICO for synthesis also provides a framework for examining similarities and differences in the characteristics of studies contributing to each analysis, facilitating qualitative synthesis of characteristics needed to interpret results. This qualitative synthesis is a particularly import feature of reviews where there is diversity in study characteristics that may explain findings (e.g. intervention complexity, different study designs).6 Such diversity sometimes triggers a decision not to use meta-analysis, and instead adopt alternative methods to synthesise and present findings. In these circumstances, it is common for authors to refer to their synthesis methods as ‘narrative’,8 reflecting the integration of the synthesis of quantitative results from studies with the qualitative synthesis of study characteristics. In this study, we distinguish between these elements and, in the section that follows, focus on the methods used to combine quantitative data on intervention effects using a statistical technique and to present the results of these analyses.” Changed section heading in Introduction from ‘Other synthesis methods’ to ‘Synthesising and presenting findings without meta-analysis’. Note that we include the term ‘presenting’ to encompass structured summary of individual study results and visual presentation of data. 5. This is a more trivial point. The authors state \"Recent guidance published in the Cochrane Handbook for Systematic Reviews of Interventions has outlined proposed options in these two areas\" - it would be helpful if these other areas were briefly indicated. Changes made: “Recent guidance published in the Cochrane Handbook for Systematic Reviews of Interventions2-4 has outlined proposed methods for specifying the PICO for each synthesis and a range of other synthesis methods.” 6. I am not sure if the discussion of reasons for not conducting meta-analysis is sufficiently nuanced. Certainly, lack of numeric data would challenge meta-analysis, but is a common issue not also excess heterogeneity between study designs? (Maybe this is less important for clinical SRs than it is in my field of environmental health.) Either way, in such cases narrative or SWIM methods come to the fore. They are complex, and I worry they are not sufficiently acknowledged in the rather broad statement that various synthesis methods such as harvest plots, etc. \"may be preferable to textual description of the results in which there is a risk that authors may privilege the results of some studies over others without appropriate justification, possibly introducing bias\". Echoing an earlier comment, I feel that this skates over some quite well-developed narrative or SWIM methodology which seeks to counter these concerns when providing textual summarisation. Introducing more of this theory I think is important for the protocol, coding the included SRs, and interpreting the methods in those SRs. (Note I use the term “narrative” in the sense of synthesis without quantitative methods, not in the sense of a narrative vs. systematic review.) Response: Regarding the point about diversity (heterogeneity), we agree that this is a common reason given for not using meta-analysis, so we have addressed this in the manuscript (see list of changes below). Regarding consideration of narrative/SWiM methods, as the reviewer notes, this issue relates closely to that covered in our response to Comment 4, where we consider the scope of narrative synthesis in relation to our project. We agree that there is benefit in expanding on these conceptual issues, as done in the new paragraph (see Comment 4). We provide further clarification of our perspective below. It is not our intention to minimise the complexity of methods encompassed by the concept of ‘narrative synthesis’. Instead, we believe that the components and process of narrative synthesis need to be disentangled to provide clear guidance for review authors on how to plan, conduct and report their methods. Following the lead of authors such as Melendez Torres,9 and our own work on the Cochrane Handbook2-5 and SWiM,7 we see three main components commonly conceived as part of a narrative synthesis: the planning work done to decide how studies will be grouped to address the review questions (the PICO for each synthesis), the qualitative analysis of study characteristics (PICO/PECO and study design features) done to prepare for synthesis, interpret and explain the quantitative synthesis findings, and the methods used to present and synthesise quantitative results from studies. Our study will examine (1) and (3) of the above. It is not within the scope of this study to examine (2) or the process by which authors integrate the findings from (2) and (3). Although these latter elements are common in reviews reporting ‘narrative synthesis’, in our view they are essential features of any review where there is diversity (and complexity) of study characteristics that must be considered in order to identify and explain variation in effects across studies. While this may be described as the ‘narrative’ synthesis that ‘tells the story’ of the evidence, it can be done irrespective of whether meta-analysis is used or not. Changes made: Addition of new paragraph addressing the concept of narrative synthesis and indicating that diversity of study characteristics may be a trigger for not using meta-analysis (as per comment 4) (‘PICO for each synthesis, final para). Added the following to address diversity: “Diversity of study characteristics and the presence of statistical heterogeneity are other reasons given for not meta-analysing, but these are more contentious. The first brings into question whether any synthesis is appropriate, the second may be addressed by using extensions to meta-analysis (e.g. meta-regression, prediction intervals) that attempt to explain or encompass heterogeneity.” (‘Synthesising and presenting findings without meta-analysis’, para 1) Edited the sentence that reads ‘…textual description of the results …’ to ‘Nevertheless, structured summary or synthesis approaches may be preferable to simply presenting an unstructured inventory of study-level results…’. We made this change to avoid any misunderstanding that by ‘textual description’ we are referring to narrative synthesis. (‘Synthesising and presenting findings without meta-analysis’, para 3) Amended our text to distinguish between structured summaries of the results of individual studies, and more unstructured summaries that are common in reviews. 7. I wonder if it is worth the authors clarifying up-front that this is a survey of SRs of quantitative data, just to be clear that qualitative research is not the target of investigation. I think I only resolved this at the point of reading the eligibility criteria. Likewise the focus on public health and health systems interventions. Response: We have revised the text throughout to specify our focus on reviews of quantitative data and public health and services interventions, including the Abstract and Objectives. 8. I am not clear as to how are the authors going to code and classify \"other\" methods? The authors have not provided a code book, so the utility and validity of their coding methodology cannot be evaluated. It seems like a potentially complex challenge, particularly if studies do not define their synthesis methods (it may be impossible for them to do so if there is no agreed way of categorising PICO-based synthesis approaches). The code definitions and criteria should be defined as far as possible in advance. I also note that in Table 2 \"Examples of data collection items\" that for summary and synthesis methods that the examples are almost exclusively related to quantitative methods and few examples of SWIM methods are given. This suggests to me that more planning around classifying these \"other\" methods is needed, to anticipate what they are and how they are accommodated in the analysis. In terms of providing a code book, as handling editor, I gave the authors of this protocol a similarly hard time as I am giving the present authors. Table A3 may be instructive. https://doi.org/10.1016/j.envint.2020.105826.1 Response: In addition to the example data collection items presented in the table, we provided a link to our draft data dictionary as Extended Data in the original version of this paper (‘Data extraction and management’ section and Data Availability statement, see https://doi.org/10.26180/5edb178961d68). The data dictionary provides the items, their response options, and a description of the items along with some guidance on coding. The dictionary covers both the data items relevant to the PICO for each synthesis (sections 2-6) and synthesis, summary and presentation methods (section 7). We believe the information provided in our data dictionary is as comprehensive as the example suggested by the reviewer. As noted in our response to Comments 4 and 6, our focus is on a subset of methods considered within the broader concept of narrative synthesis (or SWiM). Specifically, the planning of groups for synthesis (PICO for synthesis) and the methods used to present and synthesise quantitative results. The items we have included to capture the synthesis, summary and presentation methods are aligned with the methods we have outlined in Chapter 12 of the Cochrane Handbook, ‘Synthesizing and presenting findings using other methods’2 (https://training.cochrane.org/handbook/current/chapter-12#section-12-2). We have successfully used these items in a previous study.10 Changes made: Moved the reference to the data dictionary to the first paragraph of the ‘Data extraction and management’ section to increase its prominence. Added additional detail of the data items in Table 1 (note that the former Table 2 has been renumbered as it is now referenced earlier in the text), and added a reference to the full data dictionary in the table’s caption. Added the following paragraph to the ‘Data extraction and management’ section: “In seeking to map current practice, we note that terms such as ‘narrative synthesis’ can be applied to a wide range of approaches, and will seek to identify specific components in our included reviews rather than relying on broad descriptive terms. We will collect: Sources of guidance referred to in the text. Methods of summary and synthesis explicitly specified in the Methods section. Methods of summary and synthesis used in practice (whether named or used implicitly in the text). Specific elements that may appear within a text-based summary approach, such as the use of consistent effect measures across studies, the use of non-parametric summary statistics such as ranges, various methods of vote counting, and the use of PICO groupings to structure text or tables. Explicit statements by the authors that they have been unable to implement planned PICO groupings or synthesis methods, their stated reasons for this, and what changes they made to their methods in response.” 9. I am also unclear as to how the concept of \"PICO for each synthesis\" is operationalised in the data extraction methods and the synthesis approach the authors will themselves be using. Introducing this concept seems a central concern of the introduction, yet by the methods section, it seems to have faded into the background. How is this concept going to structure the extraction and analysis, such that after conducting this survey we know how authors of SRs are using PICO statements in developing the synthesis components of their SRs? Response: Details of how the PICO for synthesis is operationalised are summarised in Table 1, with full details provided in the draft data dictionary. We will undertake a descriptive analysis to characterise the extent to which authors specify their PICO for synthesis, and the basis for their PICO. For example, the percentage of reviews where intervention groups are listed by name, are defined in enough detail for replication, and have an explicit role in the planned synthesis. Similarly, these percentages will be calculated for the other PICO elements. For each PICO element and study design, we capture any groupings identified, whether an explicit role in synthesis is reported, whether the groupings were specified at the level of detail required for replication, whether the specified groupings were used in practice, whether additional new groupings were used in practice that were not specified, and statements by the authors describing any change in the planned groupings. In addition, we describe the basis of the groupings (such as whether groupings are based on existing taxonomies, or whether they are based on e.g. disease categories, health equity characteristics, time of measurement, etc.) using categories drawn from the frameworks provided for each PICO element in Chapter 3, Section 3.2 of the Cochrane Handbook for Systematic Reviews of Interventions4 (available free online at https://training.cochrane.org/handbook/current/chapter-03#section-3-2). Changes made: Added additional detail of the data items collected to examine the PICO for each synthesis in Table 1 (previously Table 2). Added the following text to the analysis section “We will calculate descriptive summary statistics to characterise the extent to which authors specify their PICO for synthesis, and the synthesis and presentation methods. For example, the percentage of reviews where intervention groups are listed by name, are defined in enough detail for replication, and have an explicit role in the planned synthesis. Similarly, these percentages will be calculated for the other PICO elements.” 10. I am not convinced that having just one person coding such complex data is sufficient. Coding decisions are likely to be difficult and require discussion. The authors seem to acknowledge this by requiring that 20% of extraction and coding be double-checked, but they provide no plan for what to do if that double-checking finds inconsistency. Is 80% agreement enough? Will the second extractor be trained to the same level as the first? If not, what does disagreement between the primary and secondary reviewer even mean? If agreement is low, do they intend to revise the coding criteria or seek to resolve disagreement? Will they double-check everything to ensure consistency across the full set of extracted data? I suspect it would be simpler, if more time-consuming, to train two extractors to an equal degree, code in parallel, then discuss the results to achieve considered consistency. Response: We recognise that a potential limitation of the review is that we will not undertake double data extraction of all reviews; this is due to limited resources. However, we are investing substantial time and effort in the piloting stage to refine the items and guidance, and achieve a shared understanding of the form (see below for the changes made in this regard). The team undertaking the data extraction have extensive experience in systematic reviews of public health and health services interventions, having written guidance for, co-authored, and edited many such reviews. While we appreciate the experience of the team does not mitigate missing information in the papers (a point which we note in the discussion of the revised protocol), it does help with making the judgments required in the data extraction. Finally, we believe that the consequences of some errors in the data extraction in a methodological review such as this are less serious than that of reviews examining the effects of interventions and environmental exposures, where misleading results arising from errors in the data extraction can impact policy decisions, patients’ health outcomes and quality of care. Changes made: We have revised the ‘Data extraction and management’ section of the paper to provide more detail on our proposed approach: “One author (MC) will extract data from all included reviews, and a second author (either SB or JM) will extract data independently on a sample of 20% of the included reviews (including those with and without meta-analysis). We will pilot test the data extraction form and coding guidance on five reviews to ensure we capture all items, to refine the items and guidance when we uncover ambiguity or a lack of clarity, and to achieve a shared understanding of the form. This will be achieved using an iterative process, where we discuss discrepancies and ambiguities as extraction is completed on each review, and revise the data extraction form and coding guidance in response to these discussions. Duplicate data extraction on the selected sample will then proceed, and agreement will be assessed at the end of this phase. For any data items in which a high degree of inconsistency is observed, duplicate data extraction will be undertaken for a further random sample of reviews. During the final, single data extraction phase, any uncertainties arising will be discussed with three authors (MC, SB, JM) and consensus reached.” (para 3) We have added the following text to the Discussion: “However, the review team has extensive experience in systematic reviews of public health and health services interventions, having written guidance for, co-authored, and edited many such reviews. While this will not mitigate missing information in the papers, it will help with making judgments required in the data extraction. Given that the aim of our study is to gain a broad understanding of current practice, we think this is unlikely to have an important impact on conclusions.” (para 3) 11. In terms of piloting, I would be much more comfortable if this was conducted and reported as part of the protocol development process. Since piloting is part of planning, I am not personally of the view that it is sufficient to state in a protocol that something will be piloted. Doing the piloting now would also help answer quite a few of the questions I have above. Response: We agree that piloting is an important aspect of a systematic review. Aspects of the data extraction for the present study (i.e. other synthesis methods items) were based on a previous study we undertook, where we had high concordance in coding across reviewers (including both experienced methodologists, and those with limited experience). We note that the current data extraction form is based on a previous study in the ‘Data extraction and management’ section. We note that the reviewer suggests that piloting should be conducted and reported as part of the protocol development process. We are not aware that there is an established convention for this. For example, it is not a requirement in Cochrane review protocols – which outline the methods for reviews evaluating the effects of healthcare interventions – to report a completed piloting process. Our protocol as submitted provides our planned methods at a point in time. We will report changes to those methods in our final report, including the finalised data dictionary arising from the piloting. We have used this approach in other methods reviews.11-13 For example, in Arnup et al,11 we undertook a review examining statistical methods used in cluster-randomised crossover trials. In a supplementary file to that paper, Table S1, we outlined changes in methods from protocol, and included the final data extraction form (which incorporated modifications from piloting). References 1. Page MJ, Shamseer L, Altman DG, Tetzlaff J, Sampson M, Tricco AC, et al. Epidemiology and Reporting Characteristics of Systematic Reviews of Biomedical Research: A Cross-Sectional Study. PLOS Medicine. 2016;13(5):e1002028. 2. McKenzie J, Brennan S. Chapter 12: Synthesizing and presenting findings using other methods. In: Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M, et al., editors. Cochrane Handbook for Systematic Reviews of Interventions. 2nd ed. Chichester (UK): John Wiley & Sons; 2019. 3. McKenzie J, Brennan S, Ryan R, Thomson H, Johnston R. Chapter 9: Summarizing study characteristics and preparing for synthesis. . In: Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M, et al., editors. Cochrane Handbook for Systematic Reviews of Interventions. 2nd ed. Chichester (UK): John Wiley & Sons; 2019. 4. McKenzie J, Brennan S, Ryan R, Thomson H, Johnston R, Thomas J. Chapter 3: Defining the criteria for including studies and how they will be grouped for the synthesis. In: Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M, et al., editors. Cochrane Handbook for Systematic Reviews of Interventions. 2nd ed. Chichester (UK): John Wiley & Sons; 2019. 5. Thomas J, Kneale D, McKenzie J, Brennan S, Bhaumik S. Chapter 2: Determining the scope of the review and the questions it will address. In: Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M, et al., editors. Cochrane Handbook for Systematic Reviews of Interventions. 2nd ed. Chichester (UK): John Wiley & Sons; 2019. 6. Thomas J, Petticrew M, Noyes J, Chandler J, Rehfuess E, Tugwell P, et al. Chapter 17: Intervention complexity. In: Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M, et al., editors. Cochrane Handbook for Systematic Reviews of Interventions. 2nd ed. Chichester (UK): John Wiley & Sons; 2019. 7. Campbell M, McKenzie JE, Sowden A, Katikireddi SV, Brennan SE, Ellis S, et al. Synthesis without meta-analysis (SWiM) in systematic reviews: reporting guideline. BMJ. 2020;368:l6890. 8. Campbell M, Katikireddi SV, Sowden A, Thomson H. Lack of transparency in reporting narrative synthesis of quantitative data: a methodological assessment of systematic reviews. J Clin Epidemiol. 2019;105:1-9. 9. Melendez-Torres GJ, O'Mara-Eves A, Thomas J, Brunton G, Caird J, Petticrew M. Interpretive analysis of 85 systematic reviews suggests that narrative syntheses and meta-analyses are incommensurate in argumentation. Research Synthesis Methods. 2017;8(1):109-18. 10. McKenzie J, Brennan S, Page M, Chau M, Kramer S, Bosch M. From summary to synthesis: a review of statistical synthesis and presentation methods used in complex reviews [poster]. Better Knowledge for Better Health | Un meilleur savoir pour une meilleure santé Abstracts of the 21st Cochrane Colloquium; 19-23 September 2013; Québec City, Canada: John Wiley & Sons; 2013. 11. Arnup SJ, Forbes AB, Kahan BC, Morgan KE, McKenzie JE. Appropriate statistical methods were infrequently used in cluster-randomized crossover trials. J Clin Epidemiol. 2016;74:40-50. 12. Page MJ, Forbes A, Chau M, Green SE, McKenzie JE. Investigation of bias in meta-analyses due to selective inclusion of trial effect estimates: empirical study. BMJ Open. 2016;6(4):e011863. 13. Turner SL, Karahalios A, Forbes AB, Taljaard M, Grimshaw JM, Cheng AC, et al. Design characteristics and statistical methods used in interrupted time series studies evaluating public health interventions: a review. J Clin Epidemiol. 2020;122:1-11."
}
]
}
] | 1
|
https://f1000research.com/articles/9-678
|
https://f1000research.com/articles/9-1213/v1
|
08 Oct 20
|
{
"type": "Systematic Review",
"title": "Maternal lipid levels in pregnant women without complications in developing risk of large for gestational age newborns: a meta-analysis",
"authors": [
"Muhammad Pradhiki Mahindra",
"Mahendra Tri Arif Sampurna",
"Muhammad Pradhika Mapindra",
"Apriska Mega Sutowo Putri",
"Muhammad Pradhiki Mahindra",
"Muhammad Pradhika Mapindra",
"Apriska Mega Sutowo Putri"
],
"abstract": "Background: Circulating into foetal circulation across the placental barrier, abnormal maternal serum lipids predispose neonates to metabolic dysfunction and thereafter affect the steroid metabolism and functions of extra-embryonic foetal tissues.\nMethods: A systematic review was conducted by searching PubMed–MEDLINE and the Cochrane library between January 2010 and January 2020. The included studies were English case control studies that described original data on at least one raw lipid measurement during pregnancy in healthy women who delivered large for gestational age (LGA) newborns and in healthy women with non-LGA newborns. The data extracted from 12 studies were pooled, and the weighted mean difference (WMD) in lipid levels was calculated using random effects models. A meta-analysis was performed to identify sources of heterogeneity and to describe the significant value of the collected studies. Results: Of 643 publications identified, a total of 12 met the inclusion criteria. Compared with women who had non-LGA newborns, those who had LGA newborns had significantly higher triglyceride (TG) levels (WMD = 0.28, 95% CI −0.02 to 0.54) and lower high density lipoprotein cholestrol (HDL-C) levels (WMD = 0.08, 95% CI −0.13 to −0.03), but not have significantly lower high-density lipoprotein cholesterol (LDL-C) levels. Moreover, the levels of total cholesterol, low-density lipoprotein cholesterol, and very low density lipoprotein cholesterol (VLDL-C) were inconsistent between both groups. Conclusions: High levels of TG and low levels of HDL-C could cause births of LGA newborns whereas maternal serum of TC, LDL-C and VLDL-C cannot be used as predictor of LGA.",
"keywords": [
"Foetal birthweight",
"healthy women",
"LGA",
"maternal lipids",
"non-LGA"
],
"content": "Introduction\n\nThe early stages of pregnancy involve endocrine and metabolic changes and is an important period for placenta formation and foetal development. Epidemiological studies have shown that excessive lipid exposure in the maternal intrauterine environment can affect the development of foetal organs and lead to maternal metabolic impairment1. Abnormalities in maternal serum lipids have been highly correlated with birth weight and may be a cause of neonatal metabolic dysfunction2. The prevalence of foetal macrosomia in developed countries ranges from 5% to 20%, and several studies have reported that gestational diabetes mellitus (GDM) and maternal obesity were strongly associated with the risks for low and high birth weights. Disturbances in maternal metabolism affect blood glucose and other maternal macronutrients, such as lipids, and subsequently affect the development of the foetus2–4. The role of triglycerides (TG) is yet to be completely understood, but a cohort study in Amsterdam reported that maternal TG concentrations during the early stages of pregnancy were linearly related with the prevalence of large for gestational age (LGA) newborns5. Overweight foetuses may develop stillbirth and are at risk for neonatal mortality5–7. Therefore, LGA newborns have increased risk for developing type 2 diabetes, cardiovascular diseases and hypertension in their adult age5–9.\n\nHigh maternal serum lipid levels have been shown to increase the likelihood of pregnancy problems, such as GDM, pre-eclampsia and pre-term delivery2. Moreover, increase in the levels of TG, total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C) and very low-density lipoprotein cholesterol (VLDL-C) can account for adverse outcomes of normal gestation. Maternal serum lipids are transferred through the placenta, suggesting that they can affect foetal sterol metabolism and the metabolic functions of extra-embryonic foetal tissues6,7,9. These studies implied how essential lipid levels are to foetal development. The impact of high maternal lipid levels on foetal birth weight remains barely recognised in clinical practice, although this is known to be a cause of cardiovascular disease and diabetes2.\n\nPrevious studies have suggested that pregnant women with GDM and normal blood glucose levels had an increased risk for delivering LGA newborns10. The positive association between early maternal hypertriglyceridemia and LGA newborns in low-risk women is well recognised in some studies6. To further investigate the relationship between maternal lipid levels and LGA newborns, we conducted a systematic review of existing cohort studies to determine the potential role of maternal lipid levels as a risk factor for uncomplicated pregnancy and LGA newborn delivery.\n\n\nMethods\n\nThis study is reported in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA P) guidelines (Figure 1; Reporting guidelines11). Only research papers written in English between January 2010 and January 2020 were included in our search. We retrieved several papers with abstracts that mentioned an association between maternal TG and LGA newborns. For additional citations, references were collected from the included articles. Then, we determined the eligibility of the included studies using a critical appraisal skill programme (CASP) checklist for cohort model to determine the quality of the included studies. Based on the assessment using CASP tools, we finally included 12 articles to assess and review. The variables extracted from the literature are shown in Table 2. Full-text articles were acquired and evaluated for eligibility.\n\nWe conducted our search, which was conducted from April 2020 to June 2020, in the following databases: PubMed (MEDLINE), Library of Michigan University and the Cochrane library. We applied search strings, including combinations of search terms, as keywords placed in the titles or abstracts of the studies (Table 1). The keywords we used for the search strategy were as follows: 1. maternal lipid profile, lipid profile or lipoprotein and 2. LGA or large for gestational age.\n\nWe included studies that assessed the relationship between maternal serum TG levels during early to late pregnancy and LGA newborn delivery by healthy women or women who had no confounding factors, such as obesity, GDM, type 1 diabetes mellitus (T1DM), type 2 diabetes mellitus (T2DM), intake of medications that can alter lipid levels and maternal obesity. Lipid profiles, including TG, TC, HDL-C, LDL-C and VLDL-C, were measured from healthy pregnant women who had the outcome LGA newborn delivery. We did not include unpublished studies, letter to the editor, commentary, supplementary materials and conference paper. We excluded studies that did not show any relevance or similarity with our study purposes. This study evaluated maternal serum TG levels measured during pregnancy but not pre-conception TG levels.\n\nMaternal serum lipid levels in mmol/L and mg/dL were compared between LGA and non-LGA newborns. The study design, study population, method of data collection, gestational age at sampling and serum lipid levels were evaluated for all the selected studies. We measure lipid levels in mmol/l to make them homogenous so all lipid level measurements being reported in mg/dl were converted to mmol/L12. The reported mean and standard deviation (SD) values in the selected studies were recorded. However, we found several studies that reported their findings in the form of median and interquartile range; we used a digital calculator to change these into estimated mean and SD13.\n\nAll authors recorded and reviewed the collected articles. The main authors determined the study design, time frame and criteria for the included studies. The other authors helped retrieve the articles and process the data with statistical analyses. We decided to include studies that had prospective and retrospective cohort designs.\n\nFirst, we measured the maternal lipid levels with the standard unit of mmol/L in all included studies; data on maternal lipid levels in mg/dL were converted to mmol/L. Thereafter, we used the Review Manager version 5.3 (RevMan) software designed by Cochrane for statistical analyses. The mean difference and analysed statistical significances of the reported maternal serum lipid levels were calculated and evaluated in terms of impact on the outcome of LGA delivery. I2 statistics was performed to assess heterogeneity.\n\n\nResults\n\nOur search retrieved 649 articles, 147 of which were independently identified as duplications. Subsequently, we decided to select 77 of 147 articles that had titles and abstracts that were related with the measurement of lipoprotein levels in pregnant women and its impact on neonates. Of the 77 articles, 40 articles that had suitable research methods and outcomes were read in full text. We found that 28 of the 40 articles did not look at the target population, leaving 13 articles that were suitable for inclusion. We cross-checked the remaining articles to ensure that original studies were reported. Detailed information on author’s name, publication year, sample size, study design, the determined lipid level and the definitions of LGA was recorded and tabulated in Microsoft Excel 2010 software (Table 2 and Table 3).\n\n*The mean was calculated by inputting the median, lowest range and highest range to the estimating calculator based on SP Hozo13.\n\naThe data was converted from mg/dL to mmol/L using a standard measuring unit12.\n\nAs shown in Table 2, the baseline characteristics of the included studies were explained. There were 12 prospective cohort studies that reported a total of 17,731 cases, 4,430 of which included LGA newborns. Maternal lipid profiles were measured in the first trimester in four articles3,9,14,15; in the second trimester in one article17; in the last trimester in three articles16,19,20 and in any of the gestational weeks in the remaining articles4,6,7,18.\n\nMost of the studies4,6,7,15 used the INTERGROWTH-21st definition of >90th percentile for LGA newborns. Some studies in China3,19 used a referred percentile standard based on a Chinese population, and some14,16,18 used their country’s definition of LGA. A detailed list of the implemented eligibility criteria for each study is shown in Table 2, and the mean values of the determined biochemical lipids are presented in Table 3. There were 12 included studies that investigated the effects of lipid profile in pregnant women who had no complications on LGA newborn delivery. The exposures of maternal lipid profile included TG (N = 12), TC (N = 9), HDL-C (N = 7), LDL-C (N = 7) and VLDL-C (N = 2). Based on the analysis of the RevMan tool, we found that the investigated studies that analysed maternal TG, TC, HDL-C and LDL-C had an I2 of more than 50%; for this reason, we used a random effects model. On the other hand, we used a fixed effect model to assess the studies that investigated maternal VLDL-C, because the I2 was below 50%.\n\nThere were 11 studies that assessed TG levels during pregnancy; 4,761 case subjects and 14,174 control subjects were included. Maternal serum TG levels were found to be significantly associated with LGA foetuses in the first trimester, according to three of the included studies3,14,15. One study reported that maternal serum TG levels in the second trimester were significantly related with the risk for LGA newborns before and after data adjustment9. On the other hand, another study on a similar population17 showed a non-significant correlation between maternal serum TG levels and the risk for LGA newborns. Furthermore, two studies7,19 reported an association between maternal serum TG levels and LGA occurrence. Each mmol/L increase in maternal serum TG was found to increase the risk for LGA newborns and reduce the probability of small gestational age newborns19. The remaining studies measured TG levels in the first, second and third trimesters and reported significant associations between maternal serum TG levels and the risk for LGA newborns4,7,18.\n\nFigure 2 shows the comparison of the mean differences of the included studies. Random effects model meta-analysis showed that the pooled weighted mean difference was 0.28 mmol/L (95% CI −0.02 to 0.54), and significant heterogeneity was observed (Tau² = 0.19; Chi²= 2460.32, I² = 100%, p = 0.03).\n\nData on 2,225 patients and 13,218 controls from nine studies3,6,9,14–17,19,20 were included to evaluate the relationship between TC and LGA newborns. In contrast to all the studies that reported an insignificant association between TC levels and LGA, Wang reported that abnormal levels of maternal TC were significantly associated with the event of LGA delivery in the first trimester3. In fact, some reports were insufficient to prove a significant correlation between TC level and the risk for LGA newborns, whereas other reports found non-significant associations between TC levels and the risk for LGA newborns in the first, second and third trimesters6,19.\n\nBased on the random effects model meta-analysis (Figure 3 and Table 4), the included studies had a pooled weighted mean difference of −0.06 mmol/L (95% CI −0.16 to 0.05) and heterogeneity (Tau2 = 0.02, Chi² = 65.27, I² = 88%) with p value = 0.26.\n\nThe analysis of HDL-C and risk for LGA newborns included 1,881 patients and 8,603 controls from seven studies3,6,14,16,17,19,20. HDL-C levels were reported to be significantly associated with the risk for LGA and SGA foetuses in the third trimester of pregnancy16,20. In addition, one study found a significant association in the second trimester6. On the other hand, one study showed HDL-C as the only lipid that was not significantly related with the birth of LGA foetuses3.\n\nFigure 4 and Table 4 show the results of the meta-analysis of the included studies. The pooled weighted mean difference was 0.08 (95% CI −0.13 to −0.03), and heterogeneity was found (Tau2 = 0.00, Chi² = 46.53, I² = 87%), with p = 0.003.\n\nSix of the included studies3,6,14,16,17,19,20, which recruited 1,881 patients and 8,603 controls, majority reported no significant correlations between LDL-C concentration and LGA newborns as a neonatal outcome. On the other hand, four studies3,6,14,17 showed that LDL-C concentration was associated with LGA newborns. This association was found during the second and third trimesters of pregnancy6. Furthermore, the study by Wang supported this association by showing that LDL-C concentrations played a significant role in the risk for LGA newborns and that three lipids (TG, TC and LDL-C) were significant contributing factors3.\n\nFigure 5 and Table 4 show that the pooled weight mean difference was −0.03 (95% CI −0.11 to 0.06) and that heterogeneity was found (Tau2 = 0.01, Chi² = 50.91, I² = 88%), with p = 0.56.\n\nStudies and available information on the impact of VLDL on LGA remain unclear. Nevertheless, two studies that included a total of 36 patients and 425 controls reported that there was no correlation between VLDL and LGA newborns16,17. Based on our meta-analysis, the level of maternal serum VLDL was not significantly associated with births of LGA newborns (p = 0.60) (Figure 6).\n\n\nDiscussion\n\nData from 12 published articles were evaluated in this systematic review to determine the relationship between lipid values measured during pregnancy and the risk for LGA newborns. Our review presented some valuable findings. We discovered that TC levels were inconsistent in both groups of women who delivered LGA and non-LGA newborns. This finding suggested that TC level as a determinant of LGA newborn delivery is clinically not useful. In support of this result, almost all studies reported that TC levels were similar across the groups. In addition, Parlakgumus17 reported that TC levels in the second trimester took a decisive role on the risk for LGA foetuses and newborns, compared with the results of many studies.\n\nMany of the studies reported increase in TG levels in women who delivered LGA newborns. Our meta-analysis concluded that maternal TG levels were significantly elevated in women who would deliver LGA neonates. Moreover, maternal HDL-C levels were lower in women with LGA newborns than in those with non-LGA newborns. Therefore, a low level of maternal HDL-C concentration was significantly associated with the risk for LGA newborns. Levels of maternal LDL-C had no significant weight on women who had LGA newborns. Therefore, LDL-C and VLDL-C levels were not significant causative factors of LGA outcomes in pregnant women who had no comorbidities.\n\nExclusion of all confounding factors, such as T1DM, T2DM, GDM, maternal obesity and excessive gestational weight gain (GWG), which can affect the increased risk for LGA newborns in women who had abnormal lipid profiles, was important. A large prospective study on more than 700 women showed significant correlations of T1DM and HbA1c ≥42 mmol/L (6%) with increased rates of LGA at 26 and 34 weeks age of gestation21. Similarly, a retrospective cohort study by Lisa et al showed considerably higher rates of LGA newborns in women with T1DM (39%) than in women with T2DM (17%); their multivariate analysis on non-Caucasian women demonstrated an increased risk for LGA newborns in women who had T1DM (OR = 4.07; 95% CI 1.46 to 11.35) and T2DM (OR = 2.47; 95% CI 1.15 to 5.32)22.\n\nA reported study on 175 women with T1DM in the United States discovered similar rates of LGA newborn delivery in women who had HbA1C of >6.5% and <6.5%, suggesting the likelihood of T1DM as a contributing factor23. A cohort study on multi-ethnic groups revealed that GDM and relatively high pregnancy BMI were linked with an increased risk for LGA newborn delivery. The prevalence of LGA newborns among women with GDM was highest in African, American and Hispanic women and lowest in Asian, Filipino and White women24.\n\nIn another study on GDM, women who had elevated fasting plasma glucose levels were at a relatively high risk for having LGA newborns25. An analysis of 23,000 women in the Hyperglycaemia and Adverse Pregnancy Outcomes study discovered that the macrosomia prevalence in non-obese women was 6.7% in 1,244 patients without GDM and 10.2% in 2,791 patients with GDM. The investigators found that the frequency of macrosomia was 50% higher in women with GDM than in women without GDM in both the non-obese and obese groups26. Moreover, abnormal pre-pregnancy BMI significantly increased the risk for LGA neonates.\n\nOur review could not exclude women with excessive GWG, because this was not reported in the majority of the included studies. Therefore, we assumed that excessive GWG may have affected our results. Some studies showed that excessive GWG in pregnant women who had no complications increased the risk for delivering LGA newborn; compared with women who had uncomplicated pregnancies, those who exceeded the GWG recommendation had three and six times higher risk for macrosomia births27. The expected association of pre-pregnancy BMI and GWG with maternal and foetal outcomes showed that GWG of >16 kg led to an increased risk for delivering LGA neonates28.\n\nA study by Lu et al reported that high second trimester GWG was significantly related with a relatively high risk for LGA newborns29. The probability of giving birth to an LGA newborn increased by 6.9% per kilogram of maternal weight gain, and the odds ratio was 1.249 for GWG beyond the recommended amount30. Similarly, another study found that the odds ratio for delivering LGA newborns was higher for non-diabetic Caucasian women with BMIs <25 or >25 than in women with GDM and normal BMIs31.\n\nThis review was the first to directly address the association between maternal lipid profiles and the risk for LGA newborns, without any confounding factors. However, it had several weaknesses. First, this review depended on the design and quality of the included studies, regardless of the baseline lipid level, which was crucial to the results of our meta-analysis. Second, our meta-analysis did not distinguish pregnant women based on trimester of pregnancy but described the effects of physiologic changes in lipid metabolism on pregnant women and the risk for LGA newborns throughout the entire pregnancy; it did not consider the confounding factors that may occur in different trimesters. Third, we did not include unpublished studies, which could beneficially support our aim to illustrate better review results. We assumed that our review results might not be sufficient to meet our expectations. Therefore, all of these reasons became researcher biases, which may have resulted in our findings.\n\nMost of the existing observational studies cannot be used to predict the definitive value of the independent contribution of lipid levels to maternal and neonatal outcomes because of the unmeasured confounding factors and methodological limitations. We recommend that future studies analyse women separately based on their non-modifiable characteristics, such as maternal age, race and inherited disorders. Furthermore, we noticed that these observational studies did not exclude women who had excessive GWG, which can contribute to the risk for LGA newborns. Future observational studies must include details on maternal lifestyles and environment to minimise population bias.\n\n\nConclusions\n\nIn conclusion, this review demonstrated notable findings from studies on the associations between maternal lipid levels and risk for LGA newborns. Our meta-analysis emphasised that high levels of TG and low levels of HDL-C may affect foetal development and cause births of LGA newborns. On the other hand, maternal serum of TC, LDL-C and VLDL-C cannot be used as predictor of LGA without the other risk factors, such as excessive GWG and insulin resistance. However, we need a better understanding of the relative contributions of other confounding factors, such as gestational age at sampling, maternal age and excessive GWG. We acknowledge that we used exclusion criteria, such as T1DM, T2DM, obesity and hypertension. Excessive GWG was not an exclusion criterion because of the limited amount of studies that excluded such population of women, although we were aware that it may contribute to LGA newborn delivery in healthy women.\n\n\nData availability\n\nFigshare: Underlying Data - Maternal Lipid Levels on Pregnant Women without Complication in Developing Risk of Large for Gestational Age Newborn: Meta-Analysis study, https://doi.org/10.6084/m9.figshare.13011941.v232.\n\nFigshare: PRISMA checklist for ‘Maternal Lipid Levels on Pregnant Women without Complication in Developing Risk of Large for Gestational Age Newborn: Meta-Analysis study’, https://doi.org/10.6084/m9.figshare.13011803.v211.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "References\n\nHeerwagen MJR, Miller MR, Barbour LA, et al.: Maternal obesity and fetal metabolic programming: a fertile epigenetic soil. Am J Physiol Regul Integr Comp Physiol. 2010; 299(3): R711–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang J, Moore D, Subramanian A, et al.: Gestational dyslipidaemia and adverse birthweight outcomes: a systematic review and meta-analysis. Obes Rev. 2018; 19(9): 1256–68. PubMed Abstract | Publisher Full Text\n\nWang C, Zhu W, Wei Y, et al.: The associations between early pregnancy lipid profiles and pregnancy outcomes. J Perinatol. 2017; 37(2): 127–33. PubMed Abstract | Publisher Full Text\n\nLiang N, Zhu H, Cai X, et al.: The high maternal TG level at early trimester was associated with the increased risk of LGA newborn in non-obesity pregnant women. Lipids Health Dis. 2018; 17(1): 1–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang X, Guan Q, Zhao J, et al.: Association of maternal serum lipids at late gestation with the risk of neonatal macrosomia in women without diabetes mellitus. Lipids Health Dis. 2018; 17(1): 1–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFarias DR, Poston L, Franco-Sena AB, et al.: Maternal lipids and leptin concentrations are associated with large-for-gestational-age births: A prospective cohort study. Sci Rep. 2017; 7(1): 804. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGeraghty AA, Alberdi G, O’Sullivan EJ, et al.: Maternal and fetal blood lipid concentrations during pregnancy differ by maternal body mass index: Findings from the ROLO study. BMC Pregnancy Childbirth. 2017; 17(1): 360. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJin WY, Lin SL, Hou RL, et al.: Associations between maternal lipid profile and pregnancy complications and perinatal outcomes: A population-based study from China. BMC Pregnancy Childbirth. 2016; 16: 60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVrijkotte TGM, Krukziener N, Hutten BA, et al.: Maternal lipid profile during early pregnancy and pregnancy complications and outcomes: The ABCD study. J Clin Endocrinol Metab. 2012; 97(11): 3917–25. PubMed Abstract | Publisher Full Text\n\nSchaefer-Graf UM, Graf K, Kulbacka I, et al.: Maternal lipids as strong determinants of fetal environment and growth in pregnancies with gestational diabetes mellitus. Diabetes Care. 2008; 31(9): 1858–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMuhammad Pradhiki M, Mahendra Tri Arif S, Muhammad Pradhika M, et al.: PRISMA checklist - Maternal Lipid Levels on Pregnant Women Without Complication in Developing Risk of Large for Gestational Age Newborn: Meta-Analysis study. figshare. Preprint. 2020. http://www.doi.org/10.6084/m9.figshare.13011803.v2\n\nRugge B, Balshem H, Sehgal R: Screening and Subclinical Hypothyroidism or Hyperthyroidism. Agency for Healthcare Research and Quality U.S. Department of Health and Human Services. 2007.\n\nHozo SP, Djulbegovic B, Hozo I: Estimating the mean and variance from the median , range , and the size of a sample. BMC Med Res Methodol. 2005; 5: 13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee SM, Kim BJ, Koo JN, et al.: Nonalcoholic fatty liver disease is a risk factor for large-for-gestational-age birthweight. PLoS One. 2019; 14(8): e0221400. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPazhohan A, Rezaee Moradali M, Pazhohan N: Association of first-trimester maternal lipid profiles and triglyceride-glucose index with the risk of gestational diabetes mellitus and large for gestational age newborn. J Matern Neonatal Med. 2019; 32(7): 1167–75. PubMed Abstract | Publisher Full Text\n\nMitra S, Nayak P, Misra S, et al.: Effect of maternal anthropometry and metabolic parameters on fetal growth. Indian J Endocrinol Metab. 2012; 16(5): 754–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nParlakgumus HA, Aytac PC, Kalayci H, et al.: First trimester maternal lipid levels and serum markers of small- and large-for-gestational age infants. J Matern Neonatal Med. 2014; 27(1): 48–51. PubMed Abstract | Publisher Full Text\n\nHarville EW, Juonala M, Viikari JSA, et al.: Preconception metabolic indicators predict gestational diabetes and offspring birthweight. Gynecol Endocrinol. 2014; 30(11): 840–4. PubMed Abstract | Publisher Full Text\n\nHou RL, Zhou HH, Chen XY, et al.: Effect of maternal lipid profile, C-peptide, insulin, and HBA1c levels during late pregnancy on large-for-gestational age newborns. World J Pediatr. 2014; 10(2): 175–81. PubMed Abstract | Publisher Full Text\n\nYe K, Bo QL, Du QJ, et al.: Maternal serum lipid levels during late pregnancy and neonatal body size. Asia Pac J Clin Nutr. 2015; 24(1): 138–43. PubMed Abstract | Publisher Full Text\n\nMorrens A, Verhaeghe J, Vanhole C, et al.: Risk factors for large-for-gestational age infants in pregnant women with type 1 diabetes. BMC Pregnancy Childbirth. 2016; 16(1): 162. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlexander LD, Tomlinson G, Feig DS: Predictors of Large-for-Gestational-Age Birthweight Among Pregnant Women With Type 1 and Type 2 Diabetes: A Retrospective Cohort Study. Can J Diabetes. 2019; 43(8): 560–6. PubMed Abstract | Publisher Full Text\n\nScifres CM, Feghali MN, Althouse AD, et al.: Effect of excess gestational weight gain on pregnancy outcomes in women with type 1 diabetes. Obstet Gynecol. 2014; 123(6): 1295–302. PubMed Abstract | Publisher Full Text\n\nSridhar SB, Ferrara A, Ehrlich SF, et al.: Risk of large-for-gestational-age newborns in women with gestational diabetes by race and ethnicity and body mass index categories. Obstet Gynecol. 2013; 121(6): 1255–62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRyan EA, Savu A, Yeung RO, et al.: Elevated fasting vs post-load glucose levels and pregnancy outcomes in gestational diabetes: a population-based study. Diabet Med. 2020; 37(1): 114–22. PubMed Abstract | Publisher Full Text\n\nSzmuilowicz ED, Josefson JL, Metzger BE: Gestational Diabetes Mellitus. Endocrinol Metab Clin North Am. 2019; 48(3): 479–93. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFerraro ZM, Barrowman N, Prud’homme D, et al.: Excessive gestational weight gain predicts large for gestational age neonates independent of maternal body mass index. J Matern Neonatal Med. 2012; 25(5): 538–42. PubMed Abstract | Publisher Full Text\n\nNohr EA, Vaeth M, Baker JL, et al.: Combined associations of prepregnancy body mass index and gestational weight gain with the outcome of pregnancy. Am J Clin Nutr. 2008; 88(6): 1705. PubMed Abstract | Publisher Full Text\n\nLu W, Zhang X, Wu J, et al.: Association between trimester-specific gestational weight gain and childhood obesity at 5 years of age: Results from Shanghai obesity cohort. BMC Pediatr. 2019; 19(1): 139. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWeschenfelder F, Lehmann T, Schleussner E, et al.: Gestational Weight Gain Particularly Affects the Risk of Large for Gestational Age Infants in Non-obese Mothers. Geburtshilfe Frauenheilkd. 2019; 79(11): 1183–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim SY, Sharma AJ, Sappenfield W, et al.: Association of maternal body mass index, excessive weight gain, and gestational diabetes mellitus with large-for-gestational-age births. Obstet Gynecol. 2014; 123(4): 737–44. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMuhammad Pradhiki M, Mahendra Tri Arif S, Muhammad Pradhika M, et al.: Underlying Data - Maternal Lipid Levels on Pregnant Women Without Complication in Developing Risk of Large for Gestational Age Newborn: Meta-Analysis study. figshare. Dataset. 2020. http://www.doi.org/10.6084/m9.figshare.13011941.v2"
}
|
[
{
"id": "72787",
"date": "03 Nov 2020",
"name": "Kian Djien Liem",
"expertise": [
"Reviewer Expertise neonatal medicine"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGeneral comments This is a well-written manuscript. The authors have tried to explore the relationships between maternal lipid levels and the risk of having a large for gestational age (LGA) offspring using a systematic review. The results are of great importance for the obstetricians to be aware of the risk factors for developing LGA newborns. With this knowledge of the risk factors, they can take measures in the pregnant woman to prevent LGA newborns. LGA should be avoided as much as possible, since it is a risk factor for developing metabolic syndrome in adult life. The authors have demonstrated that there is an association between maternal lipid levels and risk for LGA newborns, especially high levels TG and low levers HDL-C. But they don’t make any recommendation what we should do for the pregnant women in order to prevent LGA of their offspring’s. Should we determine the lipid level as a routine during pregnancy? But determination of lipid levels during pregnancy could be too late, especially when this is abnormal. Therefore should we give more attention in pre-conception care and recommend normalization of lipid levels, before the women become pregnant? I would like to challenge the authors to make some discussion on the topic about what would be the consequences of their finding.\n\nSpecific comments\nIntroduction: No comments\nMethods: No comments\nResults: Page 4, the data mentioned under Study characteristics and assessment of risk bias don’t match with the data in figure 1. In the text it is mentioned: “Our search retrieved 649 articles, 147 of which were independently identified as duplications”. But in Figure 1 it is mentioned that the number of articles identified from database search is 643 and 183 is identified as duplicate. Which numbers are the right one? Thereafter, it is mentioned in the text: “Subsequently, we decided to select 77 of 147 articles that had titles and abstracts that were related with the measurement of lipoprotein levels in pregnant women and its impact on neonates”. But this is confusing. Do the authors mean that they select 77 articles from the 147 duplication articles? It doesn’t match with figure 1. A few lines further it is mentioned: “We found that 28 of the 40 articles did not look at the target population, leaving 13 articles that were suitable for inclusion”. But in figure 1 it is mentioned that 12 articles are included. Which number is the right one? In the next paragraph it is mentioned: “There were 12 prospective cohort studies that reported a total of 17,731 cases, 4,430 of which included LGA newborns”. It seems that 12 is the right number, like the number in figure 1. Page 6 and 7: It is mentioned at the end of page 6: “Maternal serum TG levels were found to be significantly associated with LGA foetuses in the first trimester, according to three of the included studies.” . What do the authors mean with LGA foetuses in the first trimester? LGA is defined as birth weight above the 90th percentile for their gestational age and gender. So, it is impossible to speak about LGA fetuses in te first semester. Do they mean the foetuses which in utero grows much larger than average for the gestational age or fetuses with estimated fetal weight > 90th percentile? But this is different than LGA. Page 7: It is mentioned: “Each mmol/L increase in maternal serum TG was found to increase the risk for LGA newborns and reduce the probability of small gestational age newborns”. Probably is this sentence not complete. How much is the percentage increase of the risk for LGA newborns and percentage decrease of the probability of SGA newborns for each mmol/L increase in maternal serum TG. Page 7, under Total cholesterol. It is mentioned: “Wang reported that abnormal levels of maternal TC were significantly associated with the event of LGA delivery in the first trimester”. What do the authors mean with the event of LGA delivery in the first trimester. Does it mean miscarriage of a foetus, with a body weight > 90th percentile for the gestational age? Page 7 under High density lipoprotein – cholesterol. The authors mention: “HDL-C levels were reported to be significantly associated with the risk for LGA and SGA foetuses in the third trimester of pregnancy”. According to the definition, LGA and SGA can only be diagnosed after birth after weighing the newborn's body weight. What do the authors mean with LGA and SGA fetuses in the third trimester? Do they mean the foetuses which estimated fetal weight larger or less than average for the gestational age?\nDiscussion Page 8: the authors mention: “A large prospective study on more than 700 women showed significant correlations of T1DM and HbA1c ≥42 mmol/L (6%) with increased rates of LGA at 26 and 34 weeks age of gestation”. What do they mean with increased rates of LGA at 26 and 34 weeks age of gestation. Does it mean increased rates of fetuses with estimated fetal weight > 90th percentile at 26 and 34 weeks age of gestation? Page 8: the authors discuss the relationship between T1DM, T2DM and GDM with LGA newborns. This is a well-known relationship. Therefore, it is the question whether this has any added value to include in the discussion. Perhaps it might be better when the authors will discuss the influence of abnormal maternal lipid level on the body weight of the foetuses of pregnant women with T1DM, T2DM and GDM. Do pregnant women with T1DM, T2DM and GDM and abnormal maternal lipid level have a higher risk for LGA newborns than in pregnant women with T1DM, T2DM and GDM with normal lipid level?\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Yes\n\nAre the conclusions drawn adequately supported by the results presented in the review? Partly",
"responses": [
{
"c_id": "6134",
"date": "29 Jan 2021",
"name": "Mahendra Tri Arif Sampurna",
"role": "Author Response",
"response": "Thank you for your correcting this mistake. We would like to answer the questions. As one of the most simple blood measurement, lipid levels especially LGA and HDL-C could be used as a routine blood test during the pregnancy for fetal programming. Normalization of lipid levels should be one of the main targets during the pregnancy. Physical activity and dietary adjustment such as habitual fish consumption would be an effective approach to reduce maternal TG levels and increase HDL-C levels (1,2). Reference : (1) Butler, C. L., Williams, M. A., Sorensen, T. K., Frederick, I. O., & Leisenring, W. M. (2004). Relation between maternal recreational physical activity and plasma lipids in early pregnancy. American journal of epidemiology, 160(4), 350-359. (2) Williams, M. A., Frederick, I. O., Qiu, C., Meryman, L. J., King, I. B., Walsh, S. W., & Sorensen, T. K. (2006). Maternal erythrocyte omega-3 and omega-6 fatty acids, and plasma lipid concentrations, are associated with habitual dietary fish consumption in early pregnancy. Clinical biochemistry, 39(11), 1063-1070."
},
{
"c_id": "6135",
"date": "29 Jan 2021",
"name": "Mahendra Tri Arif Sampurna",
"role": "Author Response",
"response": "Thank you for correcting these mistakes in Figure 1 1. Results section, Page 4, the data mentioned under Study characteristics and assessment of risk bias don’t match with the data in figure 1. In the text it is mentioned: “Our search retrieved 649 articles, 147 of which were independently identified as duplications”. But in Figure 1 it is mentioned that the number of articles identified from database search is 643 and 183 is identified as duplicate. Which numbers are the right one? Thereafter, it is mentioned in the text: “Subsequently, we decided to select 77 of 147 articles that had titles and abstracts that were related with the measurement of lipoprotein levels in pregnant women and its impact on neonates”. But this is confusing. Do the authors mean that they select 77 articles from the 147 duplication articles? It doesn’t match with figure 1. A few lines further it is mentioned: “We found that 28 of the 40 articles did not look at the target population, leaving 13 articles that were suitable for inclusion”. But in figure 1 it is mentioned that 12 articles are included. Which number is the right one? In the next paragraph it is mentioned: “There were 12 prospective cohort studies that reported a total of 17,731 cases, 4,430 of which included LGA newborns”. It seems that 12 is the right number, like the number in figure 1. Answers : We would like to apologize because of several mistakes about the article screening process in Figure 1. we would also intend to edit the Figure 1. Regarding the statement, we would revise our statement as follows : Our search retrieved 649 articles, 147 of which were independently identified as duplications and thus leaving 502 articles. Subsequently, we decided to select 77 of 502 articles that had titles and abstracts that were related to the measurement of lipoprotein levels in pregnant women and its impact on neonates. Of the 77 articles, 40 articles that had suitable research methods and outcomes were read in full text. We found that 28 of the 40 articles did not indicate mean ± SD or median with upper and lower quartiles for maternal lipid level measurements. Thus, leaving 12 articles that were suitable for inclusion. 2. Results section, Page 6 and 7: It is mentioned at the end of page 6: “Maternal serum TG levels were found to be significantly associated with LGA foetuses in the first trimester, according to three of the included studies.” . What do the authors mean with LGA foetuses in the first trimester? LGA is defined as birth weight above the 90th percentile for their gestational age and gender. So, it is impossible to speak about LGA fetuses in te first semester. Do they mean the foetuses which in utero grows much larger than average for the gestational age or fetuses with estimated fetal weight > 90th percentile? But this is different than LGA. Answers : Thank you for your correction. We changed the confusing statement to : “ Maternal serum TG levels in the first trimester were found to be significantly associated with LGA infants, according to three of the included studies” 3. Results section, Page 7: It is mentioned: “Each mmol/L increase in maternal serum TG was found to increase the risk for LGA newborns and reduce the probability of small gestational age newborns”. Probably is this sentence not complete. How much is the percentage increase of the risk for LGA newborns and percentage decrease of the probability of SGA newborns for each mmol/L increase in maternal serum TG. Answers : Thank you very much for your corrections. Actually, this sentence is a mistake and not informative statement. We decided to delete it because it did not imply our study results 4. Results section, Page 7 under High density lipoprotein – cholesterol. The authors mention: “HDL-C levels were reported to be significantly associated with the risk for LGA and SGA foetuses in the third trimester of pregnancy”. According to the definition, LGA and SGA can only be diagnosed after birth after weighing the newborn's body weight. What do the authors mean with LGA and SGA fetuses in the third trimester? Do they mean the foetuses which estimated fetal weight larger or less than average for the gestational age? Answers : Thank you for your comment and corrections We would like to change the confusing statement to : Wang reported that abnormal levels of maternal TC in the first trimester were significantly associated with the event of LGA infants 5. Results section, Page 7 under High density lipoprotein – cholesterol. We mention: “HDL-C levels were reported to be significantly associated with the risk for LGA and SGA foetuses in the third trimester of pregnancy”. Answers : Thank you for your corrections. We would like to vhange the statement to : HDL-C levels in the third trimester of pregnancy were significantly associated with both LGA and SGA infants 6. Discussion Section, Page 8: the authors mention: “A large prospective study on more than 700 women showed significant correlations of T1DM and HbA1c ≥42 mmol/L (6%) with increased rates of LGA at 26 and 34 weeks age of gestation”. What do they mean with increased rates of LGA at 26 and 34 weeks age of gestation. Does it mean increased rates of fetuses with estimated fetal weight > 90th percentile at 26 and 34 weeks age of gestation? Answers : Thank you for the corrections. We would like to apologize for confusing statement. We would like to changed it to : A large prospective study on more than 700 women showed significant correlations of T1DM and HbA1c ≥ 42 mmol/L (6%) during 26 and 34 weeks age of gestation with increased risks of LGA newborns. 7. Discussion section, Page 8: the authors discuss the relationship between T1DM, T2DM and GDM with LGA newborns. This is a well-known relationship. Therefore, it is the question whether this has any added value to include in the discussion. Perhaps it might be better when the authors will discuss the influence of abnormal maternal lipid level on the body weight of the foetuses of pregnant women with T1DM, T2DM and GDM. Do pregnant women with T1DM, T2DM and GDM and abnormal maternal lipid level have a higher risk for LGA newborns than in pregnant women with T1DM, T2DM and GDM with normal lipid level? Answers : Thank you ver much for your suggestion. We would like to add useful information to support our study : A previous longitudinal study reported that compared the groups of women with T1DM and T2DM and healthy women found a positive association between maternal serum TG and LGA infants regardless of glycemic levels condition (1). Fasting maternal hypertriglyceridemia could be used as a significant predictor of LGA infants that is independent of maternal BMI, weight gain, and blood glucose levels (2). Reference : (1) Göbl, C. S., Handisurya, A., Klein, K., Bozkurt, L., Luger, A., Bancher-Todesca, D., & Kautzky-Willer, A. (2010). Changes in serum lipid levels during pregnancy in type 1 and type 2 diabetic subjects. Diabetes care, 33(9), 2071-2073. (2) Kitajima, M., Oka, S., Yasuhi, I., Fukuda, M., Rii, Y., & Ishimaru, T. (2001). Maternal serum triglyceride at 24–32 weeks’ gestation and newborn weight in nondiabetic women with positive diabetic screens. Obstetrics & Gynecology, 97(5), 776-780."
}
]
},
{
"id": "72781",
"date": "21 Dec 2020",
"name": "Victor Samuel Rajadurai",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors have done a systematic review of the influence of lipid levels in pregnant women and their effect on the developing fetus particularly, large for gestational age (LGA). The review has clearly addressed the rationale for the review, objectives, characteristics of the participants and outcomes. Sufficient details of the search criteria, methods, inclusion & exclusion criteria and details of the data collection have been clearly defined. Standard statistical methods, applications and interpretations have been used. The conclusions are valid and supported by the results.\nSuggested the following changes:\nLow levels of HDL Cholesterol may reflect the life style, lack of exercise and relative inactivity of the women included in the articles. Sufficient information regarding this has not been mentioned in the papers included. (To discuss in detail in a separate paragraph).\n\nUnder “strengths and Limitations”: I would suggest to delete, the third mentioned limitation as this has already been discussed under inclusion criteria.\n\nIt would be good to discuss whether increased TG levels in the absence of overt GDM or established DM (Type 1 & 2) could be a marker of pre-diabetics in some women who give birth to LGA babies.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Yes\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes",
"responses": [
{
"c_id": "6232",
"date": "29 Jan 2021",
"name": "Mahendra Tri Arif Sampurna",
"role": "Author Response",
"response": "1. Thank you for your suggestion. We would like to add this information on discussion section. “Lower TG and Higher HDL-C levels are linked to the physical inactivity, a tendency to less responsive to regular exercise. A program of exercise training is reported effective to alter the concentrations of lipoprotein, which therefore prompt the lipoprotein levels to be in the expected range (1,2). 2. Thank you very much for your corrections. We would like to delete the statement as you suggested. 3. Thank you very much for your suggestion, we have added the informative statement as you suggested in the discussion section. “Maternal lipid profiles are reported as the potential predictor for GDM in pregnant women. High concentrations of TG, TC, and LDL were found in women diagnosed with GDM throughout the second trimester (3). Another prospective cohort had also demonstrated that women who were exhibiting GDM in the second trimester, had shown higher levels of TG, TC, and LDL, and lower levels of HDL during the first trimester, even with normal glycemia and glycated hemoglobin (4). These findings emphasized that the role of lipid metabolism is crucial to contribute to the pathogenesis of such metabolic disorders.” References: (1) (1) C, Després JP, Lamarche B, Bergeron J, Gagnon J, Leon AS, Rao DC, Skinner JS, Wilmore JH, Bouchard C. Effects of endurance exercise training on plasma HDL cholesterol levels depend on levels of triglycerides: evidence from men of the Health, Risk Factors, Exercise Training and Genetics PLEASE NOTE If you are an AUTHOR of this article, please check that you signed in with the account associated with this article otherwise we cannot automatically identify your role as an author and your comment will be labelled as a “User Comment”. If you are a REVIEWER of this article, please check that you have signed in with the account associated with this article and then go to your account to submit your report, please do not post your review here. If you do not have access to your original account, please contact us. User comments must be in English, comprehensible and relevant to the article under discussion. We reserve the right to remove any comments that we consider to be inappropriate, offensive or otherwise in breach of the User Comment Terms and Conditions. Commenters must not use a comment for personal attacks. When criticisms of the article are based on unpublished data, the data should be made available. (HERITAGE) Family Study. Arteriosclerosis, thrombosis, and vascular biology. 2001 Jul;21(7):1226-32. (2) Larrydurstine J, Haskell WL. Effects of exercise training on plasma lipids and lipoproteins. Exercise and sport sciences reviews. 1994 Jan 1;22(1):477-522. References : (3) Correa PJ, Venegas P, Palmeiro Y, Albers D, Rice G, Roa J, Cortez J, Monckeberg M, Schepeler M, Osorio E, Illanes SE. First trimester prediction of gestational diabetes mellitus using plasma biomarkers: a case-control study. Journal of perinatal medicine. 2019 Feb 25;47(2):161-8. (4) Li G, Kong L, Zhang L, Fan L, Su Y, Rose JC, Zhang W. Early pregnancy maternal lipid profiles and the risk of gestational diabetes mellitus stratified for body mass index. Reproductive Sciences. 2015 Jun;22(6):712-7."
},
{
"c_id": "6233",
"date": "29 Jan 2021",
"name": "Mahendra Tri Arif Sampurna",
"role": "Author Response",
"response": "Thank you for your correcting our article.We would like to answer the questions. Results Section 1. We would like to apologize because of several mistakes about the article screening process in Figure 1. We have edited these mistakes and replaced the Figure 1, attached on the revised file \"Our search retrieved 649 articles, 147 of which wereindependently identified as duplications and thus leaving 502 articles. Subsequently, we decided to select 77 of 502 articles that had titles and abstracts that were related to the measurement of lipoprotein levels in pregnant women and its impact on neonates. Of the 77 articles, 40 articles that had suitable research methods and outcomes were read in full text\". . 2. Thank you for your correction. We changed it into: “ We found that 28 of the 40 articles did not indicate mean ± SD or median with upper and lower quartiles for maternal lipid level measurements. Thus, leaving 12 articles that were suitable for inclusion. 3. Thank you for your correction. We changed it into “ Maternal serum TG levels in the first trimester were found to be significantly associated with LGA infants, according to three of the included\" 4. Thank you for your comment. Actually, this sentence is a mistake and not informative. We decided to delete it because it did not imply our study results 5. Thank you for your comment. We changed it into : \"Wang reported that abnormal levels of maternal TC in the first trimester were significantly associated with the event of LGA infants studies” 6. Thank you for your comments. Due to mistyping, we changed it according to your suggestion \" HDL-C levels in the third trimester of pregnancy were significantly associated with both LGA and SGA infants\" Discussion Section 7. Thank you for your comments. We would like to change the confusing sentences : \"A large prospective study on more than 700 women showed significant correlations of T1DM and HbA1c ≥ 42 mmol/L (6%) during 26 and 34 weeks age of gestation with increased risks of LGA newborns\". 8. As one of the most simple blood measurement, lipid levels especially LGA and HDL-C could be used as a routine blood test during the pregnancy for fetal programming. Normalization of lipid levels should be one of the main targets during the pregnancy. Physical activity and dietary adjustment such as habitual fish consumption would be an effective approach to reduce maternal TG levels and increase HDL-C levels (1,2). 9. Thank you for your suggestion. We added useful information to support our study and put it in disucssion section. \"A previous longitudinal study reported that compared the groups of women with T1DM and T2DM and healthy women found a positive association between maternal serum TG and LGA infants regardless of glycemic levels condition (3). Fasting maternal hypertriglyceridemia could be used as a significant predictor of LGA infants that is independent of maternal BMI, weight gain, and blood glucose levels (4). References : (1) Butler, C. L., Williams, M. A., Sorensen, T. K., Frederick, I. O., & Leisenring, W. M. (2004). Relation between maternal recreational physical activity and plasma lipids in early pregnancy. American journal of epidemiology, 160(4), 350-359. (2) Williams, M. A., Frederick, I. O., Qiu, C., Meryman, L. J., King, I. B., Walsh, S. W., & Sorensen, T. K. (2006). Maternal erythrocyte omega-3 and omega-6 fatty acids, and plasma lipid concentrations, are associated with habitual dietary fish consumption in early pregnancy. Clinical biochemistry, 39(11), 1063-1070. (3) Göbl, C. S., Handisurya, A., Klein, K., Bozkurt, L., Luger, A., Bancher-Todesca, D., & Kautzky-Willer, A. (2010). Changes in serum lipid levels during pregnancy in type 1 and type 2 diabetic subjects. Diabetes care, 33(9), 2071-2073. (4) Kitajima, M., Oka, S., Yasuhi, I., Fukuda, M., Rii, Y., & Ishimaru, T. (2001). Maternal serum triglyceride at 24–32 weeks’ gestation and newborn weight in nondiabetic women with positive diabetic screens. Obstetrics & Gynecology, 97(5), 776-780"
}
]
}
] | 1
|
https://f1000research.com/articles/9-1213
|
https://f1000research.com/articles/8-1156/v1
|
22 Jul 19
|
{
"type": "Case Report",
"title": "Case Report: Don’t chew the fufu: a case report of suspected drug bodystuffing",
"authors": [
"Naya Jimenez",
"Nguyen Toan Tran",
"Pierre-Alexandre Poletti",
"Alexandra Platon",
"Francesco Meach",
"André Juillerat",
"Laurent Getaz",
"Hans Wolff",
"Naya Jimenez",
"Pierre-Alexandre Poletti",
"Alexandra Platon",
"Francesco Meach",
"André Juillerat",
"Laurent Getaz",
"Hans Wolff"
],
"abstract": "Background: Intrabody concealment of illicit substances is a common practice in the trafficking chain. Bodystuffing, which consists of precipitously swallowing packets of substances for concealment from law-enforcement officers in anticipation of impending search or arrest, is particularly dangerous. There is a risk of rupture of the loosely wrapped drug packets, which could lead to substance intoxication or even death. Case presentation: This article reports the case of a young man who was taken by law enforcement authorities to our Emergency Department for investigation of bodystuffing. Although the patient denied the facts, the initial reading of the computed tomography (CT) scan confirmed the presence of multiple images compatible with drug packets, which were mostly in the stomach. Upon admission to our secured inpatient ward for clinical surveillance of packet evacuation, the patient denied again having ingested such packets, and declared that he only ate ‘fufu’. Fufu is a traditional food of central and western Africa consisting of a starchy preparation compacted by hand into small balls. Fufu balls are usually swallowed without chewing to allow a sensation of stomach fullness throughout the day. Considering the fufu intake history, a careful reassessment of the imaging confirmed the presence of food content. Conclusions: This case study offers an example of bodystuffing false positive due to fufu. It illustrates the importance of a history of food intake that could bias the interpretation of CT scan images.",
"keywords": [
"Bodystuffing",
"bodypacking",
"prison",
"fufu/foofoo/foufou",
"radiology pitfalls"
],
"content": "Background\n\nBodypacking is a technique used in drug trafficking that consists of deliberately ingesting many drug pellets. Bodypushing refers to the intrarectal or intravaginal insertion of pellets. These are in most cases mechanically manufactured and enclosed in multiple layers of wrapping to withstand breakage during long-distance drug smuggling routes (see Figure 1)1,2. At the end of the drug trafficking chain, street dealers or consumers who carry packets of illicit substances can resort to bodystuffing, which consists of hastily swallowing them for concealment from law-enforcement officers in anticipation of impending search or arrest2,3. Contrasting with the robustness of bodypacking pellets, bodystuffing packets are loosely wrapped in cellophane or condoms4.\n\nSample of mechanically prepared pellets (length of 4–5 cm, diameter of 1.5–2 cm) (Courtesy: NT Tran).\n\nAlthough individuals who resort to bodypacking or bodystuffing are often conveniently identified as ‘bodypackers’ or ‘body-stuffers’ by law-enforcement, criminal justice, and health professionals, these shortcuts in terminology should be avoided as they carry the potential of dehumanizing individuals with suspected or confirmed bodypacking or bodystuffing. Instead, the use of person-centered language should be favored5.\n\nPeople using bodypacking can swallow up to 1 kilogram of illicit substances divided into 50 to 100 pellets, each containing 10 to 20 grams of drugs (mostly cocaine or heroin)2,3,6. They generally consume constipating substances or spasmolytic medicine to reduce peristalsis and the risk of pellet evacuation during long travel journeys. Upon arrival, laxatives or prokinetics can be administered to accelerate the drug recovery process7,8. In Switzerland, a person arrested by law-enforcement authorities for suspected bodypacking or bodystuffing is brought to a medical facility, where an unenhanced low-dose abdominal computed tomography (CT) scan is performed to ascertain the presence of pellets3,9,10. The person has the right to refuse to undergo a CT scan and cannot be constrained to it, in which case the surveillance of pellet evacuation in a medical facility offers an alternative option11.\n\nThe rupture of drug pellets is associated with high mortality risks12. For this reason, if CT scan findings are positive, the person is transferred for medical observation to a hospital unit with ad-hoc security surveillance or, where available, to a secured hospital inpatient ward specifically dedicated to the care of people who are incarcerated. Clinical observation is ensured around the clock with vital signs and neurologic surveillance performed every 2 to 4 h (Glasgow Coma Scale and pupil reflexes). The content of the first expelled pellet is analyzed, and the nature of its substance communicated to the medical team, so that they can prepare for an adequate response in case of complications. After three bowel movements without pellets or after the evacuation of the reported number of pellets, another CT scan is performed to check for complete clearance.\n\nIn addition to the loose wrapping of the packets and the reasons for swallowing them, bodystuffing has a few other differences when compared to bodypacking. Individuals who resort to bodystuffing usually swallow a smaller number of packets, and their clinical management has been a matter of debate. Some recommend discharge after 6 h of unremarkable monitoring, while others propose a health facility-based surveillance until all packets are cleared by unaided bowel movements (no medication, such as prokinetics or laxatives)13–15. The use of mineral oil laxatives is strongly discouraged because it dissolves latex and can cause packet rupture with potentially dangerous and sometimes fatal outcomes12,15.\n\n\nCase presentation\n\nThis case involved a young and healthy man from western Africa whom the police arrested for possession of cocaine in the street of Geneva, Switzerland. Police officers also suspected him of having swallowed cocaine pellets and brought him in November 2018 to our Emergency Department (ED) for investigation of bodystuffing. A low-dose abdominal CT scan was done, showing multiple foreign bodies of similar appearance in the stomach, which the on-call radiologist reported to be compatible with ingested drug packets (Figure 2a and 2b).\n\n(a) Multiple intra-digestive foreign bodies (arrows) located in the stomach and the first part of the duodenum (oblique coronal view). (b) Multiple intra-digestive foreign bodies (arrows) in the stomach (axial view).\n\nIn the ED, relevant history included the fact that he was single and homeless, had no physical or mental health conditions, and was not on any medication. Past substance use history included tobacco and cannabis smoking, cocaine, and alcohol at the rate of 1 liter of whisky per day. Vital signs were normal, and the physical exam revealed no epigastric tenderness, no abdominal rigidity, guarding, rebound tenderness, nor evidence of a palpable mass. The remaining physical examination was unremarkable, showing a patient who was alert, oriented, calm, without any intention of self-harm, but uncooperative. The patient denied having resorted to bodystuffing, even after he was informed about the positive radiological findings in his stomach.\n\nIn accordance with our guidance on the clinical management of bodystuffing, the patient was subsequently admitted to our secured hospital inpatient ward for observation. Upon his admission to the ward, he continued to deny the ingestion of drug packets and revealed that he had consumed ‘fufu’ the previous evening. Indeed, his eating habits included only one meal a day, and this was usually a heavy starchy dish from western and central Africa called ‘fufu’. Eating fufu left him feeling full for a whole day without the need of eating again. This critical information was passed onto the radiologist who carefully reviewed the images with the attending supervisor: the foreign bodies, which were previously read as compatible with images of drug pellets, were in the process of being digested! In fact, the CT scan showed images of foreign bodies with irregular borders and of different sizes (Figure 3a and 3b). Drug pellets have clearly defined and regular edges, and, if mechanically manufactured, they would have been of the same size and shape. The continuous denial of bodystuffing by the patient, his mention of fufu, the revised radiological reading, combined with the absence of acute signs of cocaine or heroin being released from dissolving bodystuffed packets were compatible with the patient’s history of fufu intake. We immediately informed the criminal justice authorities and the patient was rapidly discharged from our secured hospital inpatient ward.\n\n(a) Heterogeneous intra-digestive foreign bodies (arrows), especially in the first part of the duodenum, signifying content dissolution by digestion (axial view). (b) Dissolved gastric foreign bodies (arrow) (axial view).\n\n\nDiscussion\n\nThis case study offered an example of a false positive CT scan reading of internally concealed illicit substances, which, to our knowledge, is the first report in the peer-reviewed literature. From this case, several points are useful to consider for the management of individuals investigated for bodystuffing.\n\nFirst, the importance of a detailed patient history upon admission to the ED cannot be overstated, in particular with regard to different food types that were ingested in the past 24 h and that could influence radiological reading. A quick 24 h food intake history would have allowed our radiologist to analyze the images with relevant information that could bias the interpretation of results.\n\nSecond, foreign bodies of similar density, such as shaped food, especially if located in the stomach, can mimic images of drug pellets. The literature reported other swallowed foreign bodies that could be misread as drug pellets, including scybala (hardened masses of feces), grains, stones, apples, or other fruits4,16. In our case, it was fufu, which is also spelled ‘foofoo’ or ‘foufou’. It is a popular dish in western and central African countries, which consists of starch (e.g., from cassava, yam, or plantain) that is boiled, pounded, and rounded into balls. Fufu balls are then dipped into sauces or eaten with stews of meat, fish, or vegetable17,18. It is a common tradition to shape fufu balls with the right hand and, with the lubricating aid of diverse sauces, to swallow them without chewing to decrease the chance of feeling hungry over a whole day18,19. The CT scan measurement of Hounsfield Units, which reflect density, could help differentiate the nature of diverse ingesta. However, measuring density is not fully reliable as the nature of the substance, its purity, admixture, and compression all play a role in imaging results20.\n\nThird, false-positive interpretation of internal concealment of illicit substances should be avoided at all costs, as it poses significant clinical and ethical issues due to an individual’s deprivation of freedom and hospitalization into a law-enforced inpatient unit. In addition, unnecessary detention results in direct and indirect costs to the health system, the criminal justice authorities, and most importantly to the person under investigation, who could be psychologically and physically impacted by the incarceration.\n\n\nConclusion\n\nThis case report showed that a careful history of food intake and sharing with the radiologist relevant information that could bias the interpretation of CT scan images are essential for the management of patients presenting with bodystuffing. If there are discrepancies between the CT scan results and the patient’s history, including the swallowing of unchewed fufu or other pellet mimicking ingesta, it is recommended that the initial image interpretation be reviewed. If needed, this could be done with the expertise of professionals who are used to read images of illicit substances that are internally concealed.\n\n\nConsent\n\nInformed written consent was obtained from the patient for publication of this case report and all accompanying images.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nCappelletti S: Medico-legal issues of body packing: what do clinicians need to know? Swiss Med Wkly. 2017; 147(37–38): w14494. PubMed Abstract | Publisher Full Text\n\nTraub SJ, Hoffman RS, Nelson LS: Body packing--the internal concealment of illicit drugs. N Engl J Med. 2003; 349(26): 2519–2526. PubMed Abstract | Publisher Full Text\n\nCappelletti S, Piacentino D, Sani G, et al.: Systematic review of the toxicological and radiological features of body packing. Int J Legal Med. 2016; 130(3): 693–709. PubMed Abstract | Publisher Full Text\n\nJalbert B, Tran NT, von Düring S, et al.: Apple, condom, and cocaine - body stuffing in prison: a case report. J Med Case Rep. 2018; 12(1): 35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTran NT, Baggio S, Dawson A, et al.: Words matter: a call for humanizing and respectful language to describe people who experience incarceration. BMC Int Health Hum Rights. 2018; 18(1): 41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCappelletti S, Piacentino D, Ciallella C: Systematic Review of Drug Packaging Methods in Body Packing and Pushing: A Need for a New Classification. Am J Forensic Med Pathol. 2019; 40(1): 27–42. PubMed Abstract | Publisher Full Text\n\nJones OM, Shorey BA: Body packers: grading of risk as a guide to management and intervention. Ann R Coll Surg Engl. 2002; 84(2): 131–2. PubMed Abstract | Free Full Text\n\nMcCarron MM, Wood JD: The cocaine 'body packer' syndrome. Diagnosis and treatment. JAMA. 1983; 250(11): 1417–1420. PubMed Abstract | Publisher Full Text\n\nSchmidt S, Hugli O, Rizzo E, et al.: Detection of ingested cocaine-filled packets--diagnostic value of unenhanced CT. Eur J Radiol. 2008; 67(1): 133–138. PubMed Abstract | Publisher Full Text\n\nHeymann-Maier L, Trueb L, Schmidt S, et al.: Emergency department management of body packers and body stuffers. Swiss Med Wkly. 2017; 147: w14499. PubMed Abstract | Publisher Full Text\n\nSwiss Academy of Medical Sciences: Annex H: Medical management of persons with suspicion of bodypacking. In. Edited by Swiss Academy of Medical Sciences. Bern: Swiss Academy of Medical Sciences; 2018.\n\nKelly J, Corrigan M, Cahill RA, et al.: Contemporary management of drug-packers. World J Emerg Surg. 2007; 2: 9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYamamoto T, Malavasi E, Archer JR, et al.: Management of body stuffers presenting to the emergency department. Eur J Emerg Med. 2016; 23(6): 425–429. PubMed Abstract | Publisher Full Text\n\nWightman RS, Nelson LS: Has the die been cast? Discharge of body stuffers from the Emergency Department. Eur J Emerg Med. 2017; 24(1): 76. PubMed Abstract | Publisher Full Text\n\nGlovinski PV, Lauritsen ML, Bay-Nielsen M, et al.: Asymptomatic body packers should be treated conservatively. Dan Med J. 2013; 60(11): A4723. PubMed Abstract\n\nFlach PM, Ross SG, Ebert LC, et al.: Response to \"the detection of internal cocaine drug packs: a radiological challenge in the future?\". Eur J Radiol. 2013; 82(9): 1588–1590. PubMed Abstract | Publisher Full Text\n\nFufu. Reference Source\n\nTozay S: The Art of Making Fufu. In: Tahoma West Literary Arts Magazine. 2007; 11. Reference Source\n\nNweke F: New challenges in the cassava transformation in Nigeria and Ghana. 2004. Reference Source\n\nFlach PM, Ross SG, Ampanozi G, et al.: \"Drug mules\" as a radiological challenge: sensitivity and specificity in identifying internal cocaine in body packers, body pushers and body stuffers by computed tomography, plain radiography and Lodox. Eur J Radiol. 2012; 81(10): 2518–2526. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "58580",
"date": "12 Feb 2020",
"name": "Nageswara Mandava",
"expertise": [
"Reviewer Expertise Surgery",
"oncology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe case study titled Don't chew the fufu: a case report of suspected bodystuffing, brings to light that taking a proper history is still paramount in the management of a patient. In this case a patient who had ingested food that was meant for slow digestion, was mistaken for drug packets. CT scan which is the go to study of choice was false positive for drug packets. Even though this is just a case report, the authors reviewed the literature on concealment of drug packets and the management of these patients. Urine toxicology screening is an important tool to use in these patients, in case the drug packets have ruptured. CT scan is still the modality of choice in diagnosing various causes for an acute abdomen, including patients with swallowed drug packets. Drug packets are wrapped and have a distinct appearance of layers with sometimes air in between the layers.\n\nThe strength of the paper is that, it reviewed the current literature of bodystuffing and points out the fallibility of CT scans.\n\nThe weakness of the paper is that it does not go into further detail of managing these patients.\n\nHowever, It is still worth indexing.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes",
"responses": []
},
{
"id": "70776",
"date": "21 Sep 2020",
"name": "Nicholas J Connors",
"expertise": [
"Reviewer Expertise Medical Toxicology",
"Emergency Medicine."
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nOverall this is a very interesting case report of a fufu ingestion that looked like professionally manufactured drug pellets on CT. It brings up some important humanistic issues including terminology for patients and the most appropriate treatment of people in custody of law enforcement. I would recommend the authors clarify the difference between body packing and body stuffing, and what the medical and law enforcement team was most concerning with in this patient.\n\nAbstract: Commonly, both terms are written as two words, body packing and body stuffing rather than bodypacking. There seems to be some confusion regarding body stuffing and body packing. The authors identify them correctly: packing refers to ingestion of large quantities of drug in packets for crossing borders and are ingested with the intention that the drug will transit the GI system; body stuffers ingest drug as a means of disposal to avoid immediate detection from law enforcement with poorly packaged drug. Body stuffers are most at risk for packet rupture and drug exposure, but should a packet rupture in a body packer, there is high risk for severe sequelae due to the shear amount of drug. In the background, this distinction seems clear, but in the case presentation, it seems there was concern by law enforcement and radiology for multiple drug containing packets, which would be most consistent with body packing.\n\nBackground Should be “body pushing”.\n\n\"Although individuals who resort to bodypacking or bodystuffing are often conveniently identified as ‘bodypackers’\" Remove “conveniently”\n\n\"The person has the right to refuse to undergo a CT scan and cannot be constrained to it, in which case the surveillance of pellet evacuation in a medical facility offers an alternative option\" Remove “option”\n\nCase Presentation \"Police officers also suspected him of having swallowed cocaine pellets and brought him in November 2018 to our Emergency Department (ED) for investigation of bodystuffing.\" This sentence seems to suggest they were more concerned about body packing. If law enforcement was really concerned about stuffing, for consistency, use “packets” as ”pellets” was used in the description of body packing.\n\n\"the foreign bodies, which were previously read as compatible with images of drug pellets, were in the process of being digested!\" Remove the exclamation point.\n\n\"Drug pellets have clearly defined and regular edges, and, if mechanically manufactured, they would have been of the same size and shape.\" Again, it seems that the concern was for body packing, not stuffing.\n\n\"In the ED, relevant history included the fact that he was single and homeless, had no physical or mental health conditions, and was not on any medication. Past substance use history included tobacco and cannabis smoking, cocaine, and alcohol at the rate of 1 liter of whisky per day. Vital signs were normal, and the physical exam revealed no epigastric tenderness, no abdominal rigidity, guarding, rebound tenderness, nor evidence of a palpable mass. The remaining physical examination was unremarkable, showing a patient who was alert, oriented, calm, without any intention of self-harm, but uncooperative.\" It seems this history and physical examination was probably performed before imaging. If so please reorder and state the radiological findings after this description.\n\nDiscussion \"First, the importance of a detailed patient history upon admission to the ED cannot be overstated, in particular with regard to different food types that were ingested in the past 24 h and that could influence radiological reading. A quick 24 h food intake history would have allowed our radiologist to analyze the images with relevant information that could bias the interpretation of results.\" Yes, though if law enforcement truly suspected body packing the history would not necessarily dissuade them for the initial CT. Was the radiologist aware of the fufu ingestion prior to the second CT and could that information have been discussed with radiology prior to the exposure of a second CT?\n\n\"professionals who are used to read images of illicit substances that are internally concealed.\" Reword to “…who are experienced reading images of illicit substances…”\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Partly",
"responses": [
{
"c_id": "6298",
"date": "29 Jan 2021",
"name": "Nguyen Toan Tran",
"role": "Author Response",
"response": "Authors reply: We thank you for your thorough review, which has greatly help clarify the manuscript. We have now addressed your comments point by point. Overall this is a very interesting case report of a fufu ingestion that looked like professionally manufactured drug pellets on CT. It brings up some important humanistic issues including terminology for patients and the most appropriate treatment of people in custody of law enforcement. I would recommend the authors clarify the difference between body packing and body stuffing, and what the medical and law enforcement team was most concerning with in this patient. - Authors reply: we have clarified the text in the abstract and the main body. In short, this was a case of suspected body stuffing by the police. The CT scan revealed images compatible with drug pellets, which were more consistent with body packing than body stuffing as initially suspected. In any case, the management consisting of surveillance until the substances are expelled remained the same. However, careful review of the patient’s history, including previous food intake, and a second CT scan ruled out concealment of illicit substances Abstract: Commonly, both terms are written as two words, body packing and body stuffing rather than bodypacking. - Authors reply: we have made recommended changes throughout the manuscript. There seems to be some confusion regarding body stuffing and body packing. The authors identify them correctly: packing refers to ingestion of large quantities of drug in packets for crossing borders and are ingested with the intention that the drug will transit the GI system; body stuffers ingest drug as a means of disposal to avoid immediate detection from law enforcement with poorly packaged drug. Body stuffers are most at risk for packet rupture and drug exposure, but should a packet rupture in a body packer, there is high risk for severe sequelae due to the shear amount of drug. In the background, this distinction seems clear, but in the case presentation, it seems there was concern by law enforcement and radiology for multiple drug containing packets, which would be most consistent with body packing. - Authors reply: we clarified the text to explain the difference between body-stuffing packets and the larger and more robust body-packing pellets. We also underscored the reason for admission, which was suspicion of body stuffing. The text now reads as follows: “Intrabody concealment of illicit substances is a common practice in the trafficking chain. Body packing is a technique used in drug trafficking that consists of deliberately ingesting many drug pellets. Body stuffing consists of precipitously swallowing packets of substances, which are smaller and more fragile than body-packing pellets, for concealment from law-enforcement officers in anticipation of impending search or arrest. Therefore, body stuffing is particularly dangerous due to the rupture risk of the loosely wrapped drug packets, which could lead to substance intoxication or even death. This article reports the case of a young man who was taken by law enforcement authorities to our Emergency Department for investigation of suspected body stuffing.” Background Should be “body pushing”. - Authors reply: amendment made \"Although individuals who resort to bodypacking or bodystuffing are often conveniently identified as ‘bodypackers’ \"Remove “conveniently” - Authors reply: amendment made \"The person has the right to refuse to undergo a CT scan and cannot be constrained to it, in which case the surveillance of pellet evacuation in a medical facility offers an alternative option\" Remove “option” - Authors reply: amendment made Case Presentation \"Police officers also suspected him of having swallowed cocaine pellets and brought him in November 2018 to our Emergency Department (ED) for investigation of bodystuffing.\" This sentence seems to suggest they were more concerned about body packing. If law enforcement was really concerned about stuffing, for consistency, use “packets” as ”pellets” was used in the description of body packing. - Authors reply: we have clarified the text, which now reads as follows: “Police officers also suspected him of having swallowed cocaine packets and brought him in November 2018 to our Emergency Department (ED) for investigation of suspected body stuffing.” \"the foreign bodies, which were previously read as compatible with images of drug pellets, were in the process of being digested!\" Remove the exclamation point. - Authors reply: amendment made \"Drug pellets have clearly defined and regular edges, and, if mechanically manufactured, they would have been of the same size and shape.\" Again, it seems that the concern was for body packing, not stuffing. - Authors reply: we have clarified the text, which now reads as follows: “Drug pellets typically found in body packing have clearly defined and regular edges, and, if mechanically manufactured, would have been of the same size and shape. The continuous denial of intrabody drug concealment by the patient, his mention of fufu, the revised radiological reading, combined with the absence of acute signs of cocaine or heroin being released from dissolving or broken ingested packets or pellets were compatible with the patient’s history of fufu intake.” \"In the ED, relevant history included the fact that he was single and homeless, had no physical or mental health conditions, and was not on any medication. Past substance use history included tobacco and cannabis smoking, cocaine, and alcohol at the rate of 1 liter of whisky per day. Vital signs were normal, and the physical exam revealed no epigastric tenderness, no abdominal rigidity, guarding, rebound tenderness, nor evidence of a palpable mass. The remaining physical examination was unremarkable, showing a patient who was alert, oriented, calm, without any intention of self-harm, but uncooperative.\" It seems this history and physical examination was probably performed before imaging. If so please reorder and state the radiological findings after this description. - Authors reply: we have reordered the text accordingly. Discussion \"First, the importance of a detailed patient history upon admission to the ED cannot be overstated, in particular with regard to different food types that were ingested in the past 24 h and that could influence radiological reading. A quick 24 h food intake history would have allowed our radiologist to analyze the images with relevant information that could bias the interpretation of results.\" Yes, though if law enforcement truly suspected body packing the history would not necessarily dissuade them for the initial CT. Was the radiologist aware of the fufu ingestion prior to the second CT and could that information have been discussed with radiology prior to the exposure of a second CT? - Authors reply: As mentioned in the introduction, standards dictate that an initial CT is done for suspected intrabody concealment of illicit substances. With regard to passing relevant information to the radiologist, we have clarified the text, which now reads as follows: “A quick 24 h food intake history would have allowed our radiologist to analyze the images with relevant information that could bias the interpretation of results. Such information was not gathered and passed onto the radiologist before the first CT scan. The history of fufu intake was shared with the radiologist before the second CT scan.” \"professionals who are used to read images of illicit substances that are internally concealed.\" Reword to “…who are experienced reading images of illicit substances…” - Authors reply: amendment made."
}
]
}
] | 1
|
https://f1000research.com/articles/8-1156
|
https://f1000research.com/articles/10-54/v1
|
28 Jan 21
|
{
"type": "Data Note",
"title": "A studyforrest extension, an annotation of spoken language in the German dubbed movie “Forrest Gump” and its audio-description",
"authors": [
"Christian Olaf Häusler",
"Michael Hanke",
"Michael Hanke"
],
"abstract": "Here we present an annotation of speech in the audio-visual movie “Forrest Gump” and its audio-description for a visually impaired audience, as an addition to a large public functional brain imaging dataset (studyforrest.org). The annotation provides information about the exact timing of each of the more than 2500 spoken sentences, 16,000 words (including 202 non-speech vocalizations), 66,000 phonemes, and their corresponding speaker. Additionally, for every word, we provide lemmatization, a simple part-of-speech-tagging (15 grammatical categories), a detailed part-of-speech tagging (43 grammatical categories), syntactic dependencies, and a semantic analysis based on word embedding which represents each word in a 300-dimensional semantic space. To validate the dataset’s quality, we build a model of hemodynamic brain activity based on information drawn from the annotation. Results suggest that the annotation’s content and quality enable independent researchers to create models of brain activity correlating with a variety of linguistic aspects under conditions of near-real-life complexity.",
"keywords": [
"annotation",
"language",
"speech",
"narrative",
"naturalistic stimulus",
"fMRI",
"studyforrest"
],
"content": "Introduction\n\nCognitive and psychiatric neuroimaging are moving towards studying brain functions under conditions of lifelike complexity1,2. Motion pictures3 and continuous narratives4,5 are increasingly utilized as so called “naturalistic stimuli”. Naturalistic stimuli are usually designed for commercial purposes and to entertain their audiences. Thus, the temporal structure of their feature space is usually not explicitly known, leading to an “annotation bottleneck”6 when used for neuroscientific research.\n\nData-driven methods like inter-subject correlation (ISC)7 or independent component analysis (ICA)8 are often used to analyze such fMRI data in order to circumvent this bottleneck. However, use of data-driven methods alone falls short of associating results with particular stimulus events9. Model-driven methods, like the general linear model (GLM), which are based on stimulus annotations can be useful to test hypotheses on specific brain functions under more ecologically valid conditions, to statistically control confounding stimulus features, and to explain not just “how” the brain is responding to a stimulus but also “why”10. Studies using GLMs based on annotations of a stimulus’ temporal structure have elucidated, for example, how the brain responds to visual features of a movie11 or speech-related features of a narrative12. Furthermore, stimulus annotations can inform data-driven methods about a stimulus’ temporal dynamics, or model-driven and data-driven methods can be combined to improve the interpretability of results13.\n\nHere we provide an annotation with exact onset and offset of each sentence, word and phoneme (see Table 1 for an overview) spoken in the audio-visual movie “Forrest Gump”14 and its audio-description (i.e. the movie’s soundtrack with an additional narrator)15. fMRI data of participants watching the audio-visual movie16 and listening to the audio-description17 are the core data of the publicly available studyforrest dataset (studyforrest.org). The current publication enables researchers to model hemodynamic brain responses that correlate with a variety of aspects of spoken language ranging from a speaker’s identity, to phonetics, grammar, syntax, and semantics. This publication extends already available annotations of portrayed emotions18, perceived emotions19, as well as cuts and locations depicted in the movie20. All annotations can be used in any study focusing on aspects of real-life cognition by serving as additional confound measures describing the temporal structure and feature space of the stimuli.\n\n\nMaterials and methods\n\nWe annotated speech in the slightly shortened “research cut”17 of the movie “Forrest Gump” and its temporally aligned audio-description16 that was broadcast as an additional audio track for visually impaired listeners on Swiss public television15. The plot of the original movie is already carried by an off-screen voice of the main character Forrest Gump. In the audio-description, an additional male narrator describes essential aspects of the visual scenery when there is no off-screen voice, dialog, or other relevant auditory content.\n\nPreliminary, manual orthographic transcripts of dialogues, non-speech vocalizations (e.g. laughter or groaning) and the script for the audio-description’s narrator were merged and converted to Praat’s21 TextGrid format. This merged transcript contained rough onset and offset timings for small groups of sentences, and was further edited in Praat for manual validation against the actual content of the audio material. The following steps were performed by a single person, already familiar with the stimulus, in several passes to iteratively improve the quality of the data: approximate temporal onsets and offsets were corrected; intervals containing several sentences were split into intervals containing only one sentence; when two or more persons were speaking simultaneously the less dominant voice was dropped; low volume non-speech vocalizations or low volume background speech (especially during music or continuous environmental noise) which were subjectively assessed to be incomprehensible for the audience were also dropped.\n\nWe then used the Montreal Forced Aligner v1.0.122 to algorithmically identify the exact onset and offset of each word and phoneme. To enable the aligner to look up the phonemes embedded within each word, we chose the accompanying German pronunciation dictionary provided by Prosodylab23 that uses the Prosodylab PhoneSet to describe the pronunciation of phonemes. To improve the detection rate of the automatic alignment, the dictionary was manually updated with German words that occur in the stimuli but were originally missing in the dictionary. The pronunciation of English words and phonemes occurring in the otherwise German audio track was taken from the accompanying English pronunciation dictionary (following the ARPAbet PhoneSet). The audio track of the audio-description was converted from FLAC to WAV via FFmpeg v4.1.424 to meet the aligner’s input requirements. This WAV file, the merged transcription, and the updated dictionary were submitted to the aligner that first trained an acoustic model on the data and then performed the alignment.\n\nThe resulting timings of words and phonemes were corrected manually and iteratively in several passes using Praat v6.0.2221: in a first step, onsets and offsets on which the automatic alignment performed moderately were corrected. Some low volume sentences that are spoken in continuously noisy settings (e.g. during battle or hurricane) were removed due to poor overall alignment performance. In a second step, the complete sentences of the orthographic transcription were copied into the annotation created by the aligner. In a third step, a speaker’s identity was added for each sentence (see Table 2 for the most often occurring speakers). During every step previous results were repeatedly checked for errors and further improvements.\n\nWe employed the Python package spaCy v2.2.125 and its accompanying German language model (de_core_news_md) that was trained on the TIGER Treebank corpus26 to automatically analyze linguistic features of each word in their corresponding sentence. Non-speech vocalizations were dropped from the sentences before analysis to improve results. We then performed analyses regarding part-of-speech (i.e. grammatical tagging or word-category disambiguation), syntactic dependencies, lemmatization, word embedding (i.e. a multi-dimensional meaning representation of a word), and if the word is one of the most common words of the German language (i.e. if the word is part of a stop list).\n\nThe annotation is available in two different versions, both providing the same information: a) as a text-based Praat TextGrid file, and b) as a text-based, tab-separated value (TSV) formatted table. The following descriptions refer to the ten columns of the TSV file, namely onset, duration, person, text, pos, tag, dep, lemma, stop, vector.\n\nStart (start)\n\nThe onset of the sentence, word or phoneme. Time stamps are provided in the format seconds.milliseconds from stimulus onset.\n\nDuration (duration)\n\nThe duration of the sentence, word or phoneme provided in the format seconds.milliseconds.\n\nSpeaker identity (person)\n\nName of the person that speaks the sentence, word or phoneme. See Table 2 for the ten most often occurring speakers.\n\nText (text)\n\nThe text of a spoken sentence or word, or the pronunciation of a phoneme. Phonemes of German words follow the Prosodylab PhoneSet, English words follow the ARPAbet PhoneSet.\n\nSimple part-of-speech tag (pos)\n\nA simple part-of-speech tagging (grammatical tagging; word-category disambiguation) of words. The tag labels of this simple part-of-speech tagging follow the Universal Dependencies v2 POS tag set (universaldependencies.org). See Table 3 for a description of the labels and the respective counts of all 15 labels. Nouns that spaCy mistook for proper nouns or vice versa were corrected via script. Additionally in cells of this column, sentences are tagged as SENTENCE, and phonemes are tagged as PHONEME to facilitate filtering in potential further processing steps.\n\nDetailed part-of-speech tag (tag)\n\nA detailed part-of-speech tagging of words following the TIGER Treebank annotation scheme26 which is based on the Stuttgart-Tübingen-Tagset27. See Table 4 for a description of the labels and the respective counts of the 15 most often occurring labels (overall 43 labels). Nouns that spaCy mistook for proper nouns or vice versa were corrected via script.\n\nSyntactic dependency (dep)\n\nInformation about a word’s syntactic dependencies with other words within the same sentence. Information follows the TIGER Treebank annotation scheme26 and is given in the format: “arc label;word’s head;word’s child1, word’s child2, ...”, where the “arc label” (see Table 5) describes the type of syntactic relation that connects a ”child” (the current word) to its “head”.\n\nLemmatization (lemma)\n\nThe base form (root) of a word.\n\nCommon Word (stop)\n\nThis column’s cell provides information if the word is part of a stop list, hence one of the most common words in the German language or not (True vs. False).\n\nWord embedding (vector)\n\nA 300-dimensional word vector providing a multi-dimensional meaning representation of a word. Out-of-vocabulary words with a vector consisting of 300 dimensions of zeroes were set to # to save space.\n\nThe annotation comes in two different versions. First, as a text-based TextGrid file (annotation/ fg_rscut_ad_ger_speech_tagged.TextGrid) to be conveniently edited using the software Praat21. Second, as a text-based, tab-separated-value (TSV) formatted table (annotation/fg_rscut_ ad_ger_speech_tagged.tsv) in accordance with the brain imaging data structure (BIDS)28. The dataset and validation data are available from Open Science Framework, DataLad and Zenodo (see Underlying data)29,30,31. The source code for all descriptive statistics included in this paper is available in code/descriptive-statistics.py (Python script).\n\n\nDataset validation\n\nIn order to assess the annotation’s quality, we investigated if contrasting speech-related events to events without speech lead to increased activation in areas known to be involved in language processing32. Moreover, we tested if two similar linguistic concepts (proper nouns and nouns) providing high semantic information contrasted with a concept providing low semantic information (coordinate conjunctions) lead to increased activation in congruent brain areas.\n\nWe used a dataset providing blood oxygenation level-dependent (BOLD) functional magnetic resonance imaging (fMRI) data of 20 subjects (age 21–38 years, mean age 26.6 years, 12 male) listening to the 2 h audio-description (7 Tesla, 2 s repetition time, 3599 volumes, 36 axial slices, thickness 1.4 mm, 1.4 × 1.4 mm in-plane resolution, 224 mm field-of-view)17. Data were already corrected for motion at the scanner computer. Further, individual BOLD time-series were already aligned by non-linear warping to a study-specific T2*-weighted echo planar imaging (EPI) group template (cf.17 for exact details).\n\nAll further steps for the current analysis were carried out using FEAT v6.00 (FMRI Expert Analysis Tool)33 as part of FSL v5.0.9 (FMRIB’s Software Library)34. Data of one participant were dropped to due to invalid distortion correction during scanning. Data were temporally high-pass filtered (cut-off 150 s), spatially smoothed (Gaussian kernel; 4.0 mm FWHM), and the brain was extracted from surrounding tissue. A grand-mean intensity normalization of the entire 4D dataset was performed by a single multiplicative factor.\n\nWe implemented a standard three-level, voxel-wise general linear model (GLM) to average parameter estimates across the eight stimulus segments, and later across 19 subjects. At the first level analyzing each segment for each subject individually, we created 26 regressors (see Table 6) based on events drawn from the annotation. The 20 most often occurring detailed part-of-speech labels (nn with N=2620 to prf with N=157) were modeled as boxcar function from onset to offset of each word. The remaining other part-of-speech labels were pooled to a single new label (tag_other; N=1123) and modeled as a boxcar function from a word’s onset to offset. The 80 most often occurring phonemes (n with N=6053 to IY1 with N=32) were pooled to phonemes (N=65251) and modeled as boxcar function from a phoneme’s onset to offset. The end of each complete grammatical sentence was modeled as an impulse event (N=1651) to capture variance correlating with sentence comprehension. “No-speech” events (no-sp; N=264) serving as a control condition were created such that a sufficient number of events and a minimum separation of speech and non-speech events were achieved. Events were randomly positioned in intervals without audible speech that lasted at least 3.6 s. Each event of the no-speech condition had to have a minimum distance of 1.8 s to any onset or offset of a word, and to any onset of another no-speech event. A length of 70 ms was chosen for no-speech events matching the average length of phonemes. Lastly, we used continuous bins of information about low-level auditory features (left-right difference in volume and root mean square energy) that was averaged across the length of every movie frame (40 ms) to capture variance correlating with assumed low-level perceptual processes. Time series of events were convolved with FSL’s “Double-Gamma HRF” as a model of the hemodynamic response function to create the actual regressors. The Pearson correlation coefficients of the 26 regressors across the time course of all stimulus segments can be seen in Figure 1. Temporal derivatives were also included in the design matrix to compensate for regional differences between modeled and actual HRF. Finally, six motion parameters were used as additional nuisance regressors and the design was subjected to the same temporal filtering as the BOLD time series. The following three t-contrasts were defined: 1) words (all 21 tag-related regressors) > no-speech (no-sp), 2) proper nouns (ne) > coordinate conjunctions (kon), and 3) nouns (nn) > coordinate conjunctions (kon).\n\nRegressors were created by convolving the events with FSL’s “Double-Gamma HRF” as a model of the hemodynamic response function, temporally filtered with the same high-pass filter (cut-off 150 s) as the BOLD time series, and concatenated across runs before computing the correlation.\n\nThe second-level analysis that averaged contrast estimates across the eight stimulus segments per subject was carried out using a fixed effects model by forcing the random effects variance to zero in FLAME (FMRIB’s Local Analysis of Mixed Effects)35,36. The third level analysis which averaged contrast estimates across subjects was carried out using a mixed-effects model (FLAME stage 1) with automatic outlier deweighting36,37. Z (Gaussianised T/F) statistic images were thresholded using clusters determined by Z>3.4 and a corrected cluster significance threshold of p<.0537. Brain regions associated with observed clusters were labeled using the Jülich Histological Atlas38,39 and the Harvard-Oxford Cortical Atlas40 provided by FSL.\n\nFigure 2 depicts the results of the three contrasts (z-threshold Z>3.4; p<.05 cluster-corrected). The contrast words > no-speech yielded four significant clusters (see Table 7): one left-lateralized cluster spanning from the angular gyrus and inferior posterior supramarginal gyrus across the superior and middle temporal gyrus, including parts of Heschl’s gyrus and planum temporale. A second left cluster in (inferior) frontal regions, including precentral gyrus, pars opercularis (Brodmann Areal 44; BA44) and pars triangularis (BA45). Similarly in the right hemisphere, one cluster spanning from the angular gyrus across the superior and middle temporal gyrus but including frontal inferior regions (pars opercularis and pars triangularis). A fourth significant cluster is located in the left thalamus.\n\nSignificant clusters (Z>3.4, p<0.05 cluster-corrected) are overlaid on the MNI152 T1-weighted head template (grey). Light grey: the audio-description dataset’s field-of-view (cf.17).\n\nThe contrast proper nouns > coordinate conjunctions yielded nine significant clusters (see Table 8): one left-lateralized cluster spanning from the angular gyrus across planum temporale and superior temporal gyrus, partially covering the Heschl’s gyrus, into the anterior middle temporal gyrus. A largely congruent but smaller cluster in the right hemisphere. Two clusters in posterior cingulate cortex and precuneus of both hemispheres. Three small clusters in the right occipital pole, right Heschl’s gyrus and left superior lateral occipital pole.\n\nThe contrast nouns > coordinate conjunctions yielded four significant clusters (see Table 9): two clusters that are slightly smaller than the lateral temporal clusters of contrast nouns > coordinate conjunction. In this case, spanning from angular gyrus in the left hemisphere and from planum temporale in the right hemisphere into the anterior part of superior temporal cortex. Finally, two small right-lateralized clusters in the right posterior cingulate gyrus and right precuneus.\n\nFor the contrast words > no-speech, results show increased hemodynamic activity in a bilateral cortical network including temporal, parietal and frontal regions related to processing spoken language32,41,42. These clusters resemble results of previous studies that implemented an ISC approach to analyze fMRI data of naturalistic auditory stimuli5,43,44. We do not find significantly increased activations in midline areas (like the posterior cingulate cortex and precuneus or anterior cingulate cortex and medial frontal cortex) which showed synchronized activity across subjects in previous studies. In this regard, our results are similar to4 who implemented both an ISC and a GLM analysis. In this study, the ISC analysis showed synchronized activity in midline areas but the GLM analysis contrasting blocks of listening to narratives to blocks of a resting condition showed significantly decreased activity in these areas.\n\nThe two contrasts that contrasted nouns and proper nouns respectively to coordinate junctions yielded increased activation partially located in early sensory regions (Heschl’s Gyrus;45) and most prominently adjacent regions bilaterally (planum temporale; superior temporal gyrus;46,47). We chose nouns and proper nouns for these two contrasts because they represent linguistically similar concepts but are uncorrelated in the German language and stimulus (cf. Figure 1). We contrasted nouns and proper nouns respectively to coordinate conjunctions because nouns and proper nouns are linguistically different to coordinate conjunctions as well as uncorrelated. Despite the fact that nouns and proper nouns are uncorrelated, both contrasts lead to largely spatially congruent clusters. Results suggest that models based on our annotation of similar linguistic concepts correlate with hemodynamic activity in spatially similar areas. We confirmed the validity of these interpretation by testing if the spatial congruency could be attributed to a negative correlation of coordinate conjunctions with the modeled time series which turned out not to be the case. In summary, results of our exploratory analyses suggest that the annotation of speech meets basic quality requirements to be a basis for model-based analyses that investigate language perception under more ecologically valid conditions.\n\n\nData availability\n\nZenodo: A studyforrest extension, an annotation of spoken language in the German dubbed movie “Forrest Gump” and its audio-description (annotation). https://doi.org/10.5281/zenodo.438214329.\n\nDataset 1. The annotation (v1.0; registered) as a tab-separated-value (TSV) formatted table and a text-based TextGrid file (the native format of the software Praat).\n\nZenodo: A studyforrest extension, an annotation of spoken language in the German dubbed movie “Forrest Gump” and its audio-description (validation analysis). https://doi.org/10.5281/zenodo.438218830.\n\nDataset 2. The data of the analysis (v1.0; registered) that we ran as a validation of the annotation’s content and quality.\n\nOpen Science Framework: studyforrest-paper-speechannotation. https://doi.org/10.17605/OSF.IO/GFRME31.\n\nThe paper as LATE X document, and accompanying datasets 1 and 2 (up-to-date; unregistered) accessible as DataLad (RRID:SCR_003931) datasets.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).\n\nCOH designed, performed, and validated the annotation, and wrote the manuscript. MH provided critical feedback on the procedure and wrote the manuscript.",
"appendix": "Acknowledgements\n\nCOH is grateful to Valeri Kippes who took care of the author’s mental sanity by providing excellent training at his gym in Jülich during the mentally draining period of manual corrections of the annotation.\n\n\nReferences\n\nSonkusare S, Breakspear M, Guo C: Naturalistic Stimuli in Neuroscience: Critically Acclaimed. Trends Cogn Sci 2019; 230(8): 699–714. PubMed Abstract | Publisher Full Text\n\nEickhoff SB, Milham M, Vanderwal T: Towards clinical applications of movie fMRI. Neuroimage 2020; page 116860. PubMed Abstract | Publisher Full Text\n\nHasson U, Landesman O, Knappmeyer B, et al.: Neurocinematics: The Neuroscience of Film. Projections 2008; 20(1): 1–26. Publisher Full Text\n\nWilson SM, Molnar-Szakacs I, Iacoboni M: Beyond Superior Temporal Cortex: Intersubject Correlations in Narrative Speech Comprehension. Cereb Cortex 2008; 180(1): 230–242. PubMed Abstract | Publisher Full Text\n\nLerner Y, Honey CJ, Silbert LJ, et al.: Topographic Mapping of a Hierarchy of Temporal Receptive Windows Using a Narrated Story. J Neurosci 2011; 310(8): 2906–2915. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAliko S, Huang J, Gheorghiu F, et al.: A “Naturalistic Neuroimaging Database” for understanding the brain using ecological stimuli. BioRxiv 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHasson U, Nir Y, Levy I, et al.: Intersubject Synchronization of Cortical Activity During Natural Vision. Science 2004; 3030(5664): 1634–1640. PubMed Abstract | Publisher Full Text\n\nBartels A, Zeki S: The chronoarchitecture of the human brain–natural viewing conditions reveal a time-based anatomy of the brain. Neuroimage May 2004; 220(1): 419–433. PubMed Abstract | Publisher Full Text\n\nKauttonen J, Hlushchuk Y, Tikka P: Optimizing methods for linking cinematic features to fMRI data. Neuroimage 2015; 110: 136–148. PubMed Abstract | Publisher Full Text\n\nHamilton LS, Huth AG: The revolution will not be controlled: natural stimuli in speech neuroscience. Lang Cogn Neurosci 2020; 350(5): 573–582. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBartels A, Zeki S: Functional brain mapping during free viewing of natural scenes. Hum Brain Mapp 2004; 210(2): 75–85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRocca R, Coventry KR, Tylén K, et al.: Language beyond the language system: dorsal visuospatial pathways support processing of demonstratives and spatial language during naturalistic fast fMRI. Neuroimage 2020; 216: 116128. PubMed Abstract | Publisher Full Text\n\nLahnakoski JM, Salmi J, Jääskeläinen IP, et al.: Stimulus-Related Independent Component and Voxel-Wise Analysis of Human Brain Activity during Free Viewing of a Feature Film. PLoS ONE 2012; 70(4): e35215. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZemeckis R Forrest Gump [motion picture] United States: Paramount Pictures; 1994.\n\nKoop O, Michalski H, Beckmann R, et al.[German audio description of the motion picture]. Hörfilm e.V. Berlin and Schweizer Radio und Fernsehen produced by Bayrischer Rundfunk; 2009.\n\nHanke M, Adelhöfer N, Kottke D, et al.: A studyforrest extension, simultaneous fMRI and eye gaze recordings during prolonged natural stimulation. Sci Data 2016; 30: 160092. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHanke M, Baumgartner FJ, Ibe P, et al.: A high-resolution 7-Tesla fMRI dataset from complex natural stimulation with an audio movie. Sci Data 2014; 1: 140003. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLabs A, Reich T, Schulenburg H, et al.: Portrayed emotions in the movie “Forrest Gump” [version 1; referees: 2 approved]. F1000Res 2015; 40(92). PubMed Abstract | Publisher Full Text | Free Full Text\n\nLettieri G, Handjaras G, Ricciardi E, et al.: Emotionotopy in the human right temporo-parietal cortex. Nat Commun 2019; 100(1): 1–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHäusler CO, Hanke M: An annotation of cuts, depicted locations, and temporal progression in the motion picture “Forrest Gump”. F1000Res 2016; 5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoersma P, Weenink D: Praat: doing phonetics by computer [Computer program] Version 6.0.22.2019 retrieved 1 September 2019 from Reference Source\n\nMcAuliffe M, Socolof M, Mihuc S, et al.: Montreal Forced Aligner [Computer program] Version 1.0.1.2019 retrieved 5 September 2019 from Reference Source\n\nGorman K, Howell J, Wagner M: Prosodylab-aligner: A tool for forced alignment of laboratory speech. Canadian Acoustics 2011; 390(3): 192–193.\n\n“FFmpeg Developers”: FFmpeg tool [Computer program] Version 4.1.4.2019Reference Source\n\nHonnibal M, Montani I: spaCy 2: Natural language understanding with Bloom embeddings, convolutional neural networks and incremental parsing [Computer program] Version 2.2.1.2017Reference Source\n\nBrants S, Dipper S, Eisenberg P, et al.: TIGER: Linguistic Interpretation of a German corpus. Research on Language and Computation 2004; (4): 597–620. Publisher Full Text\n\nSchiller A, Teufel S, Stöckert C, et al.: Guidelines für das Tagging deutscher Textcorpora mit STTS (Kleines und großes Tagset). Technical report 1999.\n\nGorgolewski KJ, Auer T, Vince D, et al.The brain imaging data structure, a format for organizing and describing outputs of neuroimaging experiments. Sci Data 2016; 3: 160044. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHäusler CO, Hanke M: A studyforrest extension, an annotation of spoken language in the German dubbed movie “Forrest Gump” and its audio-description (annotation).December 2020. Publisher Full Text\n\nHäusler CO, Hanke M: A studyforrest extension, an annotation of spoken language in the German dubbed movie “Forrest Gump” and its audio-description (validation analysis).December 2020. Publisher Full Text\n\nHäusler CO, Hanke M: studyforrest-paper-speechannotation.December 2020. Publisher Full Text\n\nHickok G, Poeppel D: The cortical organization of speech processing. Nat Rev Neurosci 2007; 80(5): 393–402. PubMed Abstract | Publisher Full Text\n\nWoolrich MW, Ripley BD, Brady M, et al.: Temporal Autocorrelation in Univariate Linear Modeling of FMRI Data. Neuroimage 2001; 140(6): 1370–1386. PubMed Abstract | Publisher Full Text\n\nSmith SM, Jenkinson M, Woolrich MW, et al.: Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage 2004; 23: 208–219. PubMed Abstract | Publisher Full Text\n\nBeckmann CF, Jenkinson M, Smith SM, et al.: General multi-level linear modelling for group analysis in FMRI. Neuroimage 2003; 200(2): 1052–1063. PubMed Abstract | Publisher Full Text\n\nWoolrich MW, Behrens TEJ, Beckmann CF, et al.: Multilevel linear modelling for FMRI group analysis using Bayesian inference. Neuroimage 2004; 210(4): 1732–1747. PubMed Abstract | Publisher Full Text\n\nWoolrich M: Robust group analysis using outlier inference. Neuroimage 2008; 410(2): 286–301. PubMed Abstract | Publisher Full Text\n\nEickhoff SB, Stephan KE, Mohlberg H, et al.: A new SPM toolbox for combining probabilistic cytoarchitectonic maps and functional imaging data. Neuroimage 2005; 250(4): 1325–1335. PubMed Abstract | Publisher Full Text\n\nEickhoff SB, Paus T, Caspers S, et al.: Assignment of functional activations to probabilistic cytoarchitectonic areas revisited. Neuroimage 2007; 360(3): 511–521. PubMed Abstract | Publisher Full Text\n\nDesikan RS, Ségonne F, Fischl B, et al.: An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage 2006; 310(3): 968–980. PubMed Abstract | Publisher Full Text\n\nFriederici AD: The Brain Basis of Language Processing: From Structure to Function. Physiol Rev 2011; 910(4): 1357–1392. PubMed Abstract | Publisher Full Text\n\nPrice CJ: A review and synthesis of the first 20 years of PET and fMRI studies of heard speech, spoken language and reading. Neuroimage 2012; 620(2): 816–847. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHoney CJ, Thompson CR, Lerner Y, Hasson U: Not Lost in Translation: Neural Responses Shared Across Languages. J Neurosci 2012; 320(44): 15277–15283. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSilbert LJ, Honey CJ, Simony E, et al.: Coupled neural systems underlie the production and comprehension of naturalistic narrative speech. Proc Natl Acad Sci U S A 2014; 1110(43): E4687–E4696. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSaenz M, Langers DRM: Tonotopic mapping of human auditory cortex. Hearing Research 2014; 307: 42–52. PubMed Abstract | Publisher Full Text\n\nArsenault JS, Buchsbaum BR: Distributed Neural Representations of Phonological Features during Speech Perception. J Neurosci. 2015; 350(2): 634–642. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMesgarani N, Cheung C, Johnson K, et al.: Phonetic Feature Encoding in Human Superior Temporal Gyrus. Science 2014; (6174): 1006–1010. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "78806",
"date": "16 Feb 2021",
"name": "Giada Lettieri",
"expertise": [
"Reviewer Expertise Neuroscience"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe current Data Note reports and describe an extensive dataset including speech annotation in the audio-visual movie \"Forrest Gump\". To assess the consistency of annotation, the brain hemodynamic activity elicited by the contrast of speech events versus no speech was also investigated. The dataset here provided is richly detailed and is a great addition to the work already done in the studyforrest project. The significant effort carried out by the authors is particularly relevant for data sharing and for future investigations on language and lifelike experiences.\nIn light of all this, I would only have two suggestions:\nI think it would help readers that are not familiar with the original dataset to have in the header of tables the indication \"run 1, run 2...\" or \"segment 1, segment 2...\", instead of just the numbers from 1 to 8.\n\nIt was not fully clear to me which kind of algorithm the authors used to obtain the word-embedding. As different algorithms provide different results for word embedding, I think it would be interesting to specify which one of these was employed in their analyses (e.g., GloVe, word2vec).\n\nIn conclusion, this extension enriches the already available annotations of the original neuroimaging dataset and concur in building a comprehensive and valuable description of a naturalistic stimulation.\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes",
"responses": []
},
{
"id": "78810",
"date": "16 Feb 2021",
"name": "Martin Wegrzyn",
"expertise": [
"Reviewer Expertise fMRI",
"language"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present a word-for-word annotation of an audio version of the movie “Forrest Gump”, as used in the “studyforrest” project, for which 7T fMRT data of 20 participants is available (as described in other publications). The presented work provides a number of syntactic and semantic labels, which can be useful for studying language using this naturalistic stimulus. An additional validation of the dataset, by using a subset of its parameters as a design for fMRI analysis, illustrates its applicability. The rationale for creating the dataset is clearly described. The protocols used are appropriate and the work is technically sound. Overall, sufficient details of methods and materials are provided for the main syntactic markers of speech. However, some of the additional semantic annotations are not described in enough detail, in my opinion. The created datasets are clearly presented and are in a usable and accessible format, especially if one is a Python user.\nThis is a high-quality work that I think will substantially boost the usability of the 2014 “studyforrest” publication by Hanke and colleagues. I think naturalistic stimuli are very valuable; Especially for data sharing, their openness to different scientific questions is a strong asset. Right now (2021), many PhD-students will have a hard time collecting fMRI data or are unable to collect data altogether. I would not be surprised if this publication (together with Hanke 2014 Scientific Reports) will give rise to a dozen or more valuable publications on language-fMRI. Having worked with some “studyforrest” data myself, and having tried to annotate the stimulus, I think I can appreciate the level of dedication required to complete an annotation at the level of the present one (I know I couldn't do it). Because a naturalistic stimulus is not designed for experimental usage, many parts of it will be more ambiguous than one would like it to be. Stemming in part from my own experience, I therefore also have a number of specific questions regarding how the authors solved those challenges.\nQuestions and comments:\nAt the end of the first paragraph/beginning of the second, the authors write: “Thus, the temporal structure of their feature space is usually not explicitly known, leading to an 'annotation bottleneck'[6] when used for neuroscientific research. Data-driven methods like inter-subject correlation (ISC)[7] or independent component analysis (ICA)[8] are often used to analyze such fMRI data in order to circumvent this bottleneck. However, use of data-driven methods alone falls short of associating results with particular stimulus events [9]” I understand the point the authors are making. Namely, that it usually takes much more time to annotate the data than to collect them. However, I would like to suggest that the wording (“temporal structure”, “feature space”, “bottleneck”), especially in the first paragraph of the article, might be a bit too challenging to immediately understand. Maybe a gentler introduction to the topic would be helpful for some readers. Also, I think that the point the authors make is not 100% valid: The study by Hasson et al. (cited in [7]) actually circumvented the problem of annotating the whole stimulus, by using “reverse correlation”: That is, only interpreting the time points where activity is highest. This allowed Hasson and colleagues to interpret effects on a substantive level, just by showing the frames of the movie for which brain activity was highest. I think this is also a useful and economical approach – but certainly more limited than a full annotation.\n\nIn the description of Table 1 the authors write: “The category 'sentences' comprises complete grammatical sentences which are additionally marked in the annotation with a full stop at the end ('my feet hurt.'). It also comprises questions ('do you want a chocolate?'), exclamations ('run away!'), or non-speech vocalizations in quick succession ('ha, ha, ha'), or in isolation (e.g. 'Forrest?', 'Forrest!', 'ha') at time points when speakers switch rapidly. The category words comprises each word or non-speech vocalization (N=202).” I completely understand that the additional coding of whether a speech episode is a legal sentence in German would go beyond the scope of the present work. Also, users will be able to code this themselves, according to their needs, using the detailed tagging provided by the authors (and the dot at the end is also useful, though not a perfect marker (e.g. “lief und lief.” or “Hörfilm e. V.”)). However, I think that calling a variable or feature “sentences”, when it really is not coding the presence of a sentence, could be a source of future errors. The same goes for “words”. So maybe the authors would want to consider using different labels for those two features.\n\nRelated to the point above, I think that sometimes a single sentence has longer pauses and is then divided into multiple “sentences”. Take the narration at the very end of the movie: “Sie wird wieder angehoben, ...” , “und schwebt auf die Kamera zu.” is coded as two sentences. Researchers interested in syntax might benefit from being aware that they have some additional work to do (which is only fair!), in order to merge those speech episodes which are really a single sentence. Additionally, if they define the sentences differently, they will have to do a different tagging of the words, too. I do not suggest that the authors need to do any additional work, but I just want to point out that the label “sentences” could mislead some language researchers. Apart from sentences being divided by pauses, there is the question of punctuation. Given that the stimulus is audio only, my personal experience is that it is sometimes difficult to decide on a definitive punctuation (where commas should occur or if a comma or a full stop should be used). I think that in German commas play a bigger role in defining the syntactic structure of a sentence than they might do e.g. in English. Therefore, I would suggest that some brief discussion of the general limitations of deriving syntactic information from an audio stimulus might be a useful addition to the article (of course only if the authors agree with my premise).\n\n“Data Legend” > “Common Word (stop)”: The source or the definition of the stop list would be helpful for users who would want to cite it in their publications\n\n“Data Legend” > “Word embedding”: This is a very rich and useful resource and I was immediately able to do some interesting analyses using the 300-dimensional word vector. For example, correlating the values of all words with each other gave some very interesting and plausible results (e.g. color words clustering together etc.). I think this itself could merit a separate publication with a focus on semantics. As it currently stands, there is however no information about how these data were generated and how future users could cite them and find out more about them. Therefore, I think that a bit more information could significantly increase the impact of this resource.\n\nAlso, when playing around with it, I noticed that sometimes two different words have the exact same 300-dimensional vector. For example: strahlend-bewölkt überqueren -durchqueren braun-olivgrün gepackten-schlamm schütteln-wischen Is this correct behavior and why would two different words have the exact same values? Especially for a pair like “gepackten”-”Schlamm” that might be surprising.\n\n“Dataset validation”: maybe some brief examples for the “coordinating conjunction” and different “nouns” could be given for illustrative purposes\n\n“Dataset validation”: why were non-speech events not modeled with a boxcar, spanning their whole duration, when this was otherwise possible for single words. It would seem to me that the duration of individual words would make a boxcar function much more problematic than using a boxcar for speech-free periods (which are often short, but there are some really long ones in there as well, which might not be fully utilized when modeled with a default length of 70ms – which would amount to a stick function).\n\n“Dataset validation”: how do the non-speech events correlate with the other features (i.e. could they be included in the heatmap in Fig 1?)\n\nTables 7-9: could it be better to round the values to full mm? I would argue that even for 7T data (smoothed with 4mm FWHM), sub-mm precision for the MNI-coordinates might imply an accuracy that is not obtainable\n\nFigure 2 legend (and pages 7, 12): I recently learned that “cf.” is used to point to contrasting information (for example, if there is literature making an opposing point; https://blog.apastyle.org/apastyle/2010/05/its-all-latin-to-me.html).\nThank you!\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? Partly\n\nAre the datasets clearly presented in a useable and accessible format? Yes",
"responses": []
},
{
"id": "78808",
"date": "01 Mar 2021",
"name": "Roberta Rocca",
"expertise": [
"Reviewer Expertise naturalistic fMRI",
"natural language processing",
"psychoinformatics"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors introduce a new set of linguistic annotations for the studyforrest dataset, an open naturalistic fMRI dataset where subjects are presented audio-visual or audio-only versions of the movie Forrest Gump. This new corpus of annotations is an extremely valuable addition to the dataset. The availability of high-quality annotations of linguistic content enables researchers to easily set up and conduct analyses tapping into the neural correlates of phonetic, semantic and syntactic processing. Furthermore, making multi-level (word-, phoneme-, and sentence-level) time-stamped transcripts of the stimulus makes it possible to extract any linguistic features beyond those directly provided by the authors. Access to the new annotations is made simple through DataLad. Additionally, the authors report the results of a few validation analyses where annotated features are used in a GLM setting to retrieve established effects (e.g., neural correlates of speech and semantic processing), which speaks in favor of their reliability. Overall, the manuscript is clear, the annotation pipelines used are solid and the data sharing format is easy to navigate. It is a great contribution to the field, and to open science culture more generally. I have a few suggestions that mostly concern: a) increasing the overall clarity and readability of the manuscript; b) making it easier for readers who are not familiar with the dataset to understand how annotations map onto the various parts of the studyforrest dataset and use them; c) making it easier for readers with limited expertise in linguistic annotations to track the meaning of individual features.\n\nIntroduction\nIn the introduction, the authors make a general point about model-driven GLM analyses providing a series of advantages over data-driven methods, especially in terms of interpretability. They write that “use of data-driven methods alone falls short of associating results with particular stimulus events”, while GLMs “allow to test hypotheses on specific brain functions under more ecologically valid conditions, to statistically control for confounding stimulus features, and to explain not just “how” the brain is responding to a stimulus, but also why”. A few objections can be moved to these claims (or to the way they are presented here). While GLMs are, on the surface, more easily interpretable, their use (especially in naturalistic contexts) poses a few challenges. For example, a) global characteristics of the stimulus make interpretation of baselines tricky (and their commensurability across datasets challenging), with consequences on the interpretation of the effects of specific features; b) the choice of covariates, while allowing to control for collinearity, can also alter the interpretation of the effects of individual features – in other words, interpretation is conditional on the model; c) causal claims (aka the why) are notoriously hard to make on the basis of linear models alone. While I am sympathetic to the idea that a lot can be gained by the use of these simple tools, authors could consider nuancing some or further specifying some of their claims.\n\nMy next (related) point concerns the link between the annotations presented by the authors and GLMs. In the introduction, the authors seem to present the usability of their annotations as almost constrained to analytic contexts where GLMs are used or as heuristics tools for data-drive methods. But nothing prevents these annotations from being used as either predictor or target features in other analysis frameworks (e.g., predictive models). The authors could consider being a bit more ambitious, and slightly reframing the first paragraphs along the lines of these considerations.\n\nIn the introduction, the authors claim that “stimulus annotations can inform data-driven methods about a stimulus’ temporal dynamics”. I think the gist of it is clear, but maybe a more concrete example or wording would help.\n\nA minor point about wording, should “the current publication” be “the present publication” or “this publication”? Not a native speaker, so it may just be an issue with my English (in which case, please ignore this comment).\n\nMaterials and methods\nIt may be beneficial to mention early in the paper that the movie is presented in German. It is also a great feature of the dataset, being it one of the very few dataset – to my knowledge – with stimuli not in English, which is quite crucial for matters of cross-linguistic generalization.\n\nOne aspect that is not entirely clear from the paper is which component of the studyforrest dataset the annotations refer to / can be used for. Can they only be used for data from subjects where the movie was only presented auditorily, or can the same annotations (filtering out “narrator” rows) also be used for those parts of the dataset where participants are presented with the movie both auditorily and visually? I think it will be highly beneficial to the manuscript to a) provide a little recap of what studyforrest is, its sub-components, and where to find them; b) make it more explicit for which batches of subjects/tasks these new annotations can be used.\n\nMore of a clarification question than a suggestion: the onsets in the annotations tsv file refer to the full movie file. It should be possible to cross-reference these onsets with run- and subject-specific events files from the BIDS dataset. If so, could more details on how to get from onsets in the annotation files to run-specific GLM-ready onsets be provided (at a high level, e.g., “onsets in the annotation files can be cross-referenced with time-stamps in the event files x and x to retrieve subject-specific and run-specific onsets …”)? Disclaimer: I am relatively new to BIDS, and this may be a trivial point (in which case ignore this comment).\n\nOnce again a minor point, but the fact that annotations are BIDS-compliant and shared in tsv could already be mentioned in the introduction (or even in the abstract).\n\nIn “Annotation procedure”, the authors introduce the fact that the stimulus is also annotated at the sentence-level. However, a definition of a sentence is only given in the description for Table 1. Could a quick reference of what counts as a sentence be added to the main text?\n\nThe authors mention having dropped the “less dominant” voice when speakers overlap. Could a couple of words be added on how that was defined (volume? character prominence?)? And since some could argue these events are also potentially interesting bits of information, do instances of overlaps represent a significant portion of the stimulus, or does this occur only in few instances?\n\nOn page 4, the authors talk about feature extraction in terms of analysis (e.g., “automatically analyze linguistic features of each word in their corresponding sentence” and performed “analyses” regarding part of speech). Probably a matter of taste, but would rather refer to these steps as “extraction” of linguistic features; to me, the use of “analysis” suggests that more has been performed on the features (e.g., visualization or some form of validation) than mere feature extraction. Which type of word embeddings were extracted should be specified.\n\nSpaCy has very good documentation explaining the interpretation of each feature and how pipelines for extraction work (e.g. https://spacy.io/usage/linguistic-features). References to that (e.g., simply links to the documentation) could be added, for example in the sentence “We then performed analyses regarding part-of-speech… “ whenever each feature/pipeline is introduced. It would make it easier for non-linguists to get a feel for what features mean, and for experts to dig into the details of the specific pipelines used.\n\nThe paragraph “Annotation procedure” may benefit from adding an example, e.g., a sample sentence from the transcript with the corresponding annotation. It would help readers (especially those who do not have deep expertise in linguistics) to visualize what features are about and visualize the different levels of annotation. I understand this may be tricky to do without disrupting the flow of the text though and will leave it to the authors to decide whether to implement it or not.\n\nIn the caption for table 2, can the expression “sentences spoken by the ten most often occurring speakers” be streamlined?\n\nCan a sample of the resulting annotation dataset (whose components are described in “data legend”) be displayed as exemplification, e.g., just the head of the table or something similar?\n\nDataset validation\nCan the (run-level) design matrix be shown, possibly with labels for the regressors? Would make for a nice visual, and help readers better understand the structure of the analysis.\n\nIt is not entirely clear to me why authors annotated for the onset/duration of phonemes, but only using the 80 most frequent phonemes. What is the role of that regressor in the model, what is it meant to code and/or control for? Can this be motivated more in detail in the text?\n\nAnother aspect of the analysis which is not entirely clear to me is why “no speech” events were created ad hoc. Wouldn’t the group-level z-map for a binary regressor coding for the onset of all speech events (or even just all words) do the same job without the need for mock events? Or couldn’t a comparison between word events and events tagged as NONSPEECH in the ‘pos’ column of the dataset be an easier implementation of the same question? The analyses are convincing and sound to me considering the purpose of this paper, so I’m not necessarily suggesting to change them. But the authors could consider further clarifying the rationale for their choices.\n\nWhen mentioning annotations of low-level auditory features, the authors refer to “continuous bins of information about low-level feature”. The wording here could be made more transparent.\n\nCan the authors provide some motivation for the choice of thresholding at z>3.4?\n\nI find the representation of contrasts in Figure 2 a bit confusing, because of the combinations of high overlaps between effects and of the colormap (not a lot of contrast between the three colors). If that is not uniquely my concern, the authors could consider either splitting the map into separate figures or experimenting with alternative color combinations.\n\nIt was a pleasure to review this work. Great set of resources, and great contribution to the field!\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-54
|
https://f1000research.com/articles/10-53/v1
|
28 Jan 21
|
{
"type": "Research Article",
"title": "Complex haploinsufficient interaction between APC11 and CYCLIN A1;2/TARDY ASYNCHRONOUS MEIOSIS in embryo development and seed germination in Arabidopsis",
"authors": [
"Ming Yang",
"Yixing Wang",
"Lei Guo",
"Chun-Ming Liu",
"Yixing Wang",
"Lei Guo",
"Chun-Ming Liu"
],
"abstract": "Background: Complex haploinsufficiency is characterized by individuals with two heterozygous loci producing a phenotype that is not seen in either of the corresponding single-locus heterozygous individuals. The mutants of the anaphase-promoting complex/cyclosome (APC/C) subunit gene APC11 and the mitotic cyclin gene CYCLIN A1;2/TARDY ASYNCHRONOUS MEIOSIS (TAM) in Arabidopsis thaliana are embryo-lethal and defective in meiosis, respectively, but their heterozygous single mutants do not exhibit defective embryo development and seed germination. Methods: Crosses between two heterozygous apc11 mutant alleles and two homozygous tam mutant alleles, and between two heterozygous apc11 mutant alleles and a TAM:TAM-GFP line were conducted. Phenotypes of the F1 seeds were analyzed by light microscopy. Results: We found that F1 embryos from the crosses between heterozygous apc11-1 (APC11/apc11-1) and homozygous tam-2 (tam-2/tam-2) or between APC11/apc11-2 and tam-2/tam-2 were morphologically normal but all the seeds failed to germinate. F1 embryos from the crosses between APC11/apc11-2 and tam-1/tam-1 (weaker allele than tam-2) produced morphologically normal seeds that germinated to form mature plants. However, F1 embryos from the crosses between APC11/apc11-1 and tam-1/tam-1 were abnormal and the seeds failed to germinate. Moreover, F1 embryos from the crosses between APC11/apc11-1 and a TAM:TAM-GFP line were arrested at early developmental stages while F1 embryos from the crosses between APC11/apc11-2 and the TAM:TAM-GFP line appeared fully developed but the seeds failed to germinate. Conclusions: Our observations indicate that the apc11 and tam mutants have an allele-dependent complex haploinsufficient relationship in embryo development and seed germination.",
"keywords": [
"anaphase-promoting complex",
"mitotic cyclin",
"complex haploinsufficiency",
"cell division",
"dominant negative mutant protein"
],
"content": "Introduction\n\nHaploinsufficiency is a frequently observed phenomenon defined by the inability of a heterozygous mutant locus to produce the same phenotype as the corresponding homozygous wild-type locus in a diploid organism. Since it was first proposed in 1932 by O. L. Mohr (Stern 1943), the word “haploinsufficiency” has long been used to simply describe the dosage effect of one or more heterozygous loci. However, in some cases, two heterozygous loci co-existing in the same organism produce a phenotype not seen in either of the corresponding single heterozygotes, a phenomenon defined by Haarer et al. (2007) as complex haploinsufficiency. Establishment of a complex haploinsufficient relationship among genes can help reveal the structure of a genetic network in a biological process especially when lethality prevents the generation and/or analysis of a double mutant of two loci.\n\nCell cycle regulators form tightly knit networks in eukaryotes (Li et al. 2004; Giotti et al. 2017; Van Leene et al. 2010). However, depending on the circumstances, functional studies of the network of cell cycle regulators in development of complex multicellular organisms are hindered by either lethality of single mutants (Murphy et al. 1997; Liu and Finley 2010; Wang et al. 2012; Wang et al. 2013; Guo et al. 2016) or non-existence of a mutant phenotype likely due to redundancy of homologous genes (Jacobs et al. 1998; Ye et al. 2001; Pérez-Pérez et al. 2008). In particular, a mutant phenotype for a mitotic cyclin in embryo development and seed germination has not been observed in plants.\n\nThe genome of Arabidopsis contains 87 core cell cycle regulators, including 10 A-type and 11 B-type mitotic cyclins (Menges et al. 2005). No mutants of these cyclins have been shown to produce an embryo-defective or a seed-germination-defective phenotype, even though gene expression studies have implicated a role in embryo development and/or seed germination for most of them, including TARDY ASYNCHRONOUS MEIOSIS (TAM)/CYCA1;2 (Masubelele et al. 2005; Narsai et al. 2011; Hofmann et al. 2019). Homozygous tam-1/tam-1 and tam-2/tam-2 mutants are defective in meiotic cell cycle progression and in nuclear size regulation in trichomes and guard cells, but no obvious defect was observed in their embryo development and seed germination (Magnard et al. 2001; Wang et al. 2004; Wang et al. 2010; Jha et al. 2014; Wang and Yang 2014). The previous studies also did not reveal any defect in heterozygous TAM/tam-1 and TAM/tam-2 plants. Thus, it is likely that functional redundancy occurs among TAM and other mitotic cyclins in embryo development and seed germination.\n\nAnother major cell cycle regulator, the anaphase-promoting complex/cyclosome (APC/C), is a large E3 ubiquitin ligase complex that mediates the ubiquitination of mitotic cyclins and other cell cycle regulators, which leads to degradation of these regulators by the proteasome (Sudakin et al. 1995; Irniger et al. 1995; King et al. 1995; Yamano 2019). Degradation of cell cycle regulators via the APC/C in a temporal order allows the cell cycle to exit mitosis and prevents the reentry into the next S-phase of the cell cycle (Yamano 2019). On the other hand, cyclin As and/or Bs activate the APC/C via phosphorylation of the APC3 and APC1 subunits or an inhibitor of the APC/C activator CDC20 by cyclin-CDK complexes (Lahav-Baratz et al. 1995; Kramer et al. 2000; Rudner and Murray 2000; Dienemann and Sprenger 2004; Zhang et al. 2016). Cyclin As can also inhibit premature activation of the APC/C by phosphorylating the APC/C activators CDH1 and CDC20 (Sørensen et al. 2001; Hein and Nilsson 2016). Although TAM has not been demonstrated to be a substrate of the APC/C, the CDKA1-TAM complex can phosphorylate in vitro OSD1, a negative regulator of the APC/C in Arabidopsis (Iwata et al. 2011; Cromer et al. 2012). However, in Arabidopsis, unlike single mutants of mitotic cyclins, several single mutants of subunits of APC/C are gametophytic-lethal or embryo-lethal (Capron et al. 2003; Kwee and Sundaresan 2003; Wang et al. 2012; Wang et al. 2013; Guo et al. 2016). In particular, embryo development is arrested at the zygote stage in the homozygous apc11-1/apc11-1 and apc11-2/apc11-2 mutants of the APC11 subunit of the APC/C although heterozygous APC11/apc11-1 and APC11/apc11-2 embryos develop normally (Guo et al. 2016). The zygotic lethality of the apc11 mutants hinders the study of the effect of the mutations on subsequent development in Arabidopsis.\n\nIn an attempt to investigate how TAM and APC11 genetically interact in Arabidopsis development, we conducted genetic crosses between two tam mutants and two apc11 mutants and between a transgenic line containing the TAM:TAM-GFP transgene and the apc11 mutants. Our findings revealed a strong complex haploinsufficient relationship between TAM and APC11 in both embryo development and seed germination, and a dominant-negative effect arising from the coexistence of the apc11-1 and the tam-1 alleles or of the apc11-1 allele and the TAM:TAM-GFP transgene.\n\n\nMethods\n\nThe origins and initial characterization of the mutants and the TAM:TAM-GFP line used in this investigation were as described in our previous work (Magnard et al. 2001; Wang et al. 2004; Wang et al. 2010; Guo et al. 2016). Briefly, tam-1 and apc11-1 harbor point mutations that were generated by chemical mutagenesis (Magnard et al. 2001; Guo et al. 2016), and tam-2 and apc11-2 harbor T-DNA insertions (Wang et al. 2004; Guo et al. 2016), in the TAM and APC11 loci, respectively. The TAM:TAM-GFP line has been characterized with respect to its expression in male meiocytes and vegetative organs and its complementation of the meiotic defect in tam-1 (Wang et al. 2004; Jha et al. 2014). All seeds were air dried at least two weeks before sowing. The plants were grown in soil (Sunshine MVP growing Mix, Sungro Horticulture, Agawam, Massachusetts, USA) or on a medium containing 4.3 g/L Murashige and Skoog (MS) salt base (catalogue number 11117-074, Gibco, Waltham, MA, USA), 1% (w/v) sucrose (catalogue number S1695, Spectrum Quality Products, Gardena, CA, USA), and 0.7% (w/v) agar (catalogue number PTC001, Gaisson Laboratories, Smithfield, UT, USA) in a growth chamber at ~22°C. Daily light regime in the growth chamber was 16h fluorescent light supplemented with tungsten light (light intensity = ~50 µmol·m-2·s-1 near the plants) and 8h darkness.\n\nSix types of crosses between a heterozygous apc11 mutant and a homozygous tam mutant or the TAM:TAM-GFP line were conducted. Buds of the tam mutants and the TAM:TAM-GFP line that were about to open were emasculated and their stigmas pollinated with pollen from the APC11/apc11-1 or APC11/apc11-2 plants. At least three fertile siliques were obtained from each type of the crosses. Care was taken to remove other buds from the same inflorescences to facilitate harvesting the seeds from the crosses.\n\nThe imbibed seeds were dissected with a pair of 30G½ syringe needles under either a Leica S6D or a Nikon SMZ1000 dissecting microscope with a magnification range of 2x to 4x. Photographs of the seeds were taken using the Leica EC3 imaging system on the Leica S6D microscope. Embryos were observed and photographed on a Nikon Eclipse 80i microscope equipped with the differential interference contrast (DIC) optics and the Nikon DS-Ri1 imaging system. The images shown in the figures were lightly adjusted for contrast and brightness and made black and white on the entire image area in Adobe Photoshop CC 2014.\n\n\nResults\n\nTo investigate how apc11 mutants genetically interact with tam mutants, we crossed APC/11apc11-1 and APC11/apc11-2 individuals with tam-2/tam-2, respectively. Each of the crosses resulted in more than 50 F1 seeds. However, these seeds, after fully dried, failed to germinate in more than 14 days on the MS agar medium, in contrast with the corresponding wild-type (Col-0) seeds germinating within 2-3 days (Wang et al. 2016). The germination defect observed did not result from abnormally underdeveloped embryos because embryos dissected out from these seeds appeared fully developed (Figure 1). Because the pollen donor plants for these crosses, APC11/apc11-1 or APC11/apc11-2, were heterozygous for the apc11 mutation as the homozygous mutant plants are not available due to embryo lethality, the genotypes of the F1 seeds should be either APC11/APC11;TAM/tam-2, APC11/apc11-1;TAM/tam-2, or APC11/apc11-2;TAM/tam-2. Because seeds heterozygous for any of the three mutations alone did not exhibit such a severe germination defect (Wang et al. 2010; Guo et al. 2016), the germination defect observed indicates complex haploinsufficiency between apc11-1 and tam-2, and apc11-2 and tam-2. Moreover, because all the F1 seeds failed to germinate and approximately half of their embryos should have the genotype of APC11/APC11;TAM/tam-2, our results suggest that APC11 was also insufficient even in the APC11/APC11;TAM/tam-2 embryos.\n\n(a-c) From a cross between APC11/apc11-1 and tam-2/tam-2. (a) A representative embryo from an imbibed seed that failed to germinate. (b) and (c) Close-up views of a cotyledon and the root of the embryo in (a), respectively. (d-f) From a cross between APC11/apc11-2 and tam-2/tam-2. (d) A representative embryo from an F1 seed. (e) and (f) Close-up views of a cotyledon and the root of the embryo in (d), respectively. The embryo in (d) was larger than that in (a) presumably due to an unreduced egg frequently resulting from the tam-2 mutation (Wang et al. 2010; d'Erfurth et al. 2010). Scale bar in (a) for (a) and (d) = 200 µm, and scale bar in (b) for (b), (c), (e), and (f) = 50 µm.\n\nTo provide additional evidence for the complex haploinsufficient relationship between APC11 and TAM, we crossed the APC11/apc11-1 and APC11/apc11-2 plants with another mutant, tam-1/tam-1, respectively. The tam-1 mutation is a point mutation that does not result in unreduced gametes as tam-2 and it produces normal diploid embryos and germinated seeds (Wang et al. 2004; Wang et al. 2010). Surprisingly, F1 seeds from the crosses between APC11/apc11-1 and tam-1/tam-1 varied in size and showed variable degrees of shrinkage (Figure 2a). These seeds also failed to germinate like the F1 seeds from the crosses between APC11/apc11-1 and tam-2/tam-2. However, unlike the F1 seeds from the crosses between APC11/apc11-1 and tam-2/tam-2, small and abnormal embryos were readily found after dissection of these seeds (Figure 2c-f) even though some embryos were morphologically normal (Figure 2f). These observations indicate that apc11-1 and tam-1 have a complex haploinsufficient relationship in embryo development and seed germination, given that APC11/apc11-1 and TAM/tam-1 individuals do not have such defects (Magnard et al. 2001; Wang et al. 2004; Guo et al. 2016). The level of complex haploinsufficiency between apc11-1 and tam-1 is more severe than those between the two apc11 mutants and tam-2. The apc11-1 and apc11-2 mutations are also a point mutation and a T-DNA insertion mutation, respectively, with apc11-2 having a more severe embryo defective phenotype than apc11-1 (Guo et al. 2016). Therefore, the most severe complex haploinsufficiency was found with two relatively weak point-mutation alleles.\n\n(a) and (c-f) From a cross between APC11/apc11-1 and tam-1/tam-1. (a) Seeds of variable sizes and morphologies. (c-f) Embryos from imbibed seeds that failed to germinate, showing variable sizes and morphologies. The variation in seed or embryo size should not have resulted from variable ploidy levels of the egg cells because the tam-1 mutant, unlike the tam-2 mutant, produces only haploid gametes (Magnard et al. 2001; Ajay et al. 2014). (b) Seeds from a cross between apc11-2 and tam-1. Scale bar in (a) for (a) and (b) = 1 mm, and scale bar in (c) for (c-f) = 200 µm.\n\nThe F1 seeds from the cross between APC/11/apc11-2 and tam-1/tam-1 appeared normal (Figure 2b) and germinated and developed into mature plants. This result is consistent with tam-1 being a mild allele that did not result in complex haploinsufficiency in the double heterozygous F1 individuals during embryo development and seed germination.\n\nIn an attempt to investigate how the TAM protein is affected in the apc11 mutants, we crossed APC11/apc11-1 and APC11/apc11-2 with a homozygous TAM:TAM-GFP line in the Col-0 wild-type background, TAM-GFP/TAM-GFP, respectively. Surprisingly, for each type of the crosses, none of the more than 30 F1 seeds germinated after two weeks on the MS agar medium. Seeds from the crosses between APC11/apc11-1 and TAM-GFP/TAM-GFP appeared small andshrank (Figure 3a). Dissection of their non-germinated seeds revealed that the embryos were arrested at either theapical cell stage (Figure 3b) or the early globular stage (Figure 3c). Most of the F1 seeds from the crosses between APC11/apc11-2 and TAM-GFP/TAM-GFP were similar in size to normal seeds in two dimensions (length and width) but abnormally thin in the third dimension (depth), i.e., with a flattened appearance (Figure 3d). Some of these seeds were apparently abnormally small (Figure 3d). Dissection of the non-germinated seeds produced embryos at an advanced developmental stage (Figure 3e-g). The TAM:TAM-GFP transgene alone does not cause an embryo developmental arrest or a seed non-germination defect (Wang et al. 2004). Therefore, the defects observed suggest a dominant negative effect of the TAM-GFP protein in the F1 seeds where the APC11 protein was presumably reduced. In fact, these defects parallel those of the corresponding F1s from the crosses between APC11/apc11-1 and tam-1/tam-1, and between APC11/apc11-2 and tam-2/tam-2, respectively, although they were more severe with the TAM:TAM-GFP transgene than with the tam mutations.\n\n(a-c) From a cross between APC11/apc11-1 and TAM-GFP/TAM-GFP. (a) Small and shriveled seeds. (b) A developmentally arrested embryo from an imbibed seed that failed to germinate, showing an enlarged apical cell attached to the suspensor. (c) A developmentally arrested embryo from an imbibed seed that failed to germinate, showing an early globular embryo attached to the suspensor. (d-g) From a cross between APC11/apc11-2 and TAM-GFP/TAM-GFP. (d) Flattened seeds. (e) A representative embryo from an imbibed seed that failed to germinate. (f) and (g) Close-up views of a cotyledon and the root of the embryo in (e), respectively. Scale bar in (a) for (a) and (d) = 1 mm, scale bar in (e) = 200 µm, and scale bar in (b) for (b), (c), (f) and (g) = 50 µm.\n\n\nDiscussion\n\nOur findings of complex haploinsufficiency between pairs of apc11 mutations and tam mutations in both embryo development and seed germination are consistent with the knowledge that APC/C and mitotic cyclins reciprocally regulate each other’s functions in cell cycle progression in eukaryotes. The role of TAM in embryo development and seed germination and the role of APC11 in seed germination have not been reported before. Therefore, testing if two mutations have a complex haploinsufficient relationship is a useful approach for functional studies of genes acting in embryo development and/or seed germination. This approach can work for a mutant such as tam-1 or -2 that by itself does not have an obvious defect in embryo development and seed germination. It can also work for an embryo lethal mutant such as apc11-1 or -2 for demonstrating its effect on seed germination.\n\nIn Arabidopsis, there are many known recessive embryo lethal mutants (Meinke 2020) and other recessive mutants (e.g. mutants available from the Arabidopsis Biological Resource Center, Columbus, OH) that do not exhibit defects in embryo development and/or seed germination and yet their wild-type genes are expressed in the embryo (Narsai et al. 2011; Hoffmann et al. 2019). Any pairs of mutants from these two categories may be tested in the F1 generation for complex haploinsufficiency if the paired mutants are suspected to affect the same developmental process based on available information. This approach should circumvent the difficulty of identifying the double mutant of such a pair of mutants and be faster than the double mutant analysis because it involves only the F1 seeds.\n\nThe tam-2 and apc11-2 alleles are more severe alleles than the tam-1 and apc11-1 alleles, respectively (Magnard et al.; zWang et al. 2004; Wang et al. 2010; Guo et al. 2016), but the defects exhibited in the F1 seeds of this investigation were most severe between tam-1 and apc11-1, comparing to between tam-2 and apc11-2, tam-1 and apc11-2, and tam-2 and apc11-1. The tam-2 and apc11-2 alleles are T-DNA insertion mutants that likely do not produce the TAM and APC11 proteins, whereas the tam-1 and apc11-1 alleles are expected to produce the mutant proteins because of the nature of their point mutations generated from chemical mutagenesis. Therefore, the presence of both mutant proteins correlates with the most severe defects in comparison with the other combinations without the presence of both mutant proteins. This phenomenon, in principle, may be explained by reciprocal dominant-negative effects of the apc11-1 and the tam-1 mutant proteins. As discussed earlier, both positive and negative interplays exist between the APC/C and mitotic cyclins. This relationship may underpin the phenomenon that the presence of both mutant proteins has a greater functional impact than the presence of one, or absence of both, of the mutant proteins. Interestingly, the TAM:TAM-GFP line produced even more severe complex haploinsufficient-like phenotypes than tam-1 and tam-2. This finding further supports the idea that the presence of a malfunctioned protein, which is assumed to be the case for the TAM-GFP protein, is more devastating to the plant than without such a protein, especially when in coexistence with the apc11-1 mutant protein.\n\nBecause the crosses in the current investigation were conducted with APC11/apc11-1 and APC11/apc11-2, approximately one half of the F1 plants should be double heterozygous for the apc11 and tam mutations and the other half homozygous wild type at the APC11 locus and heterozygous for one of the tam mutations. Similar allele frequencies should occur in F1s of the crosses between the apc11 mutants and the TAM:TAM-GFP line. However, when a defect was present, it was observed in all F1 seeds of the crosses albeit the severity of embryo developmental defects, if existed, varied. This observation indicates that reconstituting homozygous APC11 from fertilization could not rescue the F1s from abnormal embryo development or seed non-germination in the TAM/tam background. A plausible explanation for this outcome in the F1s is that the pollen grains used in the crosses had a reduced level of the APC11 protein or APC11 transcript due to heterozygosity at the APC11 locus in the parental tissue, and the APC11 protein or APC11 transcript could not recover to the normal level in the haploid pollen grains even though they had a wild-type allele of APC11. If so, embryo development and seed germination seem to be sensitive to simultaneous reductions in APC11 and TAM levels or to a reduction in the APC11 level along with the presence of a malfunctioned TAM-GFP.\n\nThe roles of genes encoding core cell cycle regulators in Arabidopsis seed germination have been demonstrated for two D-type cyclins genes, CYCD4:1 and CYCD1;1 (Masubelele et al. 2005). Expression of these genes was found to be activated before the onset of cell division in the radicle, and the earliest cell divisions in the radicle preceded root emergence from the seed coat that defines the completion of germination. Cell divisions in the radicles of the cycd4:1 and cycd1;1 mutants were reduced, which correlated with the delayed root emergence from their seed coats (Masubelele et al. 2005). Interestingly, activation of expression of two A-type cyclin genes, one of which is TAM (the other CYCA3:4), was also found along with the earliest expressed D-type cyclin genes (Masubelele et al. 2005). The findings from this investigation support the roles of these A-type cyclins in seed germination. Furthermore, the non-germination phenotype associated the complex haploinsufficiency lends support to the view that cell divisions are required for seed germination. Detailed genetic analysis with combinations of the D-type and A-type cyclin mutants may shed more light on the essential role of cell division in seed germination in Arabidopsis.\n\n\nData availability\n\nFigshare: Complex haploinsufficiency between TAM and APC11.\n\nhttps://doi.org/10.6084/m9.figshare.13574891.\n\nThis project contains the following underlying data:\n\nImage files of F1 dry seeds and imbibed non-germinated seeds, and embryos dissected from imbibed non-germinated seeds.\n\nData are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).\n\n\nAuthor contributions\n\nMY conceived the idea and designed the experiments. MY and YW performed the experiments. MY analyzed the experiments. LG and C-ML contributed the apc11 mutants. MY wrote the manuscript with input from YW, LG, and C-ML.",
"appendix": "References\n\nCapron A, Serralbo O, Fülöp K, et al.: The Arabidopsis anaphase-promoting complex or cyclosome: molecular and genetic characterization of the APC2 Subunit. Plant Cell 2003; 15: 2370–2382. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCromer L, Heyman J, Touati S, et al.: OSD1 promotes meiotic progression via APC/C inhibition and forms a regulatory network with TDM and CYCA1;2/TAM. PLoS Genet 2012; 8: e1002865. PubMed Abstract | Publisher Full Text | Free Full Text\n\nd'Erfurth I, Cromer L, Jolivet S, et al.: The cyclin-A CYCA1;2/TAM is required for the meiosis I to meiosis II transition and cooperates with OSD1 for the prophase to first meiotic division transition. PLoS Genet 2010; 6: e1000989. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDienemann A, Sprenger F: Requirements of cyclin a for mitosis are independent of its subcellular localization. Curr. Biol 2004; 14: 1117–1123. PubMed Abstract | Publisher Full Text\n\nGiotti B, Anagha J, Freeman TC: Meta-analysis reveals conserved cell cycle transcriptional network across multiple human cell types. BMC Genomics 2017; 18: 30. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGuo L, Jiang L, Zhang Y, et al.: The anaphase-promoting complex initiates zygote division in Arabidopsis through degradation of cyclin B1. Plant J 2016; 86: 161–174. PubMed Abstract | Publisher Full Text\n\nHaarer B, Viggiano S, Hibbs MA, et al.: Modeling complex genetic interactions in a simple eukaryotic genome: actin displays a rich spectrum of complex haploinsufficiencies. Genes Dev 2007; 21: 148–159. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHein JB, Nilsson J: Interphase APC/C-Cdc20 inhibition by cyclin A2-Cdk2 ensures efficient mitotic entry. Nat. Commun 2016; 7: 10975. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHofmann F, Schon MA, Nodine MD: The embryonic transcriptome of Arabidopsis thaliana. Plant Reprod 2019; 32: 77–91. PubMed Abstract | Publisher Full Text\n\nIrniger S, Piatti S, Michaelis C, et al.: Genes involved in sister chromatid separation are needed for B-type cyclin proteolysis in budding yeast. Cell 1995; 81: 269–278. PubMed Abstract | Publisher Full Text\n\nIwata E, Ikeda S, Matsunaga S, et al.: GIGAS CELL1, a novel negative regulator of the anaphase-promoting complex/cyclosome, is required for proper mitotic progression and cell fate determination in Arabidopsis. Plant Cell 2011; 23: 4382–4393. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJacobs HW, Knoblich JA, Lehner CF: Drosophila Cyclin B3 is required for female fertility and is dispensable for mitosis like Cyclin B. Genes Dev 1998; 12: 3741–3751. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJha AK, Wang Y, Hercyk BS, et al.: The role for CYCLIN A1; 2/TARDY ASYNCHRONOUS MEIOSIS in differentiated cells in Arabidopsis. Plant Mol. Biol 2014; 85: 81–94. PubMed Abstract | Publisher Full Text\n\nKing RW, Peters JM, Tugendreich S, et al.: A 20S complex containing CDC27 and CDC16 catalyzes the mitosis-specific conjugation of ubiquitin to cyclin B. Cell 1995; 81: 279–288. PubMed Abstract | Publisher Full Text\n\nKramer ER, Scheuringer N, Podtelejnikov AV, et al.: Mitotic regulation of the APC activator proteins CDC20 and CDH1. Mol. Biol. Cell 2000; 11: 1555–1569. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKwee HS, Sundaresan V: The NOMEGA gene required for female gametophyte development encodes the putative APC6/CDC16 component of the anaphase promoting complex in Arabidopsis. Plant J 2003; 36: 853–866. PubMed Abstract | Free Full Text\n\nLahav-Baratz S, Sudakin V, Ruderman JV, et al.: Reversible phosphorylation controls the activity of cyclosome-associated cyclin-ubiquitin ligase. Proc. Natl. Acad. Sci. USA 1995; 92: 9303–9307. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi F, Lung T, Lu Y, et al.: The yeast cell-cycle network is robustly designed. Proc. Natl. Acad. Sci. USA 2004; 101: 4781–4786. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu D, Finley RL Jr: Cyclin Y is a novel conserved cyclin essential for development in Drosophila. Genetics 2010; 184: 1025–1035. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMagnard JL, Yang M, Chen YCS, et al.: The Arabidopsis gene TARDY ASYNCHRONOUS MEIOSIS is required for the normal pace and synchrony of cell division during male meiosis. Plant physiol 2001; 127: 1157–1166. PubMed Abstract | Free Full Text\n\nMasubelele NH, Dewitte W, Menges M, et al.: D-type cyclins activate division in the root apex to promote seed germination in Arabidopsis. Proc. Natl. Acad. Sci. USA 2005; 102: 15694–15699. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMeinke DW: Genome-wide identification of EMBRYO-DEFECTIVE (EMB) genes required for growth and development in Arabidopsis. New Phytol 2020; 226: 306–325. PubMed Abstract | Publisher Full Text\n\nMenges M, de Jager SM, Gruissem W, et al.: Global analysis of the core cell cycle regulators of Arabidopsis identifies novel genes, reveals multiple and highly specific profiles of expression and provides a coherent model for plant cell cycle control. Plant J 2005; 41: 546–566. PubMed Abstract | Publisher Full Text\n\nMurphy M, Stinnakre MG, Senamaud-Beaufort C, et al.: Delayed early embryonic lethality following disruption of the murine Cyclin A2 gene. Nat. Genet 1997; 15: 83–86. PubMed Abstract | Publisher Full Text\n\nNarsai R, Law SR, Carrie C, et al.: In-depth temporal transcriptome profiling reveals a crucial developmental switch with roles for RNA processing and organelle metabolism that are essential for germination in Arabidopsis. Plant Physiol 2011; 157: 1342–1362. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPérez-Pérez JM, Serralbo O, Vanstraelen M, et al.: Specialization of CDC27 function in the Arabidopsis thaliana anaphase-promoting complex (APC/C). Plant J 2008; 53: 78–89. PubMed Abstract | Publisher Full Text\n\nRudner AD, Murray AW: Phosphorylation by Cdc28 activates the Cdc20-dependent activity of the anaphase-promoting complex. J. Cell Biol 2000; 149: 1377–1390. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSørensen CS, Lukas C, Kramer ER, et al.: A conserved cyclin-binding domain determines functional interplay between anaphase-promoting complex-Cdh1 and cyclin A-Cdk2 during cell cycle progression. Mol. Cell Biol 2001; 21: 3692–3703. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStern C: Genic Action as studied by means of the effects of different doses and combinations of alleles. Genetics 1943; 28: 441–475. PubMed Abstract | Free Full Text\n\nSudakin V, Ganoth D, Dahan A, et al.: The cyclosome, a large complex containing cyclin-selective ubiquitin ligase activity, targets cyclins for destruction at the end of mitosis. Mol. Biol. Cell 1995; 6: 185–197. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVan Leene J, Hollunder J, Eeckhout D, et al.: Targeted interactomics reveals a complex core cell cycle machinery in Arabidopsis thaliana. Mol. Syst. Biol 2010; 6: 397. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang Y, Jha AK, Chen R, et al.: Polyploidy-associated genomic instability in Arabidopsis thaliana. Genesis 2010; 48: 254–263. PubMed Abstract | Publisher Full Text\n\nWang Y, Magnard JL, McCormick S, et al.: Progression through meiosis I and meiosis II in Arabidopsis anthers is regulated by an A-type cyclin predominately expressed in prophase I. Plant physiol 2004; 136: 4127–4135. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang Y, Yang M: Loss-of-function mutants and overexpression lines of the Arabidopsis cyclin CYCA1;2/TARDY ASYNCHRONOUS MEIOSIS exhibit different defects in prophase-I meiocytes but produce the same meiotic products. PloS One 2014; 9: e113348. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang Y, Hou Y, Gu H, et al.: The Arabidopsis APC4 subunit of the anaphase-promoting complex/cyclosome (APC/C) is critical for both female gametogenesis and embryogenesis. Plant J 2012; 69: 227–240. PubMed Abstract | Publisher Full Text\n\nWang Y, Hou Y, Gu H, et al.: The Arabidopsis anaphase-promoting complex/cyclosome subunit 1 is critical for both female gametogenesis and embryogenesis. J. Integr. Plant Biol 2013; 55: 64–74. PubMed Abstract | Publisher Full Text\n\nWang Z, Chen F, Li X, et al.: Arabidopsis seed germination speed is controlled by SNL histone deacetylase-binding factor-mediated regulation of AUX1. Nat. Commun 2016; 7: 13412. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYamano H: APC/C: current understanding and future perspectives. F1000Res 2019; 8: F1000 Faculty Rev-725. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYe X, Zhu C, Harper JW: A premature-termination mutation in the Mus musculus Cyclin-Dependent Kinase 3 gene. Proc. Natl. Acad. Sci. USA 2001; 98: 1682–1686. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang S, Chang L, Alfieri C, et al.: Molecular mechanism of APC/C activation by mitotic phosphorylation. Nature 2016; 533: 260–264. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "78412",
"date": "22 Feb 2021",
"name": "Xing-You Gu",
"expertise": [
"Reviewer Expertise Genetics and seed molecular biology"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGeneral comments:\n\nThe reported research addressed a topic known as “complex haploinsufficiency” using a genetic approach in Arabidopsis thaliana. Six crosses were made between laboratory lines that were different in the genotype for the Tam and APC11 loci involved in the regulation of cell cycling or embryo development. Female parents for the crosses were genotypes homozygous for the induced mutant tam1 or tam2 at the Tam locus or for the transgene TAM:TAM-GFP; the male parents were genotypes heterozygous for the APC11 locus, which contained an induced mutant apc11-1 or apc11- 2. Hybrid F1 seeds, i.e., those from plants of the female parents, were evaluated for germination ability and/or morphologies of the embryos. The F1 seed samples from all these crosses failed to germinate on a medium for >14 days or varied in the embryo size or morphology. These observations were explained by interactions between the two loci on the defect embryo development and the failure of germination, presumably due to the induced mutations at Tam and APC11. Results from this research will be interesting to researchers who are working on gene regulatory networks of embryo development and seed biology. A major weakness of this manuscript is lack of quantitative data for the complex phenotypes.\nSpecific comments:\nBased on descriptions in the title, abstract and introduction, “complex haploinsufficiency” appears similar to the phenomenon that has been reported for distant crosses between remote ecotypes of a species or between closely related species, in which the F1 hybrids often fail to set seeds or display partial sterility. It is suggested to define the concepts “haploinsufficiency”, “complex haploinsufficiency” and “complex haploinsufficiency interaction”, explicitly to help more readers or researchers better understand the importance of this work, and to help refine the objective and improve the title, abstract and discussions of the manuscript.\n\nThis research was conducted using a genetic approach. To help improve the M&M section and the data presentation, it is strongly suggested to develop a table to summarize key information for and from each of the 6 crosses, such as genotypes for the parents and F1s, statistics of phenotypic data (e.g., germination percentage and embryo size) from the F1s, and citations for data from the parental genotypes observed in the previous but not in this research to support the conclusions or hypotheses.\n\nThe F1 seed samples were segregating for the embryo and endosperm genotypes. More discussions are needed to address the potential influence of the segregation on the data annotation and conclusion.\n\nPage 5 “In an attempt to investigate how the TAM protein is affected in the apc11 mutants, we crossed APC11/apc11-1 and APC11/apc11-2 with a homozygous TAM:TAM-GFP line in the Col-0 wild-type background, TAM-GFP/TAM-GFP, respectively”. It is suggested to indicate the GFP signal in related figures and the difference between the F1 TAM-GFP and APC11 genotypes and the F1 TAM-GFP and APC11-1/2 genotypes.\n\nThe failure of seed germination can arise from underdeveloped embryos, or other genetic and environmental reasons, such as dormancy and the requirement of a stratification treatment in Arabidopsis. It is suggested to include these factors in experimental design, data analysis and data annotation.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "6395",
"date": "01 Mar 2021",
"name": "Ming Yang",
"role": "Author Response",
"response": "We understand the reviewer's criticisms and suggestions for further investigation into questions arising from the current results. However, for various reasons, we are not planning to continue this work. We acknowledge that the presentation of results in our paper deviates from the norm that is often found in papers in the related fields. We agree with the reviewer that additional experiments can be done to strengthen the conclusions, but we think the current results can justify for the conclusion that complex haploinsufficiency exists between the TAM locus and the APC11 locus. Our results show 100% seed non-germination in five of the six crosses, and 100% or frequent occurrences of abnormal seeds or embryos in three of the six crosses. Even though the amount of data in our paper is limited, we think that the reported findings are valuable to researchers in the related fields. Other points: Our findings are not similar to the sterility phenomenon in F1 plants from crosses between ecotypes or species. All the plant lines in our paper are in the same ecotype, Columbia (see references in our paper). Furthermore, the defective seeds were the F1 generation that never had the chance to grow up. We missed the opportunity to examine the GFP signal during embryo development in the seeds from the crosses involving TAM:TAM-GFP because we did not suspect these seeds would not germinate later. We don't think that the non-germination phenotype in our experiments resulted from environmental factors, dormancy due to incomplete drying, or the lack of a stratification treatment as all seeds were subjected to the same handling process and seeds from one of the six crosses germinated normally."
}
]
},
{
"id": "78944",
"date": "26 Feb 2021",
"name": "Takashi Hirayama",
"expertise": [
"Reviewer Expertise Plant molecular genetics",
"plant hormones",
"plant stress responses."
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study described a genetic interaction between the mutations of the genes encoding CYCLIN A1;2/TAM and APC11, of Arabidopsis. The author showed that the F1 between APC11+/- and TAM-/- failed to germinate, even though half of the seeds should be APC11+/+ TAM+/-. They used different type of alleles and obtained the similar results. The results described might be interesting, showing the abnormal genetic interaction, haploinsufficiency as the author mentioned. However, the author just described the morphological phenotypes of F1 seeds presenting the unclear photographs of seeds, embryos, and enlarged some tissues. I do not understand what the authors wanted to mention with these photographs. I think these data cannot help readers understand the relation between these two loci. If the author thinks the seed shape is important, they can categorize them and show their frequencies. That might give a clue to understand what has happened.\n\nHaploinsufficiency can be explained by the alteration of the gene or gene-product dosage. As the author described, if the APC and cyclin somehow interact each other, this can be happen due to their importance for cell proliferation. I guess, they can test this idea by introducing the WT transgene. Anyway, I think the author should re-consider the approach to the mechanisms of this genetic interactions.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": [
{
"c_id": "6394",
"date": "01 Mar 2021",
"name": "Ming Yang",
"role": "Author Response",
"response": "We understand the reviewer's criticisms and suggestions for further investigation into questions arising from the current results. However, for various reasons, we are not planning to continue this work. We acknowledge that the presentation of results in our paper deviates from the norm that is often found in papers in the related fields. We agree with the reviewer that additional experiments can be done to strengthen the conclusions, but we think the current results can justify for the conclusion that complex haploinsufficiency exists between the TAM locus and the APC11 locus. Our results show 100% seed non-germination in five of the six crosses, and 100% or frequent occurrences of abnormal seeds or embryos in three of the six crosses. Even though the amount of data in our paper is limited, we think that the reported findings are valuable to researchers in the related fields."
}
]
}
] | 1
|
https://f1000research.com/articles/10-53
|
https://f1000research.com/articles/9-599/v1
|
12 Jun 20
|
{
"type": "Research Article",
"title": "Eggshell membrane for DNA sexing of the endangered Maleo (Macrocephalon maleo)",
"authors": [
"Pramana Yuda",
"Andie Wijaya Saputra",
"Andie Wijaya Saputra"
],
"abstract": "Background: Noninvasive DNA sampling has been applied across many avian genetic studies for a variety of purposes including conservation and management of endangered birds. However, its application in megapodes is still lacking. The previous genetic studies on megapodes used either blood or fresh tissue. Here we present the first demonstration of the use of eggshell membrane for research on endangered Maleo (Macrocephalon maleo). Methods: We used 24 post-hatched eggshell membranes collected from two different sites, Tambun and Tanjung Binerean, in North Sulawesi, 12 samples in each. Two different DNA extraction methods: alkaline lysis method and gSYNCTM DNA Extraction Kit were applied. To determine the sex of Maleo, we utilized PCR-based DNA sexing using CHD genes, with the primer set 2550F/2718R. Results: We successfully extracted all samples; the mean sample concentration was 267.5 ng/µl (range 47–510.5 ng/µl) and samples were of high purity (A260/280 ratio 1.85±0.03). All samples were used to successfully identified sexes, 9 females and 15 males. Conclusions: Our research clearly illustrates that eggshell membranes can be used for DNA sexing and open the possibility to build noninvasive DNA collections over large spatial scales for population study of endangered birds.",
"keywords": [
"eggshell membrane",
"endangered birds",
"DNA sexing",
"Maleo"
],
"content": "Introduction\n\nStudies on molecular ecology have a great impact on our knowledge on ecology and evolution of animals, i.e. the phylogenetic relationships and systematics of organisms, population genetics, mating systems, micro-evolutionary processes and host-parasite interactions1–4. Oftentimes a necessary prerequisite for answering evolutionary or ecological questions is access to a good DNA sample. Birds’ blood contains nucleated red blood cells with abundant DNA, making it a preferred source of DNA5. However, obtaining blood requires the capture of a bird, which can provoke an increased level of stress and might results in unusual behavior or nest desertion. For example, blood sampling has been reported to reduce annual survival in Cliff Swallows (Petrochelidon pyrrhonota)6, although the effects of blood collection in free-living adult and immature birds is not thought to have major negative effects on adult survival, reproductive success, body condition, or behavior7.\n\nAs an alternative to invasive sampling, researchers have adopted noninvasive sampling methods such as DNA capture from molted feathers8,9, feces10,11, and egg shell membrane12,13. In addition the application of moderately invasive sampling such as buccal swabs14,15 has also increased.\n\nMegapodes (family Megapodidae) are a galliform clade, centered in Australasia16, that are known for their unique super-precocial behavior17. Their ground-living habits, large body size and large egg size make them particularly vulnerable to human persecution, habitat destruction and habitat loss: 11 out of 21 species are now considered endangered or threatened in some form18. Given this precarious conservation situation, the application of noninvasive DNA sampling techniques is crucial for megapode birds. Yet previous genetic studies on this family have used either blood19 or fresh tissue20.\n\nPreviously, all megapodes were assumed to be monogamous. The mating system is considered to correlate with sexual selection, with sexually dimorphic birds are non-monogamous and monomorphic birds are monogamous. The evolution of non-monogamous systems in birds was believed to be an adaptive solution to an unbalanced sex ratio21. The sex-ratio in Maleo (Macrocephalon maleo) is unknown, but based on previous assumptions, it is expected that the Maleo has an evenly balanced sex ratio. Even though Maleo are slightly sexual dimorphic, the available population data only report total population size and never mention sex ratio. A study on the correlation of incubation temperature and sex ratio of chicks has been carried out in the Australian brush-turkey (Alectura lathami), which revealed that at average temperature the hatched chicks in the proportion of 1:1 of male and female chicks22.\n\nThe purpose of our study was to determine whether the eggshell membrane of the endangered Maleo, a monotypic genus within the megapodes, could be successfully extracted and amplified for DNA sexing. Adult male and female Maleo are morphologically slightly different, but the chicks are not. To determine the sex in Maleo chicks, vent sexing has been conducted. Base on cloaca size and shape, a one-day-old male Maleo chick cloaca is bigger (3.96 ± 0.11 cm) and rounded, than the female cloaca (3.20 ± 0.10 cm), which is more oval in shape. The concentration of estrogen in female birds was also higher23. Until recently, no molecular technique has been applied for sex determination of Maleo.\n\nMaleo are endemic to Sulawesi, Indonesia24,25. The bird is a burrow-nesting megapode that incubates its eggs in communal nesting sites on beaches (coastal nesting grounds) and in soil heated by volcanic activity mostly at inland localities. Due to its small, severely fragmented population and continued rapid decline, the International Union for Conservation of Nature has classified Maleo as an endangered species26. Among the major threats are the over-exploitation of eggs and loss of connectivity between forest and nesting grounds27. To minimize these threats at some nesting grounds, conservation programs are currently removing eggs and hatching them in safer, semi-natural hatcheries, built close to the nesting grounds. These facilities provide an opportunity to collect noninvasive DNA samples from the eggshell membrane left in the soil or brought to the soil surface by the hatched Maleo.\n\n\nMethods\n\nPost-hatched egg-shell membranes were collected from semi-natural hatcheries of Maleo at two different nesting grounds: an inland geothermal heated nesting ground at Tambun (Bogani Nani Wartabone National Parks and a sun-heated sand beach nesting ground at Tanjung Binerean, North Sulawesi, Indonesia (Figure 1). All samples were collected from 4th April until 1st May 2018. To prevent post-sampling contamination, each sample was placed separately in a zip-lock plastic bag and stored in silica gel for delivery to laboratory. The samples were stored at -40°C until DNA extraction were conducted.\n\nWe used two different DNA extraction methods: the alkaline lysis method and gSYNC™ DNA Extraction Kit (Genaid). For the alkaline lysis method, we followed the recommended procedure for rapid preparation of mouse tails or nail lysates suitable for amplification using DNA polymerase from hyperthermophilic archaeon Pyrococcus kodakaraensis (KOD FX Neo 1103; TOYOBO Co. Ltd.). The eggshell membrane (20–25 mg; mostly with dry allantois blood vessel) was grinded using a micro-pestle in a 1.5 mL microcentrifuge tube; next, 180 µL NaOH (50 mM) was added, the suspension mixed thoroughly by vortexing and then incubated at 90°C in water-bath for 10–29 min. Following this, 20 µL Tris-HCl (1 M, pH 8.0) was added and the tube was vortexed thoroughly, then centrifuged at 12,000 RPM for 5 min. Supernatant was removed to new 1.5 mL microtube and store at freezer until used for PCR.\n\nMeanwhile the protocol for gSYNC™ DNA Extraction Kit (Genaid) followed the provided user manual with little modifications. The eggshell membrane (25 mg; mostly with dry allantois blood vessel) was grinded using a micro-pestle in a 1.5 mL microcentrifuge tube; 300 µl GST Buffer (Tris, SDS) and 30 µl Proteinase K (10 mg/ml) was added to the sample mixture, mixed thoroughly by vortexing and incubated at 60°C in water-bath overnight or until the tissue was lysed completely. Next, 200 µl GSB Buffer was added to the sample mixture, mixed thoroughly by pulse-vortexing and incubated at 70°C for 10 min. After this, 200 µl ethanol (100%) was added to the sample mixture, which was mixed thoroughly by pulse-vortexing and brief spinning of the tube to remove drops from the inside of the lid. Next, a GS Column was placed in a Collection Tube and the mixture (including any precipitate) was carefully transferred to the GS Column, which was centrifuged at 14,000 RPM for 1 min then the GS Column was placed in a new Collection Tube. Following this, 400 µl W1 Buffer was added to the GS Column and centrifuged at 14,000 RPM for 1 min then flow-through was discarded. Next, 750 µl Wash Buffer was added to the GS Column, centrifuged at full speed for 1 min, then the flow-through was discarded, the tube centrifuged at 14,000 RPM for an additional 3 min to dry the column, 50 µl of preheated Elution Buffer (pH 7.5–9.0) added to the membrane of the GS Column. The GS Column was then left to stand for 3 min, following a final centrifugation at full speed for 2 min to elute the DNA.\n\nThe eluded DNA (1 µl) was quantified using NanoVue Plus™ (Biochrom, Harvard Bioscience, Inc), at A260 nm. The 260/280 nm absorbance ratio was also measured to give an indication of purity of the DNA. Pure DNA has expected ratios of 1.7–1.9.\n\nTo determine the sex of Maleo, we applied PCR based DNA sexing by using CHD genes, with the primer set 2550F/2718R28. PCR used a 10 µl total volume containing template DNA (genomic DNA or lysate), 1.2 µl sterile dH2O, 5 µl 2x PCR buffer KOD FX Neo, 2 µl dNTPs (2 mM), (TOYOBO Co. Ltd.), 0.3 µl Primer 2550F (10 µM; 5'-GTT ACT GAT TCG TCT ACG AGA-3'), and 0.3 µl Primer 2718R (10 µM; 5'-ATT GAA ATG ATC CAG TGC TTG-3',28), and 0.2 U KOD polymerase enzyme. PCR was carried out in a Veriti™ 96-well thermal cycler (Applied Biosystems™). For genomic DNA templates, the following profile was used: 1 cycle at 94°C for 2 min followed by 35 cycles of 98°C for 10 sec, 53°C for 30 sec and 68°C for 45 sec;, and a final extension at 68°C for 7 minutes. For lysate as DNA template, the PCR profiles was the same for DNA genome, except that it was run for more cycles (40x). We employed eggshell membrane from female domestic chicken as positive control, and tube without sample as negative one.\n\nAmplification of CHD Genes were resolved on a 2% agarose gel. Electrophoresis was conducted using TAE (0.5×) buffer, stained by ethidium bromide (1%), at 100 V for 30 minutes; and 5 µl PCR product was mixed with 1 µl loading dye. After finish, the gel was visualized and analyzed on Gel Logic 200 Imaging System and Kodak Molecular Imaging Software. To confirm that the amplified fragments were the CHD genes, the PCR products of one male and one female sample, respectively, were sequenced. The gels were cut on upper and lower bands for female samples and the single band for male sample, then purified for sequencing. The sequence reactions were carried for both direction in sequencing services laboratory provided by 1st BASE Laboratories (Apical Scientific Sdn Bhd, Malaysia). The sequences were check and edited manually on Bioedit version 7.0.5.3 (Hall, 1999) and Chromas versi 2.6.5 (Technelysium Pty Ltd). Sequence similarity was probed using NCBI BLAST (Zhang et al., 2000).\n\nMann-Whitney U-test (at a significance level of 95%) was performed to assess the DNA concentration differences which were collected on Tanjung Binerean and Tambun. Analysis was performed using SPSS v17.\n\n\nResults\n\nAll eggshell membranes were successfully extracted, with mean DNA concentrations around 267.5 ng/µl (range 47–510.5 ng/µl). The average DNA concentration extracted from eggshell membrane collected from coastal nesting grounds (Tanjung Binerean: 213±179 ng/µl,) was significantly lesser than of that of inland nesting grounds (Tambun: 322±153 ng/µl, p=0.004; Data Supp.1). These results demonstrate that all samples were adequate for further PCR based analysis.\n\nThe 260/280 nm absorbance ratio of all samples ranged from 1.81 to 1.89, with an average of 1.85 (±0.03). Meanwhile the average for Tambun and Tanjung Binerean samples were, respectively, 1.85 (±0.03) and 1.84 (±0.01); Table 1). This result suggested good purity of DNA extracted from eggshell samples. However. gel visualization of extracted DNA showed smears in all samples (Figure 2), pointing to some DNA degradation.\n\nAbove: Tambun; Below: Tanjung Binereaan; LD: 100 bp DNA Ladder (SMOBiO); AK01: domestic chicken eggshell membrane.\n\nOut of 24 samples in which extracted DNA was used as a template, one did not amplify. Meanwhile all samples based on lysate were successfully amplified. There was complete agreement in gender determination across all Maleo samples that were run with different DNA template (Figure 3). Females showed two bands (545 bp and 395 bp), whereas males exhibited one band (545 bp). Three sequences of CHD1 genes have been deposited in GenBank (accession numbers MT074328, MT074329 and MT074330). Sequence similarity searches on the upper band revealed a match with CHD-Z genes of other bird species (i.e. Anser cygnoides, Anser reevesii, Anas penelope). Meanwhile the lower band matched CHD-W submissions of other birds (i.e. Gallus gallus, Crossoptilon mantchuricum, Syrmaticus reevesii).\n\nAK1, positive control, female chicken; K, negative control, no template.\n\nIn total 9 samples were identified as females and 15 were males (Table 2). Based on this limited sample, the sex ration of Maleo’s chicks in Tambun and Tanjung Binerean is biased towards males. The sex ratio males to females was 1.6, and significantly different from the hypothetic sex ratio 1:1 for Maleo.\n\n\nDiscussion\n\nThis study has demonstrated the first successful DNA isolation from eggshell membranes of a megapode bird. Our success rate (100%) compares favorably to that of previous avian eggshell membrane studies of Black-tailed Godwits (Limosa limosa), which also successfully extracted DNA from all 47 eggshell membranes13. The freshness of the samples might be one of the determinant factors of DNA extraction success. Our samples were relatively fresh, extracted 5–15 days after collection and kept in the freezers until extraction, with no concomitant extraction failure and high purity (1.85 ±0.03) and concentration (267.5 ng/µl) of DNA.\n\nPCR amplification of CHD genes succeeded in 96% of the eggshell membrane samples, with only one eggshell membrane isolate out of 24 failing to amplify. However, using lysate as DNA template for PCR resulted in 100% amplifications across 24 samples. This result show that eggshell membrane isolates yielded DNA with little amplification problems. Compared to blood DNA isolates, eggshell membrane DNA isolates of Black-tailed Godwits (Limosa limosa) also yielded fewer amplification problems13. One eggshell membrane DNA isolate out of 21 and 3 samples of Black-tailed Godwit (Limosa limosa) did not amplify for 2 and 5 of the 11 microsatellite loci. The amplification success rates was 99.1%13. Meanwhile the success rate of eggshell membrane of Sage Grouse (Centrocercus uropihasianus) for DNA sexing was only 55.6%12.\n\nThis study demonstrates that hatched eggshell membrane provides useful noninvasive DNA material as an alternative to invasive sampling in sex determination studies of Maleo. Information of the sex of the hatched eggs are important to understand demographic issues, such as the demographic consequences of offspring sex ratio bias or whether there is any sex-specific mortality or dispersal. Furthermore, this information is very important for translocation programs of endangered species, including Maleo.\n\nThis study provides additional evidence that noninvasive DNA samples yield reliable results and eliminating the need for capture and invasive sampling. Collection of post-hatched eggshell membrane of Maleo, and other megapodes does not require specific skills. This noninvasive DNA sampling also open the possibility to build participation of local community or local conservation area staff on DNA collections over large spatial scales. Furthermore, the collected samples provide sufficient samples required for population and other ecological and evolutionary study of endangered bird species.\n\n\nData availability\n\nNCBI GenBank: Macrocephalon maleo isolate MT10-FTB_UAJY_01 chromo-helicase DNA binding protein (CHDZ) gene, partial cds. Accession number MT074328.\n\nNCBI GenBank: Macrocephalon maleo isolate MB01-FTB_UAJY_02 chromo-helicase DNA binding protein (CHDZ) gene, partial cds. Accession number MT074329.\n\nNCBI GenBank: Macrocephalon maleo isolate MT10-FTB_UAJY_03 chromo-helicase DNA binding protein (CHDZ) gene, partial cds. Accession number MT074330.\n\nMendeley Data: Eggshell membrane for DNA sexing of the endangered Maleo (Macrocephalon maleo). http://doi.org/10.17632/mjp5n9pcj3.\n\nThis project contains the following underlying data:\n\nElectrophoresis photos. (Folder containing photos of DNA extraction and PCR sexing gels.)\n\nSequences. (Folder containing raw sequencing files.)\n\nData hosted with Mendeley Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Acknowledgements\n\nWe would like to acknowledge Elisabeth Purastuti and Hanom Basri (EPASS Project), Iwan Hunowu and Alfons Patandung (WCS-IP Sulawesi) and Bogani Nani Wartabone National Park staff for their support during collecting samples. We also thank to Harry Marshal and Frank E. Rheint for reviewing the manuscript.\n\n\nReferences\n\nZink RM, Blackwell-Rago RC: Species limits and recent population history in the Curve-billed Thrasher. Condor. 2000; 102(4): 881–6. Publisher Full Text\n\nYuda P: High Prevalence Level of Avian Malaria in the Wild Population of the Java Sparrow. Biota. 2009; 14(3): 198–200. Reference Source\n\nKüpper C, Augustin J, Kosztolányi A, et al.: Kentish versus Snowy Plover: phenotypic and genetic analyses of Charadrius alexandrinus reveal divergence of Eurasian and American subspecies. Auk. 2009; 126(4): 839–52. Publisher Full Text\n\nYuda P: Detection of avian malaria in wild birds at Trisik Beach of Yogyakarta, Java (Indonesia). Ann Parasitol. 2019; 65(2): 171–5. PubMed Abstract | Publisher Full Text\n\nColquitt BM, Mets DG, Brainard MS: Draft genome assembly of the Bengalese finch, Lonchura striata domestica, a model for motor skill variability and learning. Gigascience. Oxford University Press; 2018; 7(3): 1–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrown MB, Brown CR: Blood sampling reduces annual survival in Cliff Swallows (Petrochelidon Pyrrhonota). Auk. 2009; 126(4): 853–61. Publisher Full Text\n\nSheldon LD, Chin EH, Gill SA, et al.: Effects of blood collection on wild birds: An update. J Avian Biol. 2008; 39(4): 369–78. Publisher Full Text\n\nMiño CI, Del Lama SN: Molted feathers as a source of DNA for genetic studies in Waterbird Populations. Waterbirds. 2009; 32(2): 322–9. Publisher Full Text\n\nRudnick JA, Katzner TE, Bragin EA, et al.: Species identification of birds through genetic analysis of naturally shed feathers. Mol Ecol Notes. 2007; 7(5): 757–62.Publisher Full Text\n\nRytkönen S, Vesterinen EJ, Westerduin C, et al.: From feces to data: A metabarcoding method for analyzing consumed and available prey in a bird-insect food web. Ecol Evol. 2019; 9(1): 631–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJoo S, Park S: Identification of bird species and their prey using DNA barcode on feces from Korean traditional village groves and forests (maeulsoop). Animal Cells Syst (Seoul). 2012; 16(6): 488–97. Publisher Full Text\n\nBush KL, Vinsky MD, Aldridge CL, et al.: A comparison of sample types varying in invasiveness for use in DNA sex determination in an endangered population of greater Sage-Grouse (Centrocercus uropihasianus). Conserv Genet. 2005; 6(5): 867–70. Publisher Full Text\n\nTrimbos KB, Broekman J, Kentie R, et al.: Using eggshell membranes as a DNA source for population genetic research. J Ornithol. 2009; 150(4): 915–20. Publisher Full Text\n\nHandel CM, Pajot LM, Talbot SL, et al.: Use of Buccal Swabs for Sampling DNA from Nestling and Adult Birds. Wildl Soc Bull. 2006; 34(4): 1094–100. Publisher Full Text\n\nVilstrup JT, Mullins TD, Miller MP, et al.: A simplified field protocol for genetic sampling of birds using buccal swabs. Wilson J Ornithol. 2018; 130(1): 326–34. Publisher Full Text\n\nElliott A: Megapodes (Megapodiidae). (eds.) del Hoyo J, Elliott A, Sargatal J, Christie DA, de Juana E. editors. Handbook of the Birds of the World Alive. Barcelona.: Lynx Edicions; 2020. Publisher Full Text\n\nStarck JM, Ricklefs RE: editors. Avian Growth and Development Evolution within the Altrical-Precocial Spectrum. Oxford University Press, UK; 1998. Reference Source\n\nIUCN: The IUCN Red List of Threatened Species. Version. 2019-3. 2020. Reference Source\n\nBudiarsa IM, Artama IWT, Sembiring L, et al.: Diversitas genetik burung Maleo (Macrocephalon maleo) berdasarkan sekuen gen dehydrogenase sub-Unit 2 (ND2) mitokondria. Berk Penelit Hayati Ed Khusus. 2009; 3B: (June): 11–5. Reference Source\n\nBirks SM, Edwards SV: A phylogeny of the megapodes (Aves: Megapodiidae) based on nuclear and mitochondrial DNA sequences. Mol Phylogenet Evol. 2002; 23(3): 408–21. PubMed Abstract | Publisher Full Text\n\nJones DN: An evolutionary approach to megapode mating systems. Zool Verh. 1992; 278: 33–42. Reference Source\n\nGöth A, Booth D: Temperature-dependent sex ratio in a bird. Biol Lett. 2005; 1(1): 31–3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWidnyana IGNP, Sundu B, Tanari M: Sex detection in Maleo Bird (Macrocephalon Maleo Sal Muller 1846) nurtured in ex-situ conservation through body morphological and hormonal studies. Int J Vet Sci Agric Res. 2019; 1(2): 17–22. Reference Source\n\nElliott A, Kirwan GM: Maleo (Macrocephalon maleo). In: del Hoyo J, Elliott A, Sargatal, J., Christie, D.A. & de Juana E (eds. )., editors. Handbook of the Birds of the World Alive. Lynx Edicions, Barcelona. 2020. Publisher Full Text\n\nYuni LPEK, Yuda IP: The Island Biogeography of Wallacea and Krakatoa Island. In: Reference Module in Earth Systems and Environmental Sciences. Elsevier. 2019. Publisher Full Text\n\nBirdLife International: Threatened Birds of Asia: the BirdLife International Red Data Book. Cambridge, UK: BirdLife International. 2001. Publisher Full Text\n\nFroese GZL, Mustari AH: Assessments of Maleo Macrocephalon maleo nesting grounds in South-east Sulawesi reveal severely threatened populations. Bird Conserv Int. 2019; 29(4): 497–502. Publisher Full Text\n\nFridolfsson AK, Ellegren H: A simple and universal method for molecular sexing of non-ratite birds. J Avian Biol. 1999; 30: 116–21. Publisher Full Text"
}
|
[
{
"id": "69914",
"date": "08 Sep 2020",
"name": "Sena Adi Subrata",
"expertise": [
"Reviewer Expertise Wildlife molecular ecologist"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn general, the article provided a new finding that never been reported, and discussed possible contribution of the finding to the species conservation. However, there was unclear information particularly in the methods section that should be resolved before the article can be published.\nHere is my suggestion:\nIn the introduction section, a correction is needed on the family name of the bird.\n\nIn methods section, particularly genetic sampling, a further explanation is required regarding sample collection: how to collect the membranes ? How old is the samples ? Furthermore, the author should refer supernatant as lysate. Volume and DNA concentration should be explicitly mentioned when describing PCR condition. Method of sex determination of positive control (egg shell membrane of chicken) should be described. Report of statistical test should be removed because it was meaningless without argument in the discussion.\n\nThe author should mention and analyze a failed sample in the result and discussion sections. Most likely it correlated with DNA quality and quantity. When comparing sex ratio, it should be done in the same way.\n\nIn the discussion section, particularly when discussion freshness of a sample, the author should provide data on the old of the sample. In the Figure 3, a scale indicating length of the amplicon/ladder should is required.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "5979",
"date": "02 Oct 2020",
"name": "Pramana Yuda",
"role": "Author Response",
"response": "Thank you for reviewing our article and some valuable recommendations for improvement. We have incorporated the correction and recommendations in our revised article. Family name was corrected In method section we added description of sample age, collection technique, the volume and concentration of the DNA, and the more description on positive control for sex determination. We also removed the statistical test. Failed sample (MB09) in sex determination was added in ‘Sex determination’ section and discussion. Length of amplicon was added in Figure 3 The description on age and freshness of samples were added in method and discussion sections."
}
]
}
] | 1
|
https://f1000research.com/articles/9-599
|
https://f1000research.com/articles/9-1227/v1
|
12 Oct 20
|
{
"type": "Research Article",
"title": "Applying an ecological framework to examine the multiple levels of influence affecting the utilisation of private sector adult asthma services in Khartoum, Sudan: a mixed methods study",
"authors": [
"Rachael Thomson",
"Magde Noor",
"Asma Elsony",
"Magde Noor",
"Asma Elsony"
],
"abstract": "Background: Asthma is the third most common cause of hospital visits in Sudan. Sudan has a pluralistic health care system, with a strong and varied private sector. While research examining public sector asthma services exists, very little is known about which asthma services are available in the private sector.\n\nMethods: An explanatory sequential mixed-method social ecological approach was used to examine influencing factors of asthma service utilisation in the private sector, considering five levels: policy, organisational, community, familial, individual environment. Quantitative research involved surveying private healthcare facilities to describe asthma services. Qualitative research involved in-depth interviews with asthma patients to explore facility decision-making. Nine private chest clinics, 44 pharmacies, and 21 private hospitals offering asthma services in Khartoum were studied - 46 female and 28 male health providers were surveyed; 7 male and 7 female asthma patients were interviewed.\n\nResults: At the health policy level, there is no current asthma management policy for the private sector. At the organisational and health systems level, the survey found low rates of diagnostic equipment available, little asthma-specific training, and little use of asthma treatment cards, guidelines, and registers. At the community level, high levels of stigma from the community were felt by most of the patients interviewed. At the familial level, asthma was often viewed as a hereditary condition, and, as a long-term condition, there were worries about marriage potential and impact on jobs/future activities. At the individual level, patients sought frequent, short-term care at private facilities for acute attacks. The severity of the disease and the major impact it had, particularly on younger adults’, was striking.\n\nConclusions: Applying an ecological framework to examine asthma care management enables review of all levels of service provision: inclusive health policy, government commitment, high quality service delivery, uninterrupted affordable drug supply, community involvement, and patient empowerment.",
"keywords": [
"Sudan",
"Asthma",
"Health seeking",
"Access",
"Pathway to care"
],
"content": "Introduction\n\nRespiratory diseases now account for 17.4% of all deaths in sub-Saharan Africa1. In 1990, there were 74.4 million reported cases of asthma in Africa. This figure rose to 94.8 million cases in 2000 and 119.3 million in 20102. However, due to low levels of accurate diagnosis, the true prevalence of the disease is expected to be much higher. It is estimated that the number of cases will continue to increase from the currently estimated global prevalence of 330 million to reach 400 million by 20253.\n\nThe majority of people living with asthma in low and middle income countries (LMICs) have very limited access to essential prevention and treatment4. There is a high financial burden on individuals living with the disease, on their families, and on health-care systems at-large5. Inadequate treatment, coupled with high financial healthcare costs often results in high levels of disability, absenteeism, and increased risk of poverty. If the disease is left untreated, it can become much more severe and result in hospitalisation, or, in poorly controlled cases, become fatal. Severe uncontrolled asthma results in frequent hospitalisations and emergency room visits, disproportionately affecting individuals of lower socioeconomic status6. The total cost of asthma treatment and care is estimated to be at least USD $20 billion annually in LMICs5.\n\nIn addition to a lack of political commitment and a chronic shortage of asthma drugs in LMICs, there is also the issue of weak health-care systems that are unable to support patients, particularly for chronic diseases7. Other key obstacles to asthma care are availability of equipment, poor chronic care education of health-care workers, weak asthma case notification/referral systems, and issues in implementing long-term care in a poorly functioning health-care system7. In addition to weak health-care system support for asthma patients, it has been shown that there are sociocultural misconceptions attached to the disease and to the use of inhaled steroids, which significantly affects adherence to treatment2.\n\nUntreated asthma and poor management of the disease can lead to frequent visits to hospital emergency rooms8. There are many factors that lead patients to seek care in the emergency room, including asthma severity, incorrect perceptions of the disease and its medication, lack of an asthma treatment plan, over-reliance on short acting bronchodilators, changes in weather and pollution, as well as education and socioeconomic levels. Avoiding crisis care of acute asthma through long-term management and reducing the use of emergency rooms for acute asthma treatment are major goals of asthma management9. To examine this issue, changes in hospital admissions over time may be used as an indirect indicator of the burden of severe asthma10.\n\nAsthma is a major health concern in Sudan, with an estimated overall prevalence of 8.7% of the population, more than double the global prevalence of 4.3%11. Excluding maternity and deliveries, asthma is ranked as the third most common cause of hospitalisation in the country, following pneumonia and malaria12. There was a striking rise in the number of emergency room visits by asthmatics between 1998 and 2004, increasing from 20,000 to 106,0003. While asthma is a concern for the Ministry of Health in Sudan, it is not listed as a public health priority. In the public health sector there are very few peak flow meters (which measure a person’s ability to breathe out air) for diagnosis, few drugs available, and few staff trained in asthma management12. Many health professionals working in both the public and private sectors are not trained in tackling chronic asthma or long-term management of the disease5. This combination of factors leads asthma patients to seek care in the private sector. Research on asthma management in Sudan, conducted in 2003, found that 95% of patients paid the full cost of their asthma medicines; less than 2% of them received regular treatment from a single facility, and patients typically had no knowledge of their asthma management plan3. In a country where the daily wage of the lowest paid, unskilled government worker is USD $2.20 per day, the cost of one day of hospitalisation for asthma is USD $79.60, with patients also responsible for the cost of medicines and other additional costs3. These high costs of healthcare and medication often result in asthma patients self-managing their condition and only seeking care when they are extremely ill.\n\nPartly as a result of low confidence in the public health sector in Sudan, a pluralistic health system has grown and healthcare is provided by both the public and private sectors13. The government levels of funding for the public health sector is low at 6% of the GDP (gross domestic product), well below the 9% recommended by the World Health Organization (WHO), with high out-of-pocket expenditure by individuals, estimated to be above 60% of monthly salary10. The private sector includes ‘not for profit’ and ‘market based/for profit’ health-care facilities, and has expanded rapidly in the last 10 years14.\n\nAsthma is a disease that is both acute (life threatening attacks) and chronic, requiring long-term care. The health-care system is accessed at multiple levels and for different purposes; emergency care for acute attacks, while long-term management of the disease often occurs via outpatient clinics. Understanding the supply and demand factors of the health-care system and the health seeking behaviours of patients could lead to improvements in service delivery, and ultimately, better health outcomes. Figure 1 presents a framework used to guide this research and builds on social ecological theory15. It outlines the relationship between multiple levels of influence that exists between individuals, households, health institutions, the wider sociocultural environment, and the subsequent influence these have on decision-making and on understanding health-care utilisation behaviours. This research looks at the interaction between these levels of influence: individual, interpersonal, societal, organisational, and policy.\n\nAdapted from Shahabuddin et al.15 under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).\n\n\nMethods\n\nA two-phase approach was developed that would combine an identification of systemic factors driving private health-care choice and a mapping of the process involved in healthcare seeking. A quantitative health-care facility questionnaire took place from April 2014 to December 2014 and collected information on asthma services available from a large number of private facilities (breadth). Qualitative, in-depth interviews which took place from September 2014 and March 2015 explored in more detail the process and rationale of health-care decision-making of adult asthma patients (depth), with a focus on their individual decision-making regarding public or private care, and their complicated healthcare seeking process.\n\nThe Ministry of Health in Sudan provided a list of all the private facilities in Khartoum city that were listed as registered in 2014 (year of the data collection) and which facilities offered asthma services. Of the registered private health facilities that provided asthma services in 2014, there were 21 private hospitals, 9 private chest clinics, and 444 private pharmacies.\n\nPharmacy providers were stratified according to location and randomised16. The sample size for the survey was determined with a 90% confidence interval rate, and a precision estimate of +/- 15%. A random sample of 10% from each sub-district was chosen for the survey. There were 444 registered pharmacies split across three sub-districts of Khartoum (302 in Khartoum sub-district, 78 in Alshodha, and 64 in Khartoum East). Each of the three sub-districts were randomised separately using Excel to account for the difference in number of facilities. A random sample of 10% from each sub-district was chosen for the survey. Table 1 shows the number of health facilities included in the study (n=74).\n\nA facility-based cross-sectional survey of private providers in Khartoum (both health-care and drug providers) was conducted to collect information on their current level of asthma treatment provision. A structured questionnaire, developed based on WHO Health Facility Survey Guidelines17 and Global Initiative for Asthma (GINA) Guidelines9, was pre-tested. The questionnaire was tested through a pilot sample encompassing each of the facility types in the nearby city of Omdurman. Omdurman was identified as the best fit for the pilot because it is the next geographic city to Khartoum, and shares many of its characteristics. It has a high number of private facilities, a weak public health service, and it is a densely populated urban area. As a result of the pretesting, areas for improvement were identified including the need to make the survey specific to the level of facility (i.e. one survey for the hospitals, one for the clinics, and one for the pharmacies).\n\nIn order to answer the questions on access to asthma treatment, the survey focused on the following categories: Availability: levels of trained asthma health workers, asthma drugs, registers, guidelines and asthma diagnostic equipment; Accessibility: Understand the rate and type of use of services by asthma patients; Acceptability: Provider perspectives of the health- seeking behaviour of patients who use the private sector\n\nThe research team used smartphones, which stored the questionnaires, to input responses given by the health-care worker. The use of electronic questionnaires has been shown to improve collection time and accuracy, when compared to paper questionnaires18.\n\nThe private health providers were visited on an individual basis at their health facility (private hospital, clinic or pharmacy). At each health facility, the Senior Manager was met and asked for permission to conduct the research. They were then asked for the names of the health workers who fulfilled the inclusion criteria. One health care worker at each participating health facility and pharmacy was surveyed. The inclusion criteria for the survey: the provider provided care to adult asthma patients; they consented to be surveyed; they had regular contact with asthma patients and they provided an asthma diagnostic or treatment service for a fee.\n\nThe second phase of the research was based on models of healthcare seeking qualitative research19,20 and built on the findings of the Phase 1 provider survey. Exploratory qualitative research with asthma patients who sought care in the private sector in Khartoum was conducted using in-depth interviews.\n\nIdentification of participants and data collection in the qualitative component. During the health facility surveys, 14 asthma patients were identified and agreed that they would be willing to be interviewed. Table 2 shows the number patients interviewed, disaggregated by age and gender. The providers included in the Phase 1 quantitative survey assisted with identifying patients who attend their centre for the Phase 2 qualitative interviews. They were also informed that the interviews would be anonymised so that it would not be possible to tell which patients were talking about which health facility. The providers were briefed on the inclusion criteria and were asked to try and select a cross-section of patients according to age and sex.\n\nThe inclusion criteria for the in-depth interviews was as follows: They were aged 18 years and above, have an asthma diagnosis, live in Khartoum state, had sought asthma care at a private facility (hospital, clinic or pharmacy) and consented to be interviewed. It was aimed to try and interview a balance of ages and an equal mix of male/ females. The purposive sampling frame set out that four patients would be selected from each of the three provider categories (hospital, clinic, pharmacy) with an even split of males and females.\n\nHowever, this method of selection had the potential to lead to bias (such as people who are expected to give positive responses being selected by the gatekeepers), but this was felt to be the most feasible and convenient sampling method.\n\nIn order to avoid coercion, information about the study was given out by providers but patients contacted the researcher not the provider to express interest in the study, to ask questions about it and consider participation. They signed a consent form to demonstrate they were happy to participate. The patients chose whether they would prefer to be interviewed in their home, at the Research Centre or at the health facility. The interviews were recorded using dictaphones and then transcribed in Arabic by the data collection team. These transcriptions were translated into English and checked back into Arabic for consistency and accuracy of translation.\n\nThe interviews took 40 minutes and focused on individual healthcare seeking behaviour for their asthma, experiences of seeking care in the private sector, and their opinions about how their family viewed their condition, as well as the wider community thoughts regarding asthma. Due to time limitations, it was not possible to interview community or family members about their opinions on asthma care and private sector provision of care.\n\nEach individual was given a unique identifier number, comprised of a reference number, age, and gender. Each quote below is followed by a code that stands for gender (M/F) and the interviewee’s age. No two people were the same age and gender. The interviews were recorded with the patient’s consent. The interviews were transcribed by the research team in Arabic. These transcriptions were then translated by professional translators and then a sample were back translated to check for translation quality. The research team consisted of three experienced Sudanese qualitative researchers (two women and one man). Only one researcher was present at the interview. A topic guide was used to prompt and guide the interview: a copy of this is included in the extended data section.\n\nData analysis of the qualitative interviews. A three-step thematic analysis was conducted on the qualitative data21. Key themes were identified through reading of the transcripts. Interview transcripts were organised using NVIVO 11 software and linked to patient demographics. Concepts and labels were applied using a cross-sectional ‘code and retrieve’ approach22. This gave a systematic overview of the scope of the data and enabled us to make comparisons and connections.\n\nData were anonymised and stored securely throughout. Safeguarding of participants was prioritised throughout study design and implementation.\n\nEthical approval for the study was obtained from the National Health Research Ethics Committee (NHREC reference 1/7/14) in Khartoum, Sudan and the Liverpool School of Tropical Medicine Ethics Committee, United Kingdom (reference: LSTM REC 13.33). All participants gave written informed consent to participate in the questionnaire and interviews.\n\nThe participant information sheet and consent form both included statements about the possibility of publication of anonymised interview content. All participants gave their consent for the publication of material obtained in the interviews by signing the consent form.\n\nPatient/public involvement: Health-care providers, policy makers, and asthma patients were consulted in the development of the research and provided feedback on the findings of the pilot stage. Specifically, they provided input into the types of questions that were best suited to the quantitative survey, and which areas would fit best in the qualitative interviews.\n\nAsthma patients gave their full written consent to be interviewed, as well as commenting on the findings and dissemination plan. The research was disseminated at a public meeting in Khartoum in April 2018, which was attended by asthma patients, policy makers, academics, and health-care providers.\n\n\nResults: Applying the ecological framework to the findings\n\nTable 4 and Table 5 show the results of the quantitative survey and in-depth interviews in more detail.\n\nGovernment commitment to asthma as a priority health condition is highlighted in international guidelines as a key factor for improving health outcomes3. There is now a high level of international political support for an increased focus on asthma23. This support needs to filter down to the regional and country level, and lead to countries prioritizing asthma control through improved access to essential drugs, technologies, and well-trained medical personnel. Sudan does have a national asthma strategy. However, this focuses on the public sector and has only been loosely applied in practice24. This is consistent with other countries, where there has been limited adoption of international guidelines in resource-poor settings25. Implementing standard case management and strengthening health-care systems at all levels should be advocated for by the different stakeholders involved in asthma care26.\n\nStrengthening all components of the health-care system (public and private) is key to overcoming the barriers to effective asthma management and control. Good asthma management requires an uninterrupted supply of high-quality, affordable medicines, well-trained health professionals, effective diagnostic technologies, appropriate asthma guidelines, and registers to ensure accurate information recording and well-organized health-care services26. This study found that, in the private sector of Sudan, many of these essential requirements were lacking. Table 4 shows the survey findings in detail.\n\nDiagnostic equipment. According to the International guidelines, asthma diagnosis is conducted using a peak flow meter and clinical assessment9. Only 7 of the 21 (33%) hospitals reported having a peak flow meter and despite the large numbers of staff, 6 of these (29%) only had one device (with one hospital reporting owning 2 devices). The numbers were higher at the private clinics. Of the 9 clinics, 7 had peak flow meters (78%). Spirometry can be used for asthma diagnosis but it tends to be more helpful in understanding other lung health conditions that can often be disguised as asthma25. Few of the hospitals, 6 of the 21 (29%), had spirometers. Clinics were better equipped and 7 of the 9 (78%) clinics had spirometers. None of the pharmacies offered diagnostic services and so did not have peak flow meters or spirometers.\n\nThe private chest clinics surveyed were better equipped with asthma diagnostic equipment than the private hospitals. The low rates of spirometers and peak flow meters in hospitals are not unusual, and are comparable with other low-income countries27. The low rates found in the private sector were still higher than in the public sector, where there is very little asthma diagnostic equipment available12. The lack of definitive diagnosis can lead to over/under diagnosis of the condition and rely on the diagnostic skill of the health-care provider.\n\nAsthma registers. Asthma-specific registers and guidelines were in low supply in all of the facilities. Only one hospital and one clinic reported using an asthma specific register and none of the pharmacies kept asthma registers. Asthma guidelines were more widely held at half of the facilities. Very few of the facilities surveyed used specific asthma treatment registers and relied on the standard Ministry of Health outpatient registers to record asthma patients. International guidelines recommend the use of standardised asthma registers and the use of asthma treatment cards to assist with asthma management28. These were not found in the majority of the private sector facilities surveyed in Khartoum, which makes it difficult for health-care providers and policy makers to assess the number of people receiving asthma care.\n\nHealth-care provider training. A key component of asthma management guidelines is asthma-specific training for health-care personnel28. The health workers surveyed at the facilities all received training themselves on asthma as part of their general medical or pharmacy training, but only 4 of the 21 (19%) health workers at the hospitals and 3 of the 9 (33%) health providers at the clinics had undergone specific training on asthma since qualifying from medical school. One pharmacist reported asthma specific training which was carried out at Khartoum hospital. The hospital health providers’ training focused on asthma management in the emergency room and as an acute condition. The doctors at the clinics had a broader training on lung function tests and inhaler technique. Asthma is a complicated disease to diagnose and treat. Furthermore, it consists of both acute attacks and chronic symptoms. Therefore, specialist training is recommended to ensure patients are given the best support for managing their disease9.\n\nAsthma drugs. Patients highlighted that the availability of drugs was a major issue, particularly in the public sector. The cost of drugs was also viewed as a key barrier to managing their disease; both patients and providers cited that cost was the main reason for stopping treatment. When the public sector has a low availability of drugs, as seen in the public provision of asthma drugs in Khartoum, the use of the private sector increases29. A lack of legislated price control has led to wide variation in prices and was mentioned by patients as a motivator in where drugs were purchased. There are other considerable costs, such as consultation fees, hospital admissions, complementary treatments, and laboratory tests30.\n\nAll of the 21 hospitals and 3 of the 7 (33%) clinics said that they stocked asthma drugs. All the pharmacists reported stocking a range of asthma drugs. 34 of the 44 (77%) pharmacists reported stocking all 5 recommended asthma drugs (anti-inflammatory drugs, bronchodilators, beclomethasone, salbutamol, budesonide). There was a large variation in the cost of the drugs between pharmacies as listed in Table 3. This is mostly due to a difference in branded and generic versions of the drugs; branded, international brands have a far higher cost than Sudanese generic drugs. Average daily salary in Khartoum is 122 Sudanese pounds, the average price of the reliever inhalers was around 20% of a daily salary while the preventer inhalers were between 35% for non-branded types rising to over a day’s wage for the branded versions drugs31 as shown in Table 3.\n\nAsthma health education. Access to quality of health information regarding health-care providers and disease specifics has been shown to affect provider choice29. There was a distinct shortage of asthma leaflets or posters. The importance of education in inhaler technique was also highlighted as a key factor affecting patients’ treatment32,33. Increasing the role of pharmacists beyond asthma drug dispensing, to patient counselling and education, could act as a way of improving inhaler technique, thus reducing improper use and the accompanying occurrence of severe attacks32,34.\n\nCommunities play a key role in asthma management and shared beliefs often influence healthcare seeking behaviour. Stigma in communities was reported by patients and resulted in delayed healthcare seeking and case management. Stigma was particularly of great concern for younger women who reported that people treat them differently when they find out that they are asthmatic. People in the community feared that asthma is contagious or believed it to be disgusting. It was felt that it could affect marriage potential, as people do not want to marry someone with a long-term condition. Due to inhalers being perceived as addictive, people who use them can be viewed by the community as addicts. As with the individual concerns, these societal concerns were felt more strongly by younger patients and, in particular, by younger women.\n\nThe influence of community views of asthma was reported strongly and by many of the interviewees. People reported a low level of understanding in the community of the disease and how it can be managed successfully. They also highlighted little knowledge in the community as to the causes of the disease. Many of the patients described how people treat them different when they find out they have asthma.\n\n‘People are afraid of asthma because they think the patient will have an asthma attack if he gets into a small argument. But it is the opposite and he is normal like everyone else. – people are afraid of arguing with asthma patients’. (M21)\n\nThese commonly held beliefs were not always accepted by the patient themselves but they felt the beliefs were so strongly held in the community that there was little the individual could do to change these beliefs. It often led to a delay in diagnosis and a delay in treatment.\n\nThe disease was often viewed as contagious, and therefore changed how people interacted.\n\n‘there are some people who are scared of the disease they say it's contagious’ (F30)\n\nSome of the younger women also reported that because asthma is a long-term condition, people in the community perceive the disease to have a long-lasting impact on the patients’ futures.\n\n‘there are people frankly when they would approach you for marriage and know that you have asthma, they will be afraid and will say that she will always be taking me to doctors and even the children when they are born then might have asthma because of the weather….if someone wanted to marry you then if they know you have asthma, they will say that she has asthma and that she will take me to the doctor all the time so I better find someone who is healthy.’(F30)\n\nThe use of an inhaler in public was also described as something that people were cautious about, particularly younger men and women. One mother reported that her daughters were viewed negatively when they used their inhaler in public. While the mother knew the benefits of inhaler use, she highlighted that the broader community held stigmatised views of them.\n\n‘they will ask me why do I insist to say that my daughters have asthma and they refuse and they say that one of them will be addicted to the treatment. All of that is not correct and in contrary if one takes the medication then it will become less’ (F45)\n\nThere was the assumption that the community as a whole believed the best method of treatment was treating an acute attack at the hospital emergency room.\n\n‘Asthma is known, everyone knows it, as soon as I get an attack, they prepare the car and take me to the emergency room.’ (M60)\n\nThe stigmatisation of inhaler use is compounded by society’s view that the best method of treatment is at hospital emergency rooms. This fits with the low level of preventer inhaler use amongst the interviewees and their repeated visits to hospitals for acute care.\n\nMany of the patients highlighted strong family links with the disease. Yet, patients still preferred to refer to their condition as an allergy. They also described how common the disease is in Khartoum, and that it was very difficult to find someone who was not affected (despite the stigma surrounding the disease). This is consistent with the finding that asthma is the third most common reason for seeking hospital care in Sudan, and how over-burdened the health-care system is with asthma cases11.\n\nMany people reported that inherited factors played a key role in their asthma condition, and that this was the main causal factor of their condition. Genetics and family history were mentioned by over half of the patients:\n\n‘There is no problem in the family, we are in an asthma forest, my mother, grandmother and uncles have asthma. So this is a very normal thing in because it already exists in the family.’ (M48)\n\nIt led to a sense of feeling that there was little that could be done about the disease and that they should just accept living with it. The belief that asthma is hereditary is so strongly held that one young woman described the shock of being the only person in her family diagnosed with the disease.\n\n‘By god the people of my house were very shocked, to be frank with you, the shock for them was abnormal as there is no asthma in the family. We have not anyone having asthma in neither my father’s nor my mother’s side… Of course we were told it was asthma, then they said where has she got this illness from? At the end of course they can say nothing as it is from God almighty.’ (F36)\n\nFamily members were reported as key in the decision-making process about where to seek asthma treatment. Often, the decision was made entirely by the family if the patient was too ill and needed to be treated in an emergency situation. Involving patients’ families in health promotion activities could be a successful way to improve health education and knowledge about the disease. Further work that delves deeper into this and understands the relative roles different family members play is recommended.\n\nHealth beliefs affecting choice of provider. Most patients described knowing very little about asthma before their diagnosis, except that they believed it was caused by an allergy or is an inherited condition35,36. Refusing the diagnosis, and relying on the self-diagnosis of allergy, affected choice of provider and management of the disease, particularly for the younger women interviewed, and links with the increased levels of stigma in the community faced by these women.\n\nWhether asthma is viewed as separate acute episodes or a long term chronic condition seemed to influence how people sought treatment and how they perceived it affected their lives. While many respondents acknowledged how common the disease was in Sudan, they often preferred to refer to it as an allergy rather than a long-term condition. The use of the Arabic word ‘hassaseeiaa’ rather than ‘asthma’ was an indication about the reluctance to accept asthma as a diagnosis. There was a belief that an allergy was something that could be treated and was short term whereas asthma was a lifelong condition that had social and physical consequences. The patients described that there was a reluctance by health workers to discuss asthma and preferred to diagnose it as an allergic reaction.\n\n‘Even the health workers .. use the word ‘hassaseeiaa’ [allergy], they start stigmatizing, especially when you find you are a woman who is younger. They say you have a chest hassaseeiaa without using the word asthma….There are people who will hide that they have asthma, there are some who will be shocked and they will say it is allergy.’ (F18)\n\nTreating asthma as an allergy meant that many of the patients, especially female ones, only sought care when they had an attack. The stigma regarding an asthma diagnosis extended to stigma regarding the use of inhalers in public and this was highest for younger women. They reported that they would rather not use their inhaler when in public, and that if they did, it would bring unwanted attention:\n\n‘By God, frankly doctor when I got this disease and they gave me the three inhalers, I was very sensitive and I was not able to carry my inhaler with me. As we were going out and she [mother] was telling me to take my inhaler and I was very sensitive about this issue, so I told her that I can’t take the inhalers. Do you want me to puff in front of all the people? So I said I will never puff in front of people’. (F36)\n\nHigh levels of stigma regarding inhaler use can have a big impact of the lives of people living with the disease. Not using inhalers regularly mean that attacks are often more severe and more dangerous and need hospitalisation.\n\nPerceptions of care quality. Patients perceived a higher quality of care in the private sector. It was found that the public sector was viewed as overcrowded, with long waiting times, and sometimes low levels of hygiene. This was a major motivator for seeking care in the private sector. Service quality was also viewed by many individuals as much higher in the private sector.\n\nMost patients reported that the main difference between public and private care was that the public sector facilities were very overcrowded. They reported that they would often have to wait a long time for an oxygen machine as there were often two or three people waiting for each machine.\n\n‘The public hospitals are crowded. Eight or ten years ago I went to Alnaw, when I went I found they have one oxygen machine with Ventolin. I wait because there are two or three people in front of me. In that moment you will be annoyed but in the private hospital, they have a number of machines and beds.’ (M60)\n\nPublic hospitals were often described as not being clean enough and chaotic when compared with private facilities and this was a motivating factor for not seeking care in the public system. Location was also a key theme affecting the choice of provider. This was varied and depended on the severity of the attack. In a severe attack, the nearest facility was chosen.\n\n‘this [private] hospital is close to us and they will save me faster. The private hospitals have solved many problems especially for the asthma patient. The public hospitals are cheap, but there are many things that are not available.’ (F18)\n\nPatients also expressed that they would choose a health provider that was known to them and that the private sector was always their first choice:\n\nHealthcare seeking behaviour. The main reason patients reported seeking care was due to extreme illness. The severity of the disease, and that it often causes death, was a source of great concern for the patients. Many reported that they knew someone who had died of the disease and that they themselves had become unconscious from an acute attack. These attacks meant that short-term, fast acting treatment was sought, and so the patients went directly to the hospital emergency room (often repeatedly). Severity was a theme that came out strongly and was consistent with the health facility survey that highlighted many people sought care when their asthma was urgent and life threatening.\n\n‘Asthma puts you in a bad psychological state, you confront death, this is the only disease that makes you feel that you are about to die. And when you are treated you will be in a very bad psychological state because the asthma patient is always stressed, whether he has an attack or not. He can get an attack at any time. This is why I think the asthma patient is always at the edge of his life…’. (M48)\n\n‘I mean seriously this thing was a big obsession for me; it is a fatal disease and should not be taken lightly ….Like I gave up on life, I will not lie to you, it is like you find out that you will die soon, it is in God's hands’ (F30) (after being asked how she felt when she received her diagnosis)\n\nThe association between asthma and death is strongly held by a broad range of age groups and across both genders. This affected choice of provider as care was sought urgently in an emergency setting rather than as routine management of the condition. In Sudan, these patients all described their fear of asthma related death and how they focus on treating their attacks as a series of acute episodes rather as a chronic condition. Accessing care and provider choice for acute attacks can be quite different to long term management of a condition. The majority of people sought care because they were extremely ill, and the decision as to which facility to seek care in was not made by the patient, but by someone else (a family member or neighbour).\n\n‘they lifted me up as I was completely comatose, and I was not breathing at all so my family supported me and they started shouting and the neighbours came and helped with lifting me up, our neighbour took me in a rush to the hospital’ (F36)\n\nThe impact asthma had on the patient’s daily life was strongly articulated, with most people describing high levels of worry and concern and how it disrupts their plans and is an obstacle to carrying out a normal life.\n\n‘My feeling was extreme anger because he [the doctor] forbade me from playing football and grasshoppers and dust and such. I used to work with my uncle during the holidays as a building labourer, he forbade me from all of that, I just sit in the house, bored’ (M29)\n\n‘Sometimes I would find myself collapsing because I got tired but I can only thank God. I have stopped from my studies. This illness needs follow-up and to know how to know the use of the medication’ (F30)\n\n“I am scared and worried when I have an infection, I will not be awake, not among the living and not among the dead’ (F36)\n\nThe younger age group (both males and females) described higher levels of concern about having the disease and that it majorly impacted on their daily lives causing them to stop doing certain activities.\n\nHealthcare utilisation. A strong finding was the frequency with which patients sought care for their asthma and that the hospital emergency room was the main place where they sought care. An acute attack altered choice of provider as the nearest accessible facility was chosen, whether it was private or public. Self-referral to upper level heath care facilities is often a result of a perception that primary care facilities offer low-quality care and a lack of available drugs or personnel37.\n\nAccording to the International Asthma guidelines12, once an asthma diagnosis is made, a patient should receive an asthma management plan. This guides the patient on how to take their medicine, which inhalers to use and how and where to seek care. These plans have been shown to lower the need for emergency care and lessen the number of severe, life-threatening attacks12. None of the patients interviewed had such a plan and all described very complex journeys to seek care. Most of the patients only sought care when they were experiencing an attack. This care ranged from using a home oxygen kit to having to go to the hospital emergency room for treatment. For some of the patients, the journey meant that they had to seek care outside of Sudan. Repeated visits affect the patient’s ability to live a full life and also come with high levels of financial costs. This woman describes going to the hospital to receive oxygen five times a month and has to pay each time.\n\n‘There is supposed to be discounts as the attack can happen to the patient about two times in a week. And the oxygen room is for 40 pounds, in a month if I got sick five time, it is too expensive. For me it is a continuous asthma and any minor thing can affect me, dust or heavy wind and also if I made a great effort or I got mad because of a situation then I will not be able to breathe’(F30).\n\nIn the majority of asthma cases worldwide, the use of preventer inhalers substantially reduces the number of hospital visits9 and reduces severity of attacks. Most people only described the use of oxygen and reliever inhalers such as Ventolin. Only a few people reported ever having used a preventer inhaler but those that did felt that they were very beneficial.\n\n‘Before I came here I used to suffer a lot, each week I stay two or three days at home. I told him [Doctor] my history with asthma and he gave me the preventer, since then I feel a lot better. I am a completely new person after I started taking the preventer’. (M48)\n\nThe cost of both types of inhaler was highlighted as a major reason that people did not use inhalers. While for some people the high costs meant that they would not use the inhaler, others reported that they would prioritise affording it over spending money in other areas as demonstrated by the quote below:\n\n‘By God it is not good, but thanks God what can we do about it, everything is becoming very expensive and this [inhaler] is the only thing that has to be always in my hands and I can stop eating or drinking, but the inhaler has to be in my hands’ (F36)\n\nThere were a lot of costs reported in addition to the price of the inhalers, such as the use of the oxygen room, injections, hydrocortisone, doctor consultation fees, and these were considered as barriers to sustaining effective asthma management.\n\nAnother person reported that he often saw a person begging on the street outside an asthma hospital asking for inhalers, and sometimes he would try and help by buying an extra one.\n\n‘I saw people who beg for the inhalers in Hawadith street, he [one man] brings the receipts and tells you that he wants an inhaler because his inhalers are finished. Sometimes I buy it for him’…(M60)\n\nMost patients self-managed their condition and sought care when they felt they needed it. Self-medication was viewed as a low-cost alternative to health-care facilities, which charged consultation and laboratory fees. Those interviewed had no treatment plan or guidance on the steps to take to avoid asthma attacks. However, they felt they knew what to do to reduce the frequency of attacks, and on the whole, avoided going to see doctors for advice. This study did not find strong gender differences with both men and women describing managing their asthma themselves. Age did seem to be a factor with more people mentioning self-medicating who were over 30 years old and only a few under 30 years.\n\n‘It has been years since I saw a doctor…I just use my inhaler. I control my asthma and avoid infection and flu. Last time I saw a doctor was in 2000/2001’ (M60)\n\nThe use of inhalers to treat an attack was accepted, but preventative inhalers were used much less often. Pharmacists were the main point of contact for those who were self-medicating, and it is proposed that they play a key role in providing advice to consumers on the safe use of drugs and on inhaler technique34,38.\n\n\nDiscussion\n\nUsing a social ecological approach has offered an understanding of the different levels that influence asthma service delivery in the private sector in Sudan, as well as how and why asthma patients utilise these services. The phased mixed-method approach gathered complementary data to provide a fuller understanding of if, where, why, and how asthma patients seek care. The findings described above highlight that choice of provider and management of disease is more than supply and demand; decision-making involves multiple levels of influence and different perspectives that incorporate severity and health-care consequences, and that these factors all interact to affect health outcomes.\n\nAdopting a social ecological approach to examine asthma healthcare seeking behaviour and provider choice facilitates understanding regarding which factors influence decision-making by asthma patients in Khartoum. Decisions regarding asthma treatment are influenced by age, severity of disease, and perceived quality of services.\n\nA key part of successful asthma management requires patients to be empowered so that they have a good understanding of their condition, a willingness to accept the diagnosis, and a supportive social environment, thus giving them the ability to treat the condition between attacks. Health-care choices in Khartoum are currently made on an attack-by-attack basis, and short-term, acute care is sought, rather than long-term management of the disease. The private sector, while viewed as expensive, is also viewed as having a higher quality of care, with treatment being delivered faster, which is vital in the case of an acute attack.\n\nOrganisational and health-care system issues, such as a lack of available and affordable drugs need to be addressed alongside increasing demand for services and perceived quality of care administered by different providers. Asthma programmes need to promote the benefits of long-term management of the condition. Specifically, how it will reduce the frequency and severity of attacks, reduce hospital admissions, and should lead to individuals living a more normal life. Implementing the International Guidelines for Asthma Care in Khartoum in the private sector would be very beneficial addressing the different levels, including political commitment, drug availability, training, and diagnostics. Pilot studies in Sudan have shown that implementation of standard case management systems in the public sector can be very beneficial in reducing hospital visits, and have recommended a more extensive rollout of these systems12.\n\nAsthma is a severe disease that acutely and chronically affects many people in Sudan and the accompanying region. It places a large burden on a weak healthcare system and causes severe distress for those suffering from the disease. Previously, very little was known about the health seeking behaviour and service provider choice for asthma patients. This paper highlights that the pathway to care is complex and there are multiple influences on decision making. Young people, especially young women, face enormous stigma from having the disease, affecting their choice of provider and treatment options, especially with regard to a reluctance to use preventative inhalers\n\nThe current practice of seeking care at hospital emergency rooms is a result of the severity of the asthma attacks and a lack of knowledge of alternative options for care, leading to an overburdening of emergency rooms and lack of trust in outpatient facilities.\n\nDue to time and funding restrictions, this study did not interview community members with regard to their attitudes to asthma patients. This could have given more insight into the reasons behind the stigma felt by the patients.\n\n\nConclusion\n\nEncouraging private sector providers to implement standard case management should lead to a reduction in emergency room admissions and less severe attacks, as well as reduced asthma related fear and concern on the part of patients. The reasons behind patients’ choice of healthcare need to be addressed alongside the provision of appropriate services to improve outcomes for asthma patients. Targeting all layers of influence on healthcare seeking and provider choice is the most likely to create sustainable health improvements.\n\n\nData availability\n\nDue to the level of detail in the qualitative transcripts from this study (stage 2), even with direct personal identifiers removed, the stories and descriptions of participants within the transcripts makes them potentially identifiable, as well as containing sensitive and confidential patient information. We do not have ethical approval for the sharing of raw datasets from our institutional ethics committee and we have therefore deemed these raw data unsuitable for public sharing. We have attempted to share as much data as possible through quotes within the text. Data requests may be sent to Julie Irving (Julie.irving@lstmed.ac.uk), who is an institutional representative of the Liverpool School of Tropical Medicine and will hold a copy of the data and respond to external access requests. Access will be given to researchers with ethical approval from their institutions who are conducting work on this topic. To ensure long-term data storage and availability, they will be held by at least two of the authors, in addition to Ms Julie Irving.\n\nOpen Science Framework: Applying an ecological framework to examine the multiple levels of influence affecting the utilisation of private sector adult asthma services in Khartoum: A mixed methods study, https://doi.org/10.17605/OSF.IO/QRAVP39. Registered on 28 September 2020 (osf.io/v3j7x).\n\nThis project contains the following underlying data:\n\nData.xlsx (Dataset containing results for all participants of the health facility survey (hospitals, clinics, pharmacies)\n\nOpen Science Framework: Applying an ecological framework to examine the multiple levels of influence affecting the utilisation of private sector adult asthma services in Khartoum: A mixed methods study, https://doi.org/10.17605/OSF.IO/QRAVP39. Registered on 28 September 2020 (osf.io/v3j7x).\n\nThis project contains the following extended data:\n\n- Questionniares.docx (Facility structured survey; quantitative study 1)\n\n- Topic guide.docx (Interview guide; qualitative study 2)\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).",
"appendix": "Acknowledgements\n\nThis study has been made possible by the team at Epidemiological Laboratory for Public Health and Research (EPI LAB), Khartoum. We are extremely grateful for their collaboration and support throughout. We would also like to express sincere thanks to all the participants of the study, who kindly gave up their time to be interviewed and who provided such helpful and informative insights into their lived experiences of accessing healthcare in Sudan.\n\nThis study was part of the PhD thesis ‘Multiple levels of influence affecting the utilisation of adult asthma services in the private sector in Khartoum’ written by RT. The full thesis is available here: https://doi.org/10.17635/lancaster/thesis/538\n\n\nReferences\n\nBousquet J, Ndiaye M, Aï t-Khaled N, et al.: Management of chronic respiratory and allergic diseases in developing countries. Focus on sub-Saharan Africa. Allergy. 2003; 58(4): 265–83. PubMed Abstract | Publisher Full Text\n\nAdeloye D, Chan KY, Rudan I, et al.: An estimate of asthma prevalence in Africa: a systematic analysis. Croat Med J. 2013; 54(6): 519–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIUATLD: The Global Asthma Report. 2011. Reference Source\n\nLessing C, Mace C, Bissell K: The availability, pricing and affordability of three essential asthma medicines in 52 low- and middle-income countries. PharmacoEconomics. 2013; 31(11): 1063–82. PubMed Abstract | Publisher Full Text\n\nAit-Khaled N, Enarson DA, Bissell K, et al.: Access to inhaled corticosteroids is key to improving quality of care for asthma in developing countrie. Allergy. 2007; 62(3): 230–6. PubMed Abstract | Publisher Full Text\n\nAit-Khaled N, Auregan G, Bencharif N, et al.: Affordability of inhaled corticosteroids as a potential barrier to treatment of asthma in some developing countries. Int J Tuberc Lung Dis. 2000; 4(3): 268–71. PubMed Abstract\n\nAï t-Khaled N, Enarson D, Bousquet J: Chronic respiratory diseases in developing countries: the burden and strategies for prevention and management. Bull of the World Health Organ. 2001; 79(10): 971–9. PubMed Abstract | Free Full Text\n\nBilal M, Haseeb A, Khan MH, et al.: Factors associated with patient visits to the emergency department for asthma therapy in Pakistan. Asia Pac Fam Med. 2016; 15(1): 1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBeasley R: The Global Burden of Asthma Report, Global Initiative for Asthma (GINA). 2011.\n\nWHO E: Sudan Country profile: WHO Emro. 2011. Reference Source\n\nWHO: Health System Profile. World Health Organisation; 2007.\n\nEl Sony A, Chiang C, Malik E, et al.: Standard case management of asthma in Sudan: a pilot project. Public Health Action. 2013; 3(3): 247–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPantuliano S, Buchanan-Smith M, Metcalfe V, et al.: hpg. 2011. Reference Source\n\nEltilib H, Hameed N, Munim A, et al.: Management of TB in the private sector in Khartoum, Sudan: quality and impact on TB control. Red. 2010; 2: 0.9. Publisher Full Text\n\nShahabuddin A, Nostlinger C, Delvaux T, et al.: Exploring Maternal Health Care-Seeking Behavior of Married Adolescent Girls in Bangladesh: A Social-Ecological Approach. PLoS One. 2017; 12(1): e0169109. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHayes RJ, Moulton LH: Cluster randomised trials. CRC press London; 2009. Publisher Full Text\n\nWHO: Health Facility Survey Tool. World Health Organisation; 2003. Reference Source\n\nAhmed R, Robinson R, Elsony A, et al.: A comparison of smartphone and paper data-collection tools in the Burden of Obstructive Lung Disease (BOLD) study in Gezira state, Sudan. PLoS One. 2018; 13(3): e0193917. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGood B: Explanatory models and care-seeking: a critical account. Illness Behavior: Springer; 1986; 161–72. Publisher Full Text\n\nLavy V, Germain JM, Mundial B: Quality and cost in health care choice in developing countries. World Bank Washington^ eDC DC; 1994. Reference Source\n\nBoyatzis RE: Thematic analysis and code development: Tranforming qualitative information. London and New Dehli: Sage Publications; 1998. Reference Source\n\nRitchie J, Lewis J, Nicholls CM, et al.: Qualitative research practice: A guide for social science students and researchers. Sage; 2013. Reference Source\n\nPearce N, Asher I, Billo N, et al.: Asthma in the global NCD agenda: a neglected epidemic. Lancet Respir Med. 2013; 1(2): 96–8. PubMed Abstract | Publisher Full Text\n\nAsher I, Haahtela T, Selroos O, et al.: Global Asthma Network survey suggests more national asthma strategies could reduce burden of asthma. Allergol Immunopathol (Madr). 2017; 45(2): 105–14. PubMed Abstract | Publisher Full Text\n\nAhmed R, Robinson R, Mortimer K: The epidemiology of noncommunicable respiratory disease in sub-Saharan Africa, the Middle East, and North Africa. Malawi Med J. 2017; 29(2): 203–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAsher MI, Ellwood P: The global asthma report 2014. 2014. Reference Source\n\nDesalu OO, Onyedum CC, Iseh KR, et al.: Asthma in Nigeria: are the facilities and resources available to support internationally endorsed standards of care? Health policy. 2011; 99(3): 250–4. PubMed Abstract | Publisher Full Text\n\nAit-Khaled N, Enarson DA, Bencharif N, et al.: Implementation of asthma guidelines in health centres of several developing countries. Int J Tuberc Lung Dis. 2006; 10(1): 104–9. PubMed Abstract\n\nMorgan R, Ensor T, Waters H: Performance of private sector health care: implications for universal health coverage. Lancet. 2016; 388(10044): 606–12. PubMed Abstract | Publisher Full Text\n\nNunes C, Pereira AM, Morais-Almeida M: Asthma costs and social impact. Asthma Res and Prac. 2017; 3(1): 1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNumbeo: Cost of Living, Khartoum. Reference Source\n\nAbdelhamid E, Awad A, Gismallah A: Evaluation of a hospital pharmacy-based pharmaceutical care services for asthma patients. Pharm Pract (Granada). 2008; 6(1): 25–32. PubMed Abstract | Publisher Full Text | Free Full Text\n\nImad H, Yasir G: Epidemiological and clinical characteristics, spirometric parameters and response to budesonide/formoterol in patients attending an asthma clinic: an experience in a developing country. Pan Afri Med j. 2015; 21; 154. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOsman A, Hassan ISA, Ibrahim MIM: Are Sudanese community pharmacists capable to prescribe and demonstrate asthma inhaler devices to patrons? A mystery patient study. Pharm Pract (Granada). 2012; 10(2): 110–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMerghani TH, Zaki AM, Ahmed AM, et al.: Knowledge, attitude and behaviour of asthmatic patients regarding asthma in urban areas in Khartoum State, Sudan. Khartoum Medical Journal. 2012; 4(1). Reference Source\n\nPrasad R, Gupta R, Verma S: A study on perception of patients about bronchial asthma. Indian J Allergy Asthma Immunol. 2003; 17(2): 85–7.\n\nHabtom GK, Ruys P: The choice of a health care provider in Eritrea. Health Policy. 2007; 80(1): 202–17. PubMed Abstract | Publisher Full Text\n\nAwad AI, Eltayeb IB, Capps PA: Self-medication practices in Khartoum State, Sudan. Eur J Clin Pharmacol. 2006; 62(4): 317–24. PubMed Abstract | Publisher Full Text\n\nThomson R: Applying an ecological framework to examine the multiple levels of influence affecting the utilisation of private sector adult asthma services in Khartoum: A mixed methods study. 2020. http://www.doi.org/10.17605/OSF.IO/QRAVP"
}
|
[
{
"id": "72836",
"date": "28 Oct 2020",
"name": "Robin J. Green",
"expertise": [
"Reviewer Expertise Paediatric Pulmonology"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe basic error is no definition of asthma in the patients is given. What was there diagnosis based on? What severity of asthma and what medication were they using?\nIn addition it seems improbable to extrapolate conclusive findings from 14 patients.\nI would not suggest publication in this journal as the article is not methodologically sound\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": [
{
"c_id": "6270",
"date": "28 Jan 2021",
"name": "rachael thomson",
"role": "Author Response",
"response": "Thank you very much for taking the time to review this article. We appreciate your comments. It would be helpful to understand a little more about your concern on the number of patients interviewed. This is the qualitative part of the study and 14 patients is a very standard number for qualitative research. The number chosen was based on whether saturation of results was found. The analysis for the qualitative component was using the thematic framework and the coding frameworks could be used by other qualitative researchers if they wanted to replicate the work. The quantitative facility survey was much larger and the survey is available and therefore could be replicated if wanted. The choice of a mixed methods study was so that the quantitative facility survey could describe what is available in terms of services while the qualitative component aimed to gain an in-depth understanding of the motivation of seeking care and how living with asthma affected them. I hope this has clarified things a bit more for you."
}
]
},
{
"id": "75886",
"date": "08 Jan 2021",
"name": "Elopy N. Sibanda",
"expertise": [
"Reviewer Expertise Asthma",
"Allergy and Clinical Immunology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe research question addressed by the authors and its relevance to the management of patients with asthma particularly in low resource settings as exemplified by The Sudan is clear.\nThe situational context of the study is clearly articulated in the introduction.\nThe mixed methodology that brings together quantitative and qualitative aspects has its merits, however it may cloud the message. Each aspect, is important however the qualitative and quantitative analyses can down play the overall findings.\nThe authors address an important issue of the occurrence and management of asthma in the Sudan.\n\nQuestions:\nThe criteria used to include patients in the study are not clear that this doctor diagnosed or patient perceived. This is an important question because there are many other respiratory conditions that closely mimick but are not asthma such is fibrotic lung diseases. Could the authors clarify their inclusion criteriae.\n\nThe mixing of quantitative and qualitative analysis in one report although previously done, tends to belabor the points raised. There is repetitiveness of many statements, in some cases verbatim.\n\nTables 4 and 5 contain information that is extracted verbatim from the text. This makes the information repetitive.\n\nThe paragraphs \"Health Policy level\" and \"Organisational and Health Systems\" under Results do not appear to report the findings of the authors, these come out as recommendations and suggestions.\n\nThe authors could consider simple tables to capture the information where applicable, eg. under the paragraph diagnostic equipment, asthma drugs and health care provider training.\n\nStrengths and limitations does not appear to address the strengths or limitations of the work being reported.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6271",
"date": "28 Jan 2021",
"name": "rachael thomson",
"role": "Author Response",
"response": "Thank you very much for taking time to review this article. Your comments are very appreciated and it is good that you see the importance of this issue. We will go through and revise the article in line with your comments. We will aim to make the inclusion criteria much clearer. Apologies for any repetition in quotes, we will address this. We will also look at the text in the results section around health policy and organisations factors. Putting some of the data from the facility survey into a table is very helpful and we will also review the strengths and limitation section."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1227
|
https://f1000research.com/articles/4-552/v1
|
12 Aug 15
|
{
"type": "Review",
"title": "The Immune System and Responses to Cancer: Coordinated Evolution",
"authors": [
"Brendon J. Coventry",
"Martin Ashdown",
"Maciej Henneberg",
"Paul C W Davies",
"Martin Ashdown",
"Maciej Henneberg",
"Paul C W Davies"
],
"abstract": "This review explores the evolutionary interaction and co-development between immune system and somatic evolution. Over immense durations, continuous interactions between microbes, aberrant somatic cells, including malignant cells, and the immune system have successively shaped the evolutionary development of the immune system, somatic cells and microorganisms through continuous adaptive symbiotic processes of progressive immunological and somatic change providing what we observe today. The immune system is powerful enough to remove cancer and induce long-term cures. Our knowledge of how this occurs is just emerging. It is less clear why the immune system would detect cancer cells, when it is usually focused on combatting infection. Here we show the connections between immunity, infection and cancer, by searching back in time hundreds of millions of years and more to when multi-cellular organisms first began, and the immune system eventually evolved into the truly brilliant and efficient protective mechanism, the importance of which we are just beginning to now understand. What we do know is that comprehending these points will likely lead to more effective cancer therapies.",
"keywords": [
"Immune system",
"evolution",
"cancer",
"mutation",
"immune response",
"immunosurveillance",
"immunotherapy"
],
"content": "Introduction and overview\n\nIt often goes unappreciated that the adaptive immune system developed hundreds of millions of years ago, and has evolved into a truly efficient protective mechanism, the importance of which we are just beginning to now understand in science and medicine. Acute immune responses have developed alongside infection and genetic diversity, as part of the entire evolutionary process of matching organism against organism. There has been a continuous development of the immune system's capacity to protect an organism against infections through rapid genetic somatic hypermutations that also lead to a dynamic, intricate interplay between genetic endowment and somatic mutations. The immune system acts as an ultimate high fidelity 'read-out' for cellular genetic change, detecting cellular aberration at very early stages as it develops, to remove or destroy aberrant cells. Such aberration arises from infection of cells by viruses, bacteria or other microbes, DNA damage, failed repair mechanisms, mutagens, and carcinogens including UV light, toxins and chemicals, and cellular ageing. Constant dynamic interaction occurs between cells and the immune system to preserve homeostasis. Because the rate of mutation during cell division and tissue turnover far exceeds the rate of malignant tumour diagnosis, the immune system must play an efficient role in detecting and eliminating aberrant and frankly malignant cells at a developmentally early stage. The reason why cancer occurs at all in humans and animals thus remains a mystery. However, the answer likely resides with observations that in the chronic state of antigen persistence the immune system continually appears to close itself down to avoid over-activation and to conserve energy. In the acute state, with exposure to each new pathogen the immune system responds rapidly over several days and then typically retains 'memory' of that encounter, enabling more rapid responses upon subsequent exposure. If the antigen can be acutely removed from the system, the immune system returns to steady basal state via homeostatic mechanisms. However, if the antigen persists and cannot be removed from the organism, the immune system responds again with a further cycle of activity. Over-reactivity is limited by eliciting an inhibitory response after each activation response, in the form of negative feedback for biological homeostatic damping. This cyclic feedback phenomenon is seen right across many, if not all, biological systems in nature. In the chronic state, the immune system repeatedly activates in response to persistent antigenic signals. When the antigenic signal cannot be removed with a second 'round' of activation, another cycle of activation and then inhibition occurs. This repetitive cycle continues until the antigenic focus is eventually removed, or the organism dies. Although this is an efficient system in the acute setting, in the chronic setting where the problem persists indefinitely and does not appear to resolve, 'chronic inflammation' can arise which is often far less energy efficient. Vast amounts of energy can be consumed in chronic severe inflammatory conditions. In areas of the world where infection has been effectively reduced by sanitation and other public health measures, chronic inflammatory diseases have emerged as the major causes of morbidity and mortality. Clinically, this manifests as a relapsing and remitting process, often with malaise and weight loss characteristic of many chronic illnesses. This rather maladaptive process consumes massive amounts of energy, damaging surrounding tissues and cells. Over a number of generations, natural selection can lead to efficiency improvements of immune responses to specific chronic infections through co-adaptation of hosts and pathogens; examples are endemic treponematoses1,2 or tuberculosis3,4.\n\nThis article considers cancer immunology in terms of immune system evolution and chronic inflammation.\n\n\nReview\n\nThe immune system functions diversely across many organs to protect and maintain health. Importantly, the host’s immune system can regulate the genomic integrity across species and generations. Protection extends to all body barrier interfaces between the external and internal environment, where invasion of microbial agents is prevented or dealt with. The protection also acts against deleterious somatic mutations of host cells. The immune system is vital for maintenance of the health of all other body systems.\n\nEssentially, the process of DNA-based evolution, besides adapting organisms to their physical environments, has pitted organism against organism in the quest for ultimate survival. According to Darwinian principles, the surviving organisms are the most successful either in conquering and terminating competing organisms, or in reaching symbiotic balance with them. That process requires protection of host DNA and also facilitates relatively rapid genomic constitutional adaptation by acquiring and modifying useful DNA from the environment5. Indeed, the organism’s DNA is added to, modified and diversified to keep ahead of the ‘genetic superiority game’ by mutation, plasmid transfer, viral transduction, mitotic translocations, and meiotic acquisition. The immune system undergoes constant modification of innate and adaptive immunity with exposure to antigenic stimuli both at the individual and the population levels.\n\nThe mammalian immune system represents one of the final central arbiters over the course of human Darwinian evolution. Many of the advances necessary for human adaptation have been moderated, directed and shaped by the influence of the immune system. Most fundamentally, the defense against infection and therefore survival of individuals to permit reproduction and species continuity, is underpinned by immune system function. Less obvious, though equally fundamental, is the role of the immune system in maintenance of the organism’s homeostasis through removal of cells whose somatic mutations made them deleterious. Natural selection applies not only to the successful reproduction of entire organisms, but also to the clonal reproduction of cell lineages, both cancer and immunological, within an organism6.\n\nThe genes for the hypervariable regions of the antibody molecule and the genes for the hypervariable regions of the T-cell receptor, mutate at a much faster rate (hypermutation) than somatic genes under usual environmental pressure. Somatic mutation is a relatively slow process where genetic changes through selection pressure on survival and evolution usually require generations of cell divisions. The immune system genes, however, constantly rapidly mutate in order to generate diverse conformations capable of binding the multitude of antigens to which an individual is exposed. Many of those antigens might be associated with threat and danger, for example, from microbial invasion. The immune system design has necessarily evolved, through continuous successive approximation, to detect subtle molecular cell surface aberrations. This occurs through both non-specific, and specific B- and T-cell, mechanisms in an elegantly integrated manner.\n\n\nHow the genome monitors itself and evolves\n\nSomatic changes of organisms occur generally at a gradual pace as part of the slow, but effective, evolutionary process through such mechanisms as random mutation, natural selection and viral infection. For example human morphological characteristics, such as stature, brain size and tooth size change at rates ranging from 0.3 darwins to 65 darwins7. Microbial DNA sequences, for example from retroviruses like HIV, Herpes viruses and Mycobacterium tuberculosis, have been identified in the human genome, and these genes must have been structurally incorporated over time from repeated exposure, interaction and exchange between mammalian and microbial DNA8,9. Human Endogenous Retroviruses (HERVs) are estimated to make up 8% of the human genome, though fragmented and replication incompetent, it bears testament to long and intimate genetic interactions between a parasite with a few genes and 10,000 nucleotides, and a host of some 22,000 genes and some 2.85 billion nucleotides10,11. Interestingly, the (uninfected) C57 black mouse has several whole genomic copies of the LMP56 retrovirus in its germline12. Clearly, the retrovirus became inserted into the murine genetic complement in the mammal's evolutionary past13–15. When infected with the virus in the experimental situation, the mouse develops a chronic immunodeficiency disease, the clinical course of which parallels HIV/AIDS in humans16. It is now suggested that this chronic disease state is due to the murine immune system failing to differentiate between self and non-self, such that it homeostatically attenuates or down-regulates the response against the virus in vivo17,18. Failure to resolve the disease is due to persisting viral (self) antigens. The experimental similarity to the immune response in murine cancer models is strikingly compelling. In the case of cancer in the mouse the persisting antigens are due to the growing cancer which appears to exert a similar attenuating effect19.\n\nOver the millennia the constant exchange of genetic material between host and environmental microorganisms has offered incremental adaptive advantage to both organisms, but in fundamentally different ways, perhaps comprising the ultimate symbiotic relationship, since both have evolved and survived20. However, many organisms can expand rapidly, possess mechanisms for evasion of host defences, and can mutate at a rate that far outpaces somatic evolutionary change via much faster division/reproduction rates. This may explain the immune system’s evolved ability to match these rapid microbial mutational rates to more effectively neutralize them via innate mechanisms, antibody production and cellular responses. Examples are the microorganisms that rapidly expand and produce outbreaks of disease in humans, animals, plants and insects, sometimes with transmission across species. Rapid, immediate 'revolutionary' adaptive change is advantageous to keep the immune system ahead of microbial mutation, virulence and growth21. To oppose mutated, infected and otherwise aberrant cells, the immune system has a number of adaptive and protective mechanisms. These include somatic hypermutation genes for generation of hypervariable region binding domains for antibody molecules by plasma (B-) cells, and for hypervariable T-cell surface receptors by T-cells for rapid response to antigen exposure. In this way, adaptive immune responses can rapidly generate multiple molecules with variable affinity for binding whole or fragmented antigens. An analogy would be 'random number generation' to break unknown digital codes, or in contemporary terms to 'hack into' a computer system across encrypted firewalls22–24.\n\nWithout adequate host organism defence, infection would cause cellular damage and death. Humans are estimated to harbour some 1014 microbes, mostly bacteria, while we consist of only 1013 mammalian cells25–27. It might therefore be argued that in a cellular sense we are more bacterial than mammalian in constitution. Let's however, adhere to the notion that the host is the mammalian component. The human body, like any other multicellular organism, should be treated as a complex ecosystem whose balance is dynamically maintained by feedback interactions amongst its parts.\n\nTo understand the human immune system, we must appreciate that each facet of the immune system has evolved concurrently as life itself has evolved. The mammalian genome, therefore constantly monitors itself through the actions of the immune system, both non-specific and adaptive. This is in order to achieve a state of evolving homeostasis to achieve progressive protection of the genome, and of cellular and tissue function, as the environmental, microbial and other pressures continually change.\n\n\nHistory of immune system development and cellular aberration\n\nLife on earth commenced between 3 and 4 Ga (giga/billion years ago) as unicellular organisms adapted to survive environmental hazards through rapid reproduction and repopulation. Some 1.2 Ga algal mats developed as the first multicellular organisms, and then about 1 Ga more complex chlorophyll-containing organisms evolved. About 450 Ma (mega/million years ago) even more complex plants developed and acquired fundamental innate static immune systems largely through intracellular anti-microbial molecules to resist infection principally from fungi, bacteria and viruses.\n\nAdaptive immunity developed rather precipitously around 450 Ma in primitive fish and amphibians, and with reptiles, about 300 Ma, this evolved rapidly for protection against infection.\n\nMammalian life began about 120 Ma, with immune system evolution to meet the need for local and systemic protection from invasive microorganisms, and placentation20. Indeed, for effective adaptive symbiosis the mammalian immune system must have developed evolutionary tolerance for specific microorganisms since some organisms conferred adaptive advantages and others did not.\n\nOver a mere 60 years or so, we have investigated the intricate interplay between non-specific (innate) and more specific (adaptive) immune mechanisms for fundamental evolutionary and developmental advantage. Often viewed as separate arms of the immune response, it is clear that they are rarely mutually exclusive or separate. The division arose for experimental explanatory research reasons, rather than physiological ones, but are inextricably inseparable.\n\nGenomic intrinsic mutational pressures and exogenous infection of cells are significant forces capable of exerting phenotypic change to produce cell membrane 'aberration'. During cellular transformation to dysplasia, metaplasia and malignancy, cell membrane changes are detectable. Since gene mutations occur about 1 in every 106 cell divisions, the risk of cellular aberration is high in rapidly dividing tissues, with some leading to malignant transformation. The immune system is the only system capable of high level detection and action, and must therefore detect aberrant cells early and remove them exceedingly effectively and efficiently, otherwise, the rates of cancers would exceed that observed clinically. About 106 cells form a 5mm diameter tumour from some 30 divisions (assuming a regular process applies).\n\n\nFundamental reactivity to aberrant antigens\n\nAberrations, arising from multiple events such as infection of cells, cellular injury, trauma, ageing or from genetic mutation, are reflected by cell surface expression of aberrant proteins, lipids (especially glycolipids) and carbohydrates. Detection of aberration through both non-specific and specific adaptive mechanisms is essential for destruction and removal of abnormal cells to restore tissue integrity. Membrane profile alterations from normal to dysplastic and malignant transformation are evident using magnetic resonance spectroscopy28–30. The immune system is carefully tuned to detect relatively subtle changes in proteins through the standard HLA systems via Class I and II molecules, and the far less explored CD1 system for the detection of lipid, glycolipid and carbohydrate molecules31. In addition, the Fc receptor mechanism of the non-specific arm of the immune system detects foreign and altered cells. Activation of granulocytes, macrophages, B-cells and T-cells pushes the immune system in one direction or in the other, producing either overall responsiveness/activation, or inhibition/tolerance. Increasingly, it is being appreciated that all levels of the immune system can either respond or inhibit. Therefore, infected, damaged or malignant cells can be either actively eliminated or tolerated. Clinically, this is precisely what is observed, in a variety of infections and malignancies. Indeed, chronic inflammatory states have emerged as the predominant illnesses affecting many individuals, including persistent infections, autoimmunity and malignancy. Diseases such as cancer, cardiovascular disease and diabetes are now appreciated as chronic persistent inflammatory states, capable of modulation by factors such as anti-inflammatory medication and immune modulation.\n\n\nThe cancer cell as an evolutionary entity\n\nCancer cells are often portrayed as profoundly defective 'rogue' cells. Certain acquired key mutations permit loss of cellular control in division and adhesion, to evade immune destruction. The extent of genetic heterogeneity occurring within the cancer mass(es), both primary and metastatic, appears considerable32–36.\n\nClearly, the cancer cell appears as an adaptive and highly evolved entity able to switch on certain genes to survive the onslaught of radiation and chemotherapy, despite having genetic/chromosomal errors. So, in this sense it is a very sophisticated survival machine. So much so, that a cancer cell is often described as being “immortalized”. For example, the HeLa cell line, from Henrietta Lacks who died of cervical cancer in 1951 has been cultivated for decades in tissue culture worldwide, with some 20 tonnes grown to date37,38.\n\nCancer remains a major protracted health problem globally despite decades of apparent sophisticated research and monies spent, with relatively minor reductions in mortality from advanced cancers of most types39. Indeed, perhaps the strategies and “paradigms” currently used for cancer research and therapeutic intervention might be incorrect. In 2010, the successes of cancer research efforts were again questioned40,41, while in 2008 the USA National Cancer Institute, in frank admission of glacial progress, sought insights from the physical sciences into cancer biology (via 12 new so-called Physical Science-Oncology Centers; PS-OCs), hoping for radically new thinking42. Novel ideas emerged from the PS-OC programme like the atavistic theory, where cancer is viewed not so much as a “dream run” of genetic accidents conferring extraordinary capabilities, but as a \"default state\" in reaction to an insult or stress, where cells abandon many recently-evolved capabilities to run on ancient core functionality – a sort of basic “safe mode” for cells. In other words, cancer is an inbuilt response to damage (or poor tissue environments) rather than a product of it43,44. Thus rather than cancer being a modern biological phenomenon, it has very deep evolutionary roots - confirmed by the fact that cancer is found across most classes of multicellular life, including simple organisms like hydra that possess only two cell types45.\n\nCancer represents a reversion to a more primitive eukaryotic cellular state. In the single-celled world, cells are effectively immortal, and their prime imperative is replication in the face of diverse challenges. Proliferation is thus the default state of unicellular life and it has had 4 billion years to evolve mechanisms to preserve it when threatened. A major transition in biology occurred between about 1.0 and 1.5 billion years ago with the evolution of multicellularity, and later with primitive metazoan multicellularity somewhere in the Cambrian period approximately 550 Ma. In many multicellular organisms somatic cells outsource their immortality to specialized germ cells, and accept apoptosis as the price. However, this ancient contract is vulnerable to 'cheating' (as with all cooperative biological systems) and so it must be policed by layer upon layer of regulatory control. When the control mechanisms are damaged or compromised, reversion to unconstrained proliferation may ensue, manifesting as a neoplasm. Cancer is thus an ever-present threat – an accident waiting to happen – because of “pre-programmed” deeply-entrenched, highly-protected and ancient genome parts which can be variously triggered, including by random damage. Like a genie in a bottle, the bottle can be shattered in many ways, but once the genie escapes it executes its agenda with ruthless efficiency and determinism. Thus cancer follows a broadly predictable pattern of behaviour across organ types and species, with primary tumours followed by EMT, motility, dissemination via the vasculature, colonization of remote organs, MET and secondary growth, suggesting it is a very basic biological phenomenon and not an aberration.\n\nThe atavistic theory makes some specific and testable predictions about cancer progression. In defaulting to an ancestral phenotype, cancer is more comfortable in, and may even engineer niches to recreate, conditions resembling the Proterozoic oceans in which multicellularity evolved. For example, the Proterozoic environment at that time was hypoxic (the second great oxygenation event did not take place until about 800 million years ago). Sure enough, cancer metabolism prefers the ancestral, but less energy efficient mechanism of anaerobic fermentation (glycolysis) over the more recently-evolved oxidative phosphorylation. This atavistic reversion to an ancient mode of metabolism is known as the Warburg Effect and has been widely recognized, even if unexplained, since the 1930s46. Another example concerns the long-recognized resemblance between embryo development and tumours, common features being hypoxia, cell motility (EMT), angiogenesis, invasiveness and rapid proliferation. But ontogeny roughly mimics phylogenetic evolution (von Bauer’s laws of ontogeny), so a reversion to a more primitive evolutionary state closely resembles a reversion to an embryonic developmental state. It is widely known that developmental genes tend to be inappropriately up-regulated in cancer47, and these are in turn the ancient genes controlling the basic body plan.\n\nIn a refinement for the broad-brush (and fairly uncontentious) reversion theory, Lineweaver, Davies and Vincent44 are examining the evolutionary ages of the genes that are up-regulated in cancer. This emerging field is known as phylostratigraphy. The atavism theory predicts that genes which are up-regulated in cancer (oncogenes) should be systematically older than those that are down-regulated (tumour suppressor genes), and that this skewed distribution should become more pronounced as a function of cancer progression in individual organisms. The theory thus makes a new prediction: that in cancer, there should be a correlation between gene ages and (anomalous) gene activity.\n\nIf this general trend towards a more primitive state is correct, it exposes an Achilles Heel of cancer. Reversion involves changes in cell functionality. The atavism theory claims that the gain of function in cancer is really regain of pre-existing ancient functionality. In contrast to the standard somatic mutation theory of cancer, in which neoplasms are treated as if created anew in each organism (and acquire their astonishing similarity via high-speed convergent evolution within the organism in a matter of months or years), the atavism theory asserts that cancer never invents anything new but merely appropriates, or co-opts, or re-acquires, existing biological functionalities that are deeply pre-programmed into the cells’ genetic and epigenetic pathways. Conversely, loss of function in cancer occurs when cells revert to a more primitive phenotype, because in so doing they jettison, or lose, or decouple from more-recently evolved (and usually more sophisticated) capabilities. Among the (relatively) more recently-evolved biological capabilities is the adaptive immune system. The atavism theory predicts that, as cancer advances, the neoplasm progressively loses contact with adaptive immunity and becomes, in effect, immunosuppressed. In the atavism theory, cancer immunosuppression – which is well known – represents a loss of function (due to a reversion to a phenotype that predates the evolution of adaptive immunity about 400 million years ago) rather than a gain of function conferring a survival trait (i.e. ability to evade immune attack). But immunosuppression is a two-edged sword. It may confer protection from immune attack, but it is also an obvious weakness, making the tumour environment vulnerable to infectious agents.\n\nThe history of the interaction of bacteria, viruses and cancer is a very long and somewhat confused one, since William Coley obtained some amazing clinical results over a century ago48. Some infections will boost the immune system and bring additional pressure on cancer cells, but some agents will directly infect the cancer cells preferentially in their immunosuppressed niches, for example oncolytic viruses. A variety of new approaches49–55 to immunotherapy exploits these features. The atavism theory predicts that advanced cancer will be particularly vulnerable to certain infectious agents, and specific treatment regimes have been advocated to take advantage of that aspect44,56.\n\n\nHomeostatic regulation of immune reactivity and cancer\n\nThe relapsing and remitting behaviour of many chronic inflammatory states, such as arthritis, inflammatory bowel diseases, multiple sclerosis, and thyroiditis is well recognised. Diabetes, cardiovascular diseases and cancers of all types are now being considered similarly. The fluctuating, oscillating nature of these diseases has largely confounded our understanding to date and remained frustratingly unexplained, but is indicative that the immune system must be transitioning between stimulation/activation and suppression/tolerance phases repeatedly to produce the observed clinical picture. Moreover, oscillatory behaviour is highly characteristic of any homeostatic biological system under negative feedback control. This cyclical dynamic is a physical expression of physiological control to maintain relative constancy of the milieu intérieur, as recognised by Claude Bernard around 1867, and later Walter Cannon. Physiological constancy, or homeostatic control, of the body's immune status requires proportioned synchrony between effector stimulation and regulatory functions to be operational. Many cyclical examples, such as the diurnal temperature cycles, peri-monthly menstrual cycles, and 24-hour cortisol cycles have been elucidated by close serial monitoring.\n\nThe association between cancer and the host immune response has been recognised for over a century57–62. In animals, North et al. and more recently Klatzmann et al., demonstrated that the time of delivery of cytotoxic agents after tumour transplantation was crucial in determining whether tumour regression occurred or not63–72. Early clinical observations of inflammation and cancer regression were made by those treating cancer57–60, particularly the development of infection/fever after surgery. Chronic inflammation has been associated with cancer development, for example chronic ulceration and Marjolin's squamous cell cancer of the skin.\n\nThe immune system has innate and adaptive arms. C-Reactive Protein (CRP) is a non-specific functional analogue of immunoglobulin that binds to self/non-self cellular breakdown products of inflammation to initiate the adaptive immune responses73,74. T & B cells respond to cellular changes due to infection, damage or mutagenesis. To fine tune and limit these responses, the ensuing immune response is down-regulated paradoxically by the same cytokines and receptors that initiated it, but on functionally different cell types. Regulatory T-cells play a major role in this homeostatic attenuation and experimental and clinical evidence has shown that when these cells are either removed or blocked, cancer can completely regress, while autoimmune conditions may develop or worsen.\n\nIn recent years, it has become clear that the immune system recognises and processes both self- and non-self antigens to either respond or tolerate the antigen, but that homeostatic balance usually prevails.\n\nImmune responses can therefore be thought of as a “bi-stable” system existing in either of two principal states (responsive or tolerant). Antigen is the prime mover for either of these two states, and cytokines, most notably interleukin-2 (IL2), provide the feedback loop in the time domain to govern the direction. If antigen is continuously supplied to such a system (due to tumor cell growth/turnover) logic and physiology dictate that this response must oscillate73–82. Bi-stable oscillatory systems are characteristic of any homeostatic system with a feedback loop (Figure 1).\n\n\nAnti-cancer agents and immune responses\n\nCytotoxic agents inhibit cell division to therapeutically damage and kill tumor cells. However, cancer cells divide asynchronously. About 20–30% of malignant cells within many solid cancers are dividing at any one time-point (greater rates of division occur in some cancers such as childhood leukaemia and testicular carcinoma). Regimens have evolved often with weekly dosing of sequential 'lines' (1st, 2nd, 3rd etc) of treatments or in combinations. Repetitive dosing of agents inducing multiple cycles of cell damage and antigen release (vaccination events) from the tumour is emerging as highly significant56,75–77.\n\nCells of the immune system rapidly divide, but they divide synchronously and alternately (effector then regulatory) at different times sequentially to initiate then terminate an immune response in the time domain73–82.\n\nCytotoxic agents, unless applied discriminately, can aimlessly ablate different groups of proliferating immunological cells, as well as any proliferating tumor cells.\n\nIt is now clear that the immune system is not ignorant to the presence of tumors and that the normal homeostatic regulatory mechanisms are at the seat of the problem. This explains why immuno-modulatory agents, such as IL2, CTLA4, PD1/L1 monoclonal antibodies can deliver random dramatic complete responses in a limited percentage of late-stage cancer patients by interfering with the pre-existing homeostatic suppression/tolerance81–88. All of these agents can induce tolerance. The lack of efficacy of these agents in most patients is explained by induction of tolerance with some doses via regulatory T-cells while activating with other doses, the net balance of which can determine overall clinical outcome. Interestingly, autoimmunity can result from 'overdrive' of the immune system by some immunostimulatory agents and this has often been associated with better clinical responses against the cancer.\n\n\nImproving results of natural selection\n\nMost aberrant cells appearing in the human body as a result of somatic mutations are detected and disposed of by the immune system. Some are not and can produce pathology, with the majority of clinical cases of cancer occurring in older patients. This is explicable by the fact that natural selection operates principally by differential reproduction, consequently it is unable to operate for biological characteristic selection in non-reproductive (older) individuals. Thus, over the generations immune responses to malignant cells appearing in young people became adjusted by natural selection and, statistically speaking, operate efficiently, while such responses in older age were not “reachable” by natural selection for adjustment. This principle is not only applicable to specific immune responses, but encompasses the entire regulation of homeostatic balance of an organism. In practical terms, clinical intervention should imitate adaptation by selection of immunological processes occurring in younger organisms, to support, adjust and enhance natural operation of immune systems of older patients.\n\n\nConcluding remarks and implications\n\nAlthough knowledge has developed deeply concerning the immune system and cancer immunology, our contemporary understanding needs to be placed in evolutionary perspective. Our immune systems are the adaptive result of the necessity for defence against persistent selective pressures from environmental microbial pathogens. Over the millennia, the immune system and other body cells have undergone a continuous adaptive symbiotic process of synchronous, coordinated, cooperative, progressive immunological and somatic evolutionary change to provide what we observe today. Gradual evolution of adaptive immunity against infected and aberrant cells now explains many of the observations regarding cancer immunity and clinical responses. It is gradually being appreciated that normal immune regulatory mechanisms are holding back a primed immune response from selectively killing cancer cells. With an appreciation that immuno-modulation of pre-existing endogenous immune responses appears to occur with most cancer therapies, there is the serious prospect that serial immune monitoring might define optimal time-points for targeted administration of therapies to engineer effective complete clinical responses in a much more predictable, reliable and durable manner in the future. If achievable, increased long-term survival from advanced cancer, with reduced toxicity, might become a reality by harnessing the immuno-modulatory capacity of many currently existing therapeutic agents. The cost savings would be truly enormous89.\n\n\nAuthors’ information\n\nThe authorship represents a unique collaboration between diverse disciplines with contributors having backgrounds and qualifications in cancer surgery, immunology, immunotherapy (BJC), basic science (MLA), evolutionary biology, anatomy, anthropology (MH), and the physical sciences, cancer biology (PD). As such, this work aims to approach the problem of cancer development and immune system recognition/responses uniquely from a scientific evolutionary perspective to explain many of the clinical observations that have been already made to date. Emanating from this understanding, new approaches and therapies might then be fruitfully generated for science and clinical medicine.\n\n\nList of abbreviations\n\nIL2 interleukin-2\n\nCTLA4 cytotoxic T-lymphocyte associated protein 4\n\nPD-1 programmed cell death protein 1\n\nPD-L1 programmed death-ligand 1\n\nHLA human leukocyte antigen\n\nCD1 cluster of differentiation 1\n\nDNA deoxyribonucleic acid\n\nEMT epithelial-mesenchymal transition\n\nMET mesenchymal–epithelial transition\n\nPS-OC Physical Science-Oncology Centers\n\nHIV human immunodeficiency virus\n\nAIDS acquired immunodeficiency syndrome\n\nHREV human endogenous retroviruses\n\nMa mega/million years\n\nGa giga/billion years",
"appendix": "Author contributions\n\n\n\nBJC devised and wrote the main text; MLA contributed to the discussion on HIV/retroviruses and hyper-variable T-cell receptors; MH contributed to evolutionary discussions; PCWD contributed to cancer cell evolution. All authors prepared, read and approved the manuscript to reach the final content.\n\n\nCompeting interests\n\n\n\nThe authors declare that they have no financial or non-financial competing interests, nor any conflicts of interests.\n\n\nGrant information\n\nThis work was supported in part by NIH grant U54 CA143682.\n\n\nReferences\n\nHackett CJ: On the Origin of the Human Treponematoses (Pinta, Yaws, Endemic Syphilis and Venereal Syphilis). Bull World Health Organ. 1963; 29: 7–41. PubMed Abstract | Free Full Text\n\nHackett CJ: An introduction to diagnostic criteria of syphilis, treponarid and yaws (treponematoses) in dry bones, and some implications. Virchows Arch A Pathol Anat Histol. 1975; 368(3): 229–41. PubMed Abstract | Publisher Full Text\n\nHolloway KL, Henneberg RJ, de Barros Lopes M, et al.: Evolution of human tuberculosis: a systematic review and meta-analysis of paleopathological evidence. Homo. 2011; 62(6): 402–58. PubMed Abstract | Publisher Full Text\n\nHolloway KL, Staub K, Rühli F, et al.: Lessons from history of socioeconomic improvements: a new approach to treating multi-drug-resistant tuberculosis. J Biosoc Sci. 2014; 46(5): 600–20. PubMed Abstract | Publisher Full Text\n\nAlizon S, de Roode JC, Michalakis Y: Multiple infections and the evolution of virulence. Ecol Lett. 2013; 16(4): 556–67. PubMed Abstract | Publisher Full Text\n\nGreaves M, Maley CC: Clonal evolution in cancer. Nature. 2012; 481(7381): 306–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHenneberg M: The rate of human morphological microevolution and taxonomic diversity of hominids. Studies in Historical Anthropology. 2006; 4.2004: 49–59. Reference Source\n\nRasmussen HB: Interactions between Exogenous and Endogenous Retroviruses. J Biomed Sci. 1997; 4(1): 1–8. PubMed Abstract | Publisher Full Text\n\nKrone B, Kölmel KF, Henz BM, et al.: Protection against melanoma by vaccination with Bacille Calmette-Guerin (BCG) and/or vaccinia: an epidemiology-based hypothesis on the nature of a melanoma risk factor and its immunological control. Eur J Cancer. 2005; 41(1): 104–117. PubMed Abstract | Publisher Full Text\n\nHohn O, Hanke K, Bannert N: HERV-K(HML-2), the Best Preserved Family of HERVs: Endogenization, Expression, and Implications in Health and Disease. Front Oncol. 2013; 3: 246. PubMed Abstract | Publisher Full Text | Free Full Text\n\nInternational Human Genome Sequencing Consortium: Finishing the euchromatic sequence of the human genome. Nature. 2004; 431(7011): 931–45. PubMed Abstract | Publisher Full Text\n\nJones RB, Garrison KE, Mujib S, et al.: HERV-K-specific T cells eliminate diverse HIV-1/2 and SIV primary isolates. J Clin Invest. 2012; 122(12): 4473–89. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChattopadhyay SK, Lander MR, Rands E, et al.: Structure of endogenous murine leukemia virus DNA in mouse genomes. Proc Natl Acad Sci U S A. 1980; 77(10): 5774–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKubo Y, Nakagawa Y, Kakimi K, et al.: Molecular cloning and characterization of a murine AIDS virus-related endogenous transcript expressed in C57BL/6 mice. J Gen Virol. 1994; 75(pt 4): 881–8. PubMed Abstract | Publisher Full Text\n\nKubo Y, Kakimi K, Higo K, et al.: Possible origin of murine AIDS (MAIDS) virus: conversion of an endogenous retroviral p12gag sequence to a MAIDS-inducing sequence by frameshift mutations. J Virol. 1996; 70(9): 6405–9. PubMed Abstract | Free Full Text\n\nJolicoeur P: Murine acquired immunodeficiency syndrome (MAIDS): an animal model to study the AIDS pathogenesis. FASEB J. 1991; 5(10): 2398–405. PubMed Abstract\n\nBeilharz MW, Sammels LM, Paun A, et al.: Timed ablation of regulatory CD4+ T cells can prevent murine AIDS progression. J Immunol. 2004; 172(8): 4917–25. PubMed Abstract | Publisher Full Text\n\nMeng L, Tompkins M, Miller M, et al.: Lentivirus-activated T regulatory cells suppress T helper cell interleukin-2 production by inhibiting nuclear factor of activated T cells 2 binding to the interleukin-2 promoter. AIDS Res Hum Retroviruses. 2014; 30(1): 58–66. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOndondo B, Jones E, Godkin A, et al.: Home sweet home: the tumor microenvironment as a haven for regulatory T cells. Front Immunol. 2013; 4: 197. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee A, Nolan A, Watson J, et al.: Identification of an ancient endogenous retrovirus, predating the divergence of the placental mammals. Philos Trans R Soc Lond B Biol Sci. 2013; 368(1626): 20120503. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWright BE, Schmidt KH, Minnick MF: Kinetic models reveal the in vivo mechanisms of mutagenesis in microbes and man. Mutat Res. 2013; 752(2): 129–37. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFrench DL, Laskov R, Scharff MD: The role of somatic hypermutation in the generation of antibody diversity. Science. 1989; 244(4909): 1152–1157. PubMed Abstract | Publisher Full Text\n\nTomlinson IM, Walter G, Jones PT, et al.: The imprint of somatic hypermutation on the repertoire of human germline V genes. J Mol Biol. 1996; 256(5): 813–17. PubMed Abstract | Publisher Full Text\n\nJones JM, Gellert M: The taming of a transposon: V(D)J recombination and the immune system. Immunol Rev. 2004; 200: 233–48. PubMed Abstract | Publisher Full Text\n\nBianconi E, Piovesan A, Facchin F, et al.: An estimation of the number of cells in the human body. Ann Hum Biol. 2013; 40(6): 463–71. PubMed Abstract | Publisher Full Text\n\nSavage DC: Microbial ecology of the gastrointestinal tract. Annu Rev Microbiol. 1977; 31: 107–133. PubMed Abstract | Publisher Full Text\n\nGerritsen J, Smidt H, Rijkers GT, et al.: Intestinal microbiota in human health and disease: the impact of probiotics. Genes Nutr. 2011; 6(3): 209–240. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMountford CE, Grossman G, Reid G, et al.: Characterization of transformed cells and tumors by proton nuclear magnetic resonance spectroscopy. Cancer Res. 1982; 42(6): 2270–6. PubMed Abstract\n\nMackinnon WB, Huschtscha L, Dent K, et al.: Correlation of cellular differentiation in human colorectal carcinoma and adenoma cell lines with metabolite profiles determined by 1H magnetic resonance spectroscopy. Int J Cancer. 1994; 59(2): 248–61. PubMed Abstract | Publisher Full Text\n\nYakoub D, Keun HC, Goldin R, et al.: Metabolic profiling detects field effects in nondysplastic tissue from esophageal cancer patients. Cancer Res. 2010; 70(22): 9129–36. PubMed Abstract | Publisher Full Text\n\nCoventry B, Heinzel S: CD1a in human cancers: a new role for an old molecule. Trends Immunol. 2004; 25(5): 242–248. PubMed Abstract | Publisher Full Text\n\nBurrell RA, McGranahan N, Bartek J, et al.: The causes and consequences of genetic heterogeneity in cancer evolution. Nature. 2013; 501(7467): 338–45. PubMed Abstract | Publisher Full Text\n\nMunoz J, Swanton C, Kurzrock R: Molecular profiling and the reclassification of cancer: divide and conquer. Am Soc Clin Oncol Educ Book. 2013: 127–34. PubMed Abstract | Publisher Full Text\n\nMurugaesu N, Chew SK, Swanton C: Adapting clinical paradigms to the challenges of cancer clonal evolution. Am J Pathol. 2013; 182(6): 1962–71. PubMed Abstract | Publisher Full Text\n\nJamal-Hanjani M, Thanopoulou E, Peggs KS, et al.: Tumour heterogeneity and immune-modulation. Curr Opin Pharmacol. 2013; 13(4): 497–503. PubMed Abstract | Publisher Full Text | Free Full Text\n\nS Datta R, Gutteridge A, Swanton C, et al.: Modelling the evolution of genetic instability during tumour progression. Evol Appl. 2013; 6(1): 20–33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNcayiyana DJ: The extraordinary story of the life after death of Henrietta Lacks. S Afr Med J. 2011; 101(3): 141. PubMed Abstract\n\nNjoku DB: The immortal life of Henrietta Lacks. Anesth Analg. 2013; 117(1): 286. PubMed Abstract | Publisher Full Text\n\nKiberstis PA, Travis J: Celebrating a glass half-full. Science. 2006; 312(5777): 1157. Publisher Full Text\n\nTuma RS: Large trials, small gains: is change on the way? J Natl Cancer Inst. 2010; 102(16): 1216–1217, 1223. Publisher Full Text\n\nUS National Institues of Health. President’s Cancer Panel. 2007– 2008 annual report. 2012. Reference Source\n\nDavies P: Rethinking cancer. Physics World. 2010; 23(6): 28. Reference Source\n\nDavies PC, Lineweaver CH: Cancer tumors as Metazoa 1.0: tapping genes of ancient ancestors. Phys Biol. 2011; 8(1): 015001. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLineweaver CH, Davies PC, Vincent MD: Targeting cancer’s weaknesses (not its strengths): Therapeutic strategies suggested by the atavistic model. Bioessays. 2014; 36(9): 827–35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDomazet-Lošo T, Klimovich A, Anokhin B, et al.: Naturally occurring tumours in the basal metazoan Hydra. Nat Commun. 2014; 5: 4222. PubMed Abstract | Publisher Full Text\n\nWarburg O: On the origin of cancer cells. Science. 1956; 123(3191): 309–314. PubMed Abstract | Publisher Full Text\n\nAdamson ED: Oncogenes in development. Development. 1987; 99(4): 449–71. PubMed Abstract\n\nColey WB: The treatment of malignant tumors by repeated inoculations of erysipelas: with a report of ten original cases. Am J Med Sci. 1893; 105(5): 487–511. Reference Source\n\nKim SH, Castro F, Paterson Y, et al.: High efficacy of a Listeria-based vaccine against metastatic breast cancer reveals a dual mode of action. Cancer Res. 2009; 69(14): 5860–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGravekamp C, Paterson Y: Harnessing Listeria monocytogenes to target tumors. Cancer Biol Ther. 2010; 9(4): 257–65. PubMed Abstract | Publisher Full Text\n\nNguyen VH, Kim HS, Ha JM, et al.: Genetically engineered Salmonella typhimurium as an imageable therapeutic probe for cancer. Cancer Res. 2010; 70(1): 18–23. PubMed Abstract | Publisher Full Text\n\nGalmbacher K, Heisig M, Hotz C, et al.: Shigella mediated depletion of macrophages in a murine breast model is associated with tumor regression. PLoS One. 2010; 5(3): e9572. PubMed Abstract | Publisher Full Text | Free Full Text\n\nQuispe-Tintaya W, Chandra D, Jahangir A, et al.: Nontoxic radioactive Listeriaat is a highly effective therapy against metastatic pancreatic cancer. Proc Natl Acad Sci USA. 2013; 110(21): 8668–73. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCoster TS, Hoge CW, VanDeVerg LL, et al.: Vaccination against shigellosis with attenuated Shigella flexneri 2a strain SC602. Infect Immun. 1999; 67(7): 3437–43PubMed Abstract | Free Full Text\n\nDonnelly O, Harrington K, Melcher A, et al.: Live viruses to treat cancer. J R Soc Med. 2013; 106(8): 310–4. PubMed Abstract | Publisher Full Text\n\nCoventry BJ, Lilly CA, Hersey P, et al.: Prolonged repeated vaccine immuno-chemotherapy induces long-term clinical responses and survival for advanced metastatic melanoma. J Immunother Cancer. 2014; 2: 9. Publisher Full Text\n\nColey WB: Contribution to the knowledge of sarcoma. Ann Surg. 1891; 14(3): 199–220. PubMed Abstract | Free Full Text\n\nColey WB: Treatment of inoperable malignant tumors with toxins of erysipelas and the bacillus prodigiosus. Trans Am Surg Assn. 1894; 12: 183–212.\n\nColey WB: Disappearance of a recurrent carcinoma after injections of mixed toxins. Ann Surg. 1912; 55: 897–8.\n\nErysipelas and prodigiosus toxins (Coley). JAMA. 1934; 103(14): 1070–71. Publisher Full Text\n\nHoption Cann SA, van Netten JP, van Netten C, et al.: Spontaneous regression: a hidden treasure buried in time. Med Hypotheses. 2002; 58(2): 115–9. PubMed Abstract | Publisher Full Text\n\nHoption Cann SA, van Netten JP, van Netten C: Dr William Coley and tumour regression: a place in history or in the future. Postgrad Med J. 2003; 79(938): 672–680. PubMed Abstract | Free Full Text\n\nNorth RJ, Awwad M: T cell suppression as an obstacle to immunologically-mediated tumor regression: elimination of suppression results in regression. Prog Clin Biol Res. 1987; 244: 345–58. PubMed Abstract\n\nAwwad M, North RJ: Immunologically mediated regression of a murine lymphoma after treatment with anti-L3T4 antibody. A consequence of removing L3T4+ suppressor T cells from a host generating predominantly Lyt-2+ T cell-mediated immunity. J Exp Med. 1988; 168(6): 2193–206. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAwwad M, North RJ: Cyclophosphamide (Cy)-facilitated adoptive immunotherapy of a Cy-resistant tumour. Evidence that Cy permits the expression of adoptive T-cell mediated immunity by removing suppressor T cells rather than by reducing tumour burden. Immunology. 1988; 65(1): 87–92. PubMed Abstract | Free Full Text\n\nAwwad M, North RJ: Sublethal, whole-body ionizing irradiation can be tumor promotive or tumor destructive depending on the stage of development of underlying antitumor immunity. Cancer Immunol Immunother. 1988; 26(1): 55–60. PubMed Abstract | Publisher Full Text\n\nHill JO, Awwad M, North RJ: Elimination of CD4+ suppressor T cells from susceptible BALB/c mice releases CD8+ T lymphocytes to mediate protective immunity against Leishmania. J Exp Med. 1989; 169(5): 1819–27. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAwwad M, North RJ: Cyclophosphamide-induced immunologically mediated regression of a cyclophosphamide-resistant murine tumor: a consequence of eliminating precursor L3T4+ suppressor T-cells. Cancer Res. 1989; 49(7): 1649–54. PubMed Abstract\n\nNorth RJ, Awwad M, Dunn PL: The immune response to tumors. Transplant Proc. 1989; 21(1 Pt 1): 575–7. PubMed Abstract\n\nNorth RJ, Awwad M: Elimination of cycling CD4+ suppressor T cells with an anti-mitotic drug releases non-cycling CD8+ T cells to cause regression of an advanced lymphoma. Immunology. 1990; 71(1): 90–5. PubMed Abstract | Free Full Text\n\nAwwad M, North RJ: Radiosensitive barrier to T-cell-mediated adoptive immunotherapy of established tumors. Cancer Res. 1990; 50(8): 2228–33. PubMed Abstract\n\nDarrasse-Jèze G, Bergot AS, Durgeau A, et al.: Tumor emergence is sensed by self-specific CD44hi memory Tregs that create a dominant tolerogenic environment for tumors in mice. J Clin Invest. 2009; 119(9): 2648–62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCoventry BJ, Ashdown ML, Quinn MA, et al.: CRP identifies homeostatic immune oscillations in cancer patients: a potential treatment targeting tool? J Transl Med. 2009; 7: 102. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAshdown ML Coventry BJ: A Matter of Time. Australasian Science. 2010; 18–20. Reference Source\n\nCoventry BJ, Hersey P, Halligan A-M, et al.: Immuno-Chemotherapy Using Repeated Vaccine Treatment Can Produce Successful Clinical Responses in Advanced Metastatic Melanoma. Journal of Cancer Therapy. 2010; 1: 205–213. Publisher Full Text\n\nCoventry BJ, Ashdown ML, Markovic SN: Immune Therapies for Cancer: Bimodality—The Blind Spot to Clinical Efficacy—Lost in Translation. J Immunother. 2011; 34: 717.\n\nCoventry BJ, Ashdown ML: Complete clinical responses to cancer therapy caused by multiple divergent approaches: a repeating theme lost in translation. Cancer Manag Res. 2012; 4: 137–149. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCoventry BJ, Ashdown ML: The 20th anniversary of interleukin-2 therapy: bimodal role explaining longstanding random induction of complete clinical responses. Cancer Manag Res. 2012; 4: 215–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcNally A, Hill GR, Sparwasser T, et al.: CD4+ CD25+ regulatory T cells control CD8+ T-cell effector differentiation by modulating IL-2 homeostasis. Proc Natl Acad Sci U S A. 2011; 108(18): 7529–7534. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoyman O, Sprent J: The role of interleukin-2 during homeostasis and activation of the immune system. Nat Rev Immunol. 2012; 12(3): 180–190. PubMed Abstract | Publisher Full Text\n\nJain N, Nguyen H, Chambers C, et al.: Dual function of CTLA-4 in regulatory T cells and conventional T cells to prevent multiorgan autoimmunity. Proc Natl Acad Sci U S A. 2010; 107(4): 1524–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nA delicate balance: tweaking IL-2 immunotherapy. Nat Med. 2012; 18(2): 208–209. PubMed Abstract | Publisher Full Text\n\nHodi FS, O’Day SJ, McDermott DF, et al.: Improved survival with ipilimumab in patients with metastatic melanoma. N Engl J Med. 2010; 363(8): 711–723. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPrieto PA, Yang JC, Sherry RM, et al.: CTLA-4 blockade with ipilimumab: long-term follow-up of 177 patients with metastatic melanoma. Clin Cancer Res. 2012; 18(7): 2039–2047. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTopalian SL, Hodi FS, Brahmer JR, et al.: Safety, activity, and immune correlates of anti-PD-1 antibody in cancer. N Engl J Med. 2012; 366(26): 2443–54. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrahmer JR, Tykodi SS, Chow LQ, et al.: Safety and activity of anti-PD-L1 antibody in patients with advanced cancer. N Engl J Med. 2012; 366(26): 2455–65. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOtt PA, Hodi FS, Robert C: CTLA-4 and PD-1/PD-L1 Blockade: New Immunotherapeutic Modalities with Durable Clinical Benefit in Melanoma Patients. Clin Cancer Res. 2013; 19(19): 5300–5309. PubMed Abstract | Publisher Full Text\n\nWolchok JD, Kluger H, Callahan MK, et al.: Nivolumab plus ipilimumab in advanced melanoma. N Engl J Med. 2013; 369(2): 122–33. PubMed Abstract | Publisher Full Text\n\nMurphy KM, Topel RH: The value of health and longevity. J Polit Econ. 2006; 114(51): 871–904. Reference Source"
}
|
[
{
"id": "9959",
"date": "13 Aug 2015",
"name": "Jonathan M. Austyn",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article contains some very interesting ideas and concepts. However, to this reviewer (and with the greatest respect to the authors), it tends to read perhaps as ‘stream of consciousness’ writing rather than a considered scientific article. The authors show abundant enthusiasm in their writing, but this often appears to mask real scientific rigour. It is very difficult to identify a key hypothesis (or hypotheses), the structure of the sections seems to lack definition or focus, and it is sometimes almost impossible to understand what the main conclusions are from each. Furthermore, in a number of places, there seems to be a significant lack of scientific accuracy. For example, regarding some statements regarding the immune system, the authors do not clearly discriminate between somatic recombination (which applies to both T cell receptors and B cell receptors), somatic hypermutation (which applies only to the latter), hypervariable regions (which are present in both) and mutation (per se). As another example, in their discussion of atavism theory, the authors use the term ‘tumour immunosuppression’ but appear to apply this to the tumour rather than the host, where it really belongs. Even the authors’ brief review of the evolution of the immune system seems to lack sufficient focus going from species to species during evolutionary time. There are also some rather vexing ‘throwaway statements’. As just one example, regarding cancer, the authors state “The reason why cancer occurs at all in humans and animals thus remains a mystery”. While it is certainly true that much remains unknown, the authors might usefully consider further Burnet’s concept of tumour surveillance, particularly in its more recent form comprising immunoediting, equilibrium and escape phases. It might also be valuable to reconsider, in relation to the immune system’s apparent capacity to eliminate tumours whether or not the immune system actually helps to eliminate the infectious agents (e.g. viruses) that can cause tumorigenesis rather than actual or potentially malignant cells (and, on that point, it is perhaps a little surprising that there is no mention of generic ‘DAMPs’ and ‘PAMPs’). A further criticism is that the choice of immunological mechanisms under discussion sometimes feels rather random – why, for example, do the authors specifically focus on Fc receptors rather than complement receptors, or C-reactive protein rather than the many other molecules that play similar or related roles? Finally, regarding the single diagram that is presented, it is completely unclear why the authors have chosen to illustrate an ‘IL-2 [sic] feedback loop’; without any justification this seems to be over-simplistic in the extreme. Nevertheless, to return to the initial point: this article does appear to contain some very interesting material that would potentially be of value to readers of the journal. To do these the greatest justice, however, really does seem to require a very careful, focussed, and considered rewriting of the present text into a completely revised and possibly restructured article.",
"responses": []
},
{
"id": "10255",
"date": "10 Sep 2015",
"name": "Angus Dalgleish",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI was asked to review this article, with full access to previous reviewer's report, Jonathan M Austyn, from the John Radcliffe Hospital. I agree that the article contains some very interesting ideas and concepts, which will be very stimulatory to a wide readership, particularly those studying cancer and practicing cancer (oncologists) who are slowly becoming aware that the immune system is important in the control of cancer and hence the development and evolution.The review touches on some very broad and very interesting concepts, particularly with regards to evolution over time and the evolutionary difference between innate and adaptive immune systems and he has some of the concepts that are relevant to the hypothesis but perhaps not been elucidated that clearly, especially with regards to the concept of atavism and this has been pointed out by Jonathan Austyn with regards subtle differences in somatic recombination and hypermutation and mutation.I also agree with rather more reaching, throwaway statements, such as 'cancer occurs in all humans and animals, thus remaining a mystery when several reasons have already been explored, such as random mutations and the concept of escaping tumour surveillance.cancerWhereas references made to many infectious agents and the fact that they have ended up making up the vast majority of the intron genome, which has been shown to be due to the incorporation of many virus and bacterial sequences. The hypothesis depends very much on the shaping of these agents and the immune system but does not deal adequately with how the immune is shaped by exposure to infectious agents during infant and childhood development. Much of the references of this article is in large chunks, whereas the authors' papers are listed in large batches, for example; the work by North and colleagues are all listed together and referenced together.I feel that in addressing this, that they have not acknowledged the work of others who have spent a long time working out the affects of basic infectious diseases on the immune system and how this translates to chronic infectious diseases and one author in particular stands out for his work on this and that is Graham Rook, he has published much on this about the exposure of pathogens and the dirt theory, how it impacts on chronic diseases in later life in the western world and, indeed, a review by him and this reviewer (Rook and Dalgleish, 2011) published in Immunological reviews in 2011 goes into great detail with regards to its impact on cancer.Similarly, the impact of both infectious and non-infectious chronic activation/inflammation, which is relevant to many of the issues raised in this article, have not been addressed sufficiently, especially as it is so relevant to many of the speculations discussed. Again, there are a large number of authors and contributors who have gone from broad brush theories of the association of chronic inflammation and cancer, to those who have gone into great detail pointing out how it impacts on the molecular level of mutations in suppressor genes, such as P53, all the way through to immunological hypoxic pathways, etc., and I do not recognise this from the reference list which, as mentioned, is very block buster in its approach with an author's several contributions all being listed together throughout.I also feel that, with regards to the homeostatic component focusing purely in Interleukin-2, that although it is true that it is one of several cytokines which contribute to activation and tolerance that a little bit more background should be given.In conclusion, I think this is a very valuable piece of work of great interest to the rapidly proliferating and emerging population of cancer specialists who are slowly becoming aware that the immune system is extremely important in the management of this disease, a concept that has been ignored for the last few decades and foreign to most oncologist practising today. However, to make it more impactful and a 'must read' article it does require considerable focusing on the aspects raised by the referees and tighter structure, as it does seem to read rather like a speculative lecture in its current format.With regards to referencing, there is a referral to Klazmann et al. and there is no Klazmann et al. in the references and this is a further example that the references need to be very carefully looked into, although a paper where he is the senior author is listed, I do not think that it is appropriate therefore to refer to it in the text as Klazmann et al.",
"responses": []
}
] | 1
|
https://f1000research.com/articles/4-552
|
https://f1000research.com/articles/9-161/v1
|
03 Mar 20
|
{
"type": "Research Article",
"title": "Growth, production and feed conversion performance of the gurami sago (Osphronemus goramy Lacepède, 1801) strain in different aquaculture systems",
"authors": [
"Azrita Azrita",
"Netti Aryani",
"Ainul Mardiah",
"Hafrijal Syandri",
"Azrita Azrita",
"Netti Aryani",
"Ainul Mardiah"
],
"abstract": "Background: Giant gourami (Osphronemus goramy, Osphronemidae), belonging to the gurami sago strain, is an important economic fish species that was newly released for domestication in 2018 in Indonesia. The present study aimed to determine the growth, production and feed conversion efficiency of gurami sago strain in different aquaculture systems. Methods: A mean of 240 juveniles were stocked (mean, 54.53 g and 13.88 cm) into concrete ponds, floating net cages and earthen freshwater ponds (12 m3) with three replicates of each. The juveniles were fed a floating commercial pellet diet containing 30% crude protein and 5% crude lipids. Feed was supplied at 3% of fish biomass per day throughout the 90 days of the experiment. The research was conducted in the area surrounding Lake Maninjau of West Sumatera Province, Indonesia. Results: After 90 days, the mean weight of fish reared in concrete ponds was 166.86 g, floating net cages was 179.51 g and earthen freshwater ponds was 149.89 g. The mean final biomass was 37.64 kg for concrete ponds, 41.27 kg for floating net cages, and 33.72 kg for earthen freshwater ponds. The specific growth rates (%/day) for concrete ponds, floating net cages and earthen freshwater ponds were 0.67, 0.75 and 0.62, respectively. The feed conversion rates were 1.45 for concrete ponds, 1.30 for floating net cages and 1.87 for earthen freshwater ponds. The net yields (kg mˉ3) were 2.05 for concrete ponds, 2.27 for floating net cages, and 1.73 for earthen freshwater ponds. The exponents (b) of the length–weight relationship were calculated for concrete ponds (1.0146), floating net cages (1.2641), and earthen freshwater ponds (1.0056). Conclusion: The study showed that the growth performance, production and feed conversion efficiency of the gurami sago strain could be considered a new candidate strain for floating net cage aquaculture in the future.",
"keywords": [
"Giant gourami",
"aquaculture systems",
"juveniles",
"growth",
"environment factors"
],
"content": "Introduction\n\nAquaculture activities are responsible for the supply of fish for human consumption. To meet the demand for food from aquaculture production, competition uses natural resources, such as land and water1–3. Many studies state that aquaculture production depends on many factors, including its species, aquaculture system, technical efficiency, production inputs, and infrastructure4–6.\n\nIn 2018, the total aquaculture production in Indonesia was 16,032,122 metric tons, of which 3,374,924 metric tons (21.05%) were obtained from freshwater fish culture, 9,884,670 metric tons (61.65%) were obtained from marine water fish culture, and 2,772,528 metric tons (17.29%) were obtained from brackish water fish culture7. Meanwhile, the species of freshwater aquaculture commodities that have been developed in Indonesia are Nile tilapia, Clarias catfish, Pangasius catfish, common carp, and giant gourami. These species have contributed 37.93%, 33.35%, 12.38%, 9.28% and 6.96% of the total freshwater aquaculture production, respectively7.\n\nAlthough the contribution from giant gourami was lowest (6.96%), the local gurami tambago and gurami galunggung strains have been cultured intensively8,9. The giant gourami that belongs to the local gurami sago strain has never been cultured intensively. This species is the result of newly released domesticated fish in 201810, which is still limited in West Sumatera Province of Indonesia11. Gurami sago is an herbivorous species that can consume a variety of plants, such as sente leaves (Alocasia macrorrhiza), kale (Brassica oleracea), cassava leaves (Manihot esculenta), and other young terrestrial plants. In addition, this species can eat commercial pellets and tolerate crowded aquaculture production systems, such as earthen freshwater ponds and synthetic sheet ponds9,11,12.\n\nThe gurami sago strain has been detected as candidate species for middle-scale commercial culture in Indonesia11. This species grows well in nursery ponds and reach a market size of 200 to 300 g per fish and a size of 50 to 100 g per fish as ornamental fishes. This characteristic creates commercial interest as a new species in an effort to develop freshwater fish farming in the future. Concrete ponds and floating net cages were options in the development of gurami sago culture. Many studies have found that continuous water flow systems in concrete ponds, synthetic sheet ponds, tanks, canvas tanks, pens and many other systems could be an alternative for fish aquaculture because these systems provide a high degree of control that can allow for high production5,12–16.\n\nIn the last decade, cage systems have received more attention from both researchers and producers. Fish farming in cages can be practiced intensively17,18. High production can be achieved at a low cost19,20. Fish farming in cages can achieve maximum growth with a high survival rate18,21,22. However, cage fish farming has advantages and disadvantages that must be considered before choosing a production system. The main disadvantages of fish farming in the floating net cages of lakes are that they are not ideal for land use and may cause massive fish deaths23,24. Meanwhile, the advantages of floating net cage aquaculture include high water circulation, solid waste not accumulating near cages, low water quality variation, and no electrical power required for water aeration18,21,22,25,26.\n\nFish production systems in many countries use a variety of methods, e.g., carp in earthen freshwater ponds27, giant gourami in earthen freshwater ponds and synthetic sheet ponds9,12. Nile tilapia in the ponds and cages28, and golden pompano in the floating cages29. Because the rearing of the gurami sago strain is relatively new, there are no parameters or best methods available to predict the growth performance, survival and feed conversion efficiency in a commercial rearing system. Therefore, knowledge about the contribution of gurami sago to each aquaculture system is very important to analyze. The current study was conducted to assess the growth performance, production, economic food conversion rate and waste load of feed of gurami sago strains in different aquaculture systems namely, concrete ponds, floating net cages and earthen freshwater ponds.\n\n\nMethods\n\nThere are no required permits from the government of the Republic of Indonesia to culture the gurami sago (O.goramy) strain in this study in concrete ponds, floating net cages and earthen freshwater ponds in the area surrounding Lake Maninjau of West Sumatera Province of Indonesia. The study was founded by LPPM (Research and Community Service) University of Bung Hatta under the Indonesia Endowment Fund for Education, Ministry of Finance, Republic of Indonesia, through the competitive grants scheme called the Productive Innovative Research (Policy/Governance) 2019 with the contract number PRJ-99/LPDP/2019. This grant included ethical approval and permits to collect fish samples including permission to rear this species. The animals used in this study did not suffer during the experiment. Gurami sago was transported to concrete ponds, floating net cages and earthen freshwater ponds for rearing for 90 days, fed commercial pellets and measured for growth performance every 30 days. At the end of the experiment, the gurami sago were still in good condition.\n\nThe study was conducted at the Research Center of Faculty of Fisheries and Marine Science, Bung Hatta University located in the area of Lake Maninjau, Koto Malintang village, Tanjung Raya sub-district, District Agam of West Sumatera Province, Indonesia. The geographical coordinates were S:00º12'26.63\"-S:00º25'02.80\" and E:100º07'43.74\"-E:100º 16'22.48\" and the altitude was 461 m above sea level. At the location, concrete ponds, earthen freshwater ponds and floating net cages were available.\n\nEach concrete pond has a size of 4×2 m, a depth of 1.5 m and a volume of 12 m3. It has 50 mm of middle drainage, which is covered with a net of 0.5 cm mesh to prevent juveniles from escaping and predators from entering. The water was pumped from borehole wells at a velocity of 5 litres per minute.\n\nEach floating net cage has a size 4 × 2 m, a depth of 1.5 m and a volume of 12 m3, and these cages were built from resistant PVC plastic. Each cage was constructed using a monofilament net with 10 mm mesh. The floating net cages were set up in Lake Maninjau near the fish farm (maximum depth of 9 m and an average water current of 25 cm per sec). The surface of the floating net cages was covered with nets stretched (25 mm mesh) to avoid bird predators.\n\nEach earthen freshwater pond has a size of 4 × 2 m, a depth of 1.5 m and a volume of 12 m3. It had 50 mm of central drainage and was covered with a net of 0.5 cm mesh to prevent fish jumping and predator entry during the rearing activity. The water was pumped from wells at a velocity of 5 litres per minute.\n\nThe experiment ran for 90 days beginning on 01 April and ending on 29 June 2019. Approximately 3,000 gurami sago juveniles weighing approximately 50 g were obtained from a hatchery in the Luhak sub-district in the district of Lima Puluh Kota. Fish were acclimatized with 1000 juveniles per each pond (concrete pond, floating net cages and earthen freshwater pond). Fish were acclimatized to the floating net cages (5 × 5 × 3 m) for one month prior to the experiment. In the initial growth phase, three concrete ponds, three floating net cages and three earthen freshwater ponds of 12 m3 (three replicates) were stocked with 240 juveniles each, with a density of approximately 20 fish/m3. The average initial weights and lengths of juveniles were 54.51±0.45 g and 13.81±0.02 cm (mean ± SD), respectively. The length was measured using a ruler with an accuracy level of 0.1 cm. The weight of each individual was measured with an electronic balance (OHAUS, Model CT 1200-S, USA).\n\nFish were fed twice daily (09:00 AM and 17:00 PM) with commercial floating pellet feed (JapfaComfeed Indonesia Ltd; 30% crude protein, 5% crude lipids, 6% crude ash and 13% crude fibre)18. The amount of feed provided was as much as 3% per day based on fish biomass during the experiment. Every 30 days, samples were taken from ponds to monitor fish growth and to adjust the feed amount. Twenty-four fish samples were obtained from each concrete pond, floating net cage and earthen freshwater pond. Fish were captured at 07.00 AM with gillnets, which have a net bag with a suitable mesh size. Then, fish were anaesthetized orally with tricaine methanesulfonate (MS-222, ethyl 4-aminobenzoate methanesulfonate 98%, Sigma Aldrich Co, USA, MO; 50 mg Lˉ1), based on the dosage used for Hemibagrus wyckii30.\n\nWater parameters were recorded monthly in the concrete ponds, floating net cages and earthen freshwater ponds. The water temperature (⍜C) and dissolved oxygen (DO; mg Lˉ1) were measured with an oxygen metre (YSI model 85). The pH values were determined using a pH metre (Digital Mini-pH Metre, 0-14PH, IQ Scientific, Chemo-science (Thailand) Co., Ltd, Thailand). The levels of ammonia (NH3; mg.Lˉ1), nitrite-nitrogen (NO2-N; mg Lˉ1), nitrate-nitrogen (NO3-N; mg Lˉ1), chemical oxygen demand (COD; mg Lˉ1), biological oxygen demand (BOD5; mg Lˉ1), alkalinity (mg Lˉ1), hardness (mg Lˉ1), total dissolved solids (TDS; mg Lˉ1) and total suspended solids (TSS; mg Lˉ1) were measured in each aquaculture system with replication according to standard procedures31. The nets of the floating cages were cleaned routinely to maintain water circulation in the fish rearing areas. The walls of the floating net cages were cleaned by divers in the water.\n\nThe gurami sago were reared for 90 days, and the survival rate was estimated by checking the aquaculture systems every day and recording the results. Dead fish were removed immediately. The survival rate percentage was calculated by subtracting the number of dead fish from the initial number of the stock. The parameters were analyzed according to Aryani et al.8, Kibra and Haque27 and Mokoro et al.32 with the following equations:\n\nAbsolute growth rate (AGR; g dayˉ1) or (Wt-Wi)/t, where Wt = final weight, Wi = initial weight, and t = time (day);\n\nSpecific growth rate (SGR, % dayˉ1) = (lnW1-lnW2/t × 100)\n\nGross yield (kg mˉ3) = total number of fish at harvest × average final weight/cage capacity\n\nNet yield (kg mˉ3) = (harvested biomass - stocked biomass/cage capacity)\n\nFeed conversion efficiency (FCE) = [fish weight gain (g)/total feed ingested (g)]\n\nApparent feed conversion rate (AFCR) = supplied feed/increase fish weight\n\nEconomic AFCR = cost/kg of fish weight × feed cost\n\nWaste load of feed = [feed intake (kg)] – [final biomass (kg)]\n\nFor each aquaculture system, the final total length (cm) and final total weight (g) were used to determine the relationship of W = aLb, where W is the total wet weight (g), L is the total length (cm) and a and b are variables of the length–weight relationships (LWRs) equations. These variables were estimated by the least square regression method. A t-test was used for comparison of the b values obtained in the linear regressions with the isometric value by equation33: ts = (b – 3)/Sb, where ts is the t-test value, b is the slope and Sb is the standard error of the slope (b). The comparison of the obtained values of the t-test with the respective table critical values allowed for the determination of whether the b values were statistically significant as well as their inclusion in the isometric range (b=3) or allometric range (negative allometric; b<3 or positive allometric; b>3). The degree of correlation between the variables was computed to determine the coefficient, R2. Fulton’s condition index was calculated as K=100(W/L3)33, where K = Fulton’s condition index, W = weight, and L= length.\n\nThe data were analyzed using SPSS software (version 16.0 for Windows; SPSS Inc., Chicago, IL). Kolmogorov-Smirnov statistics were used to test data normality. Then, Levine’s test was used to analyse the absolute residuals from homogeneity. One-way ANOVA was used to analyze the effect of each treatment, followed by post hoc Duncan's multiple range tests34. The 95% confidence level (p<0.05) was considered as the threshold to identify significant differences. All means are given with ± standard deviation (±SD). The canonical discriminant functions were used to analyze the water quality grouping between rearing systems.\n\n\nResults\n\nThe overall survival rate of fish in different aquaculture systems was greater than 89.44%. The culture system had a significant effect (p<0.05) on the mean final body weight (g), final biomass (kg), weight gain (g), gross yield (kg m ˉ3), net yield (kg m ˉ3), absolute growth rate (g day ˉ1), specific growth rate (% day ˉ1), AFCR, and economic food conversion rate (US$/kg gain) after 90 days of culture (Table 1). In contrast, the culture system did not significantly (p>0.05) affect the mean final total length, feed intake (kg) or Fulton’s K. The economic AFCRs were US$1.45 for concrete ponds, US$1.30 for floating net cages and US$1.87 for earthen freshwater ponds.\n\nWithin a row, means followed by different letters are significantly different (p<0.05). TL: total length. *USD 1.00 = IDR 14,350.\n\nDuring the 90 days of the experiment, the gurami sago reared in floating net cages grew faster than those reared in concrete ponds and earthen freshwater ponds (Figure 1). At the end of the experiment, the fish reared in the floating net cages had a larger size distribution than that of the fish reared in the concrete ponds and earthen freshwater ponds throughout the 90 day trial (Figure 2). The mean final body weights of the gurami sago reared in concrete ponds, floating net cages and earthen freshwater ponds were 166.86 g, 179.51 g, and 149.89 g, respectively. The net yield was 2.05 kg mˉ3 for concrete ponds, 2.27 kg mˉ3 for floating net cages and 1.73 kg mˉ3 for earthen freshwater ponds during the 90 days of rearing. The FCE and waste load at 90 days of culture were significantly (p<0.05) affected by the different rearing systems. A summary of the FCR, FCE and waste load feed from the five aquaculture species is presented in Table 2.\n\nThe FCE for giant gourami culture is 0.77 (1.0 kg feed fish results in 0.77 kg of fish). This value suggests that the waste load is 0.23 kg (1.0 kg feed – 0.77 kg fish). The above calculation can be applied to other species. FCR, feed conversion rate; FCE, feed conversion efficiency.\n\nThe growth rates of gurami sago based on body weight were described according to the following exponential equation: W = 60.875e0.0498t (with R2 = 0.83) for the concrete pond, W = 48.580e0.0613t (with R2 = 0.75) for the floating net cage and W = 55.7050e0.0623t (with R2 = 0.75) for the earthen freshwater pond. The length-weight relationships for the gurami sago reared in concrete ponds were shown by W = 7.9368L1.0146 (with R2 = 0.83, Figure 3) and by W = 3.7760L1.2641 (with R2 = 0.75, Figure 4) for the floating net cages and by W = 9.3106L1.0056 (with R2 = 0.75, Figure 5) for the earthen freshwater ponds. The three b-values of each aquaculture system differed from 3.0 (b<3, p<0.05) indicating negative allometric growth. The Fulton’s condition index in the concrete pond, floating net cages and earthen freshwater pond were 2.45, 1.91, and 3.36, respectively.\n\nEach point represents one sampled fish (N=24).\n\nEach point represents one sampled fish (N=24).\n\nEach point represents one sampled fish (N=24).\n\nIn this study, the water quality from each aquaculture system during the experiment period showed significant differences (p<0.05) in terms of TDS, TSS, DO, COD, BOD, ammonia, nitrites, nitrates, pH, alkalinity and hardness, only water temperature did not show a significant difference. Furthermore, in the principal component analysis, PC1 accounted for 83.33% of the 12 parameters of water quality, which had a positive correlation with all water quality parameters. This result shows that value has an effect on the water quality parameters in aquaculture systems. Alkalinity, hardness, pH, and dissolved oxygen make high contributions to the aquaculture system (Table 3). The plot of PC1 and PC2 shows highly isolated water quality parameters between concrete ponds, floating net cages and earthen freshwater ponds (Figure 6).\n\nExtraction Method: Principal component analysis (PCA).\n\n\nDiscussion\n\nThe aquaculture industry needs environmentally friendly aquatic ecosystems. Therefore, aquaculture practices must use aquaculture systems that minimize waste loads and increase added value1,3,38,39. In fact, the diversification of aquaculture systems with the efficient use of land resources can increase aquaculture production28,40.\n\nThese culture comparisons of concrete ponds, floating net cages and earthen freshwater pond systems are relevant to determine their relative per unit volume performance for juvenile rearing of gurami sago and to recommend an alternative to diversify aquaculture and contribute to the development of commercial production in the future. The comparisons between concrete ponds, floating net cages and earthen freshwater ponds were relevant to determine their performance per unit volume of aquaculture system. The rearing of gurami sago is an alternative diversity of aquaculture that can contribute to the development of commercial production in the future.\n\nGurami sago was successfully reared in concrete ponds, floating net cages and earthen freshwater ponds. However, their growth performance was best in the floating net cages. The high survival rate of gurami sago was found in the floating net cages, which was similar to the gurami tambago strain8 and gurami sago in the synthetic sheet ponds12. On the other hand, the survival rates of gurami sago in earthen freshwater ponds (89.44%) were higher than those of carps (65.74%) and stinging catfish (69.00%) in freshwater ponds27.\n\nThe growth rate of gurami sago, with an average initial weight of 54.18 g, was faster in floating net cages than in concrete ponds and earthen freshwater ponds, with specific growth rate (SGR, % day-1) values of 0.67, 0.75 and 0.62, respectively. In contrast, Budi et al.41 stated that giant gourami belonging to the local gurami soang strain in the laboratory with initial weight of 15.83 g had faster growth with an SGR value of 2.13% dayˉ1. The specific growth rate of fish seems to be influenced by the initial weight, strains and aquaculture systems. The economic AFCR value of fish fed in floating net cages was lower than that of fish fed in concrete ponds and earthen ponds. Therefore, it can reduce the cost of feed and increase the economic benefits to producers. This condition indicates that the culture of gurami sago in floating net cages gives fish a chance to consume more feed. However, this AFCR was lower than that of Nile tilapia42,43, and giant gourami8, and higher than the African catfish AFCR value44.\n\nIn this study, the growth performance of different gurami sago individuals in each aquaculture system was caused by differences in water quality. The PCA shows that there are differences in water quality among concrete ponds, floating net cages and earthen freshwater ponds. The alkalinity, hardness, and pH might affect the growth performance of gurami sago in aquaculture systems. Boyd et al.45 stated that the productivity of aquatic ecosystems and aquaculture production can be influenced by water quality, such as alkalinity, hardness and pH. Many studies have found that growth performance can be affected by water temperature46,47, DO level48 and nitrite-nitrogen27.\n\nThe aquaculture system influences the production of gurami sago. The highest production was found in the floating net cages, with a value of 3.36 kg mˉ3. However, its production was lower than that of other freshwater cages, for example 4.19 to 10.70 kg mˉ3 for the strain gurami tambago (O. goramy)8, 25.4 to 26.3 kg mˉ3 for pirarucu (Arapaima gigas)49, 88.5 kg mˉ3 for silver perch, (Bidyanus bidyanus)50 and 11.60 to 16.03 kg mˉ3 for spotted rose snapper (Lutjanus guttatus)36. It seems that different levels of aquaculture production can be influenced by species diversity, stocking density and duration of aquaculture. Giant gourami can produce a maximum profit after 324 days of aquaculture51.\n\nHerein, we recommend gurami sago strain aquaculture in concrete ponds, floating net cages and earthen freshwater ponds for 324 days. According to De Oliveira Continho et al,52 fish reared in cages can increase the variation in weight production. In contrast, the freshwater cages have been marred by increasing the frequencies of fish mortality, causing negative implications to finances and the environment23,24,53. Bosma and Verdegem54 reported that the direct risks related to aquaculture in ponds were habitat destruction, suboptimal freshwater consumption, organic pollution, eutrophication, and water contamination with pesticides. These factors can cause production to decline and cause low economic value.\n\nIn this study, after the analysis of growth performance and production, we also analyzed the length–weight relationship and condition factor (K) from aquaculture systems. The exponent of the length–weight relationship - or per Froese55, the allometric coefficient (b) - calculated was 1.0146 for concrete ponds, 1.2641 for floating net cages and 1.0056 for earthen freshwater ponds. Gurami sago grown in different aquaculture systems showed negative allometric growth. These values were smaller than 2.94 for the culture of Tilapia zillii56 and 2.99 and 2.93 for Pangasianodon hypophthalmus and Clarias gariepinus, respectively57. The K-values were not different among concrete ponds, floating net cages and earthen freshwater ponds. The finding explains that no different morphological factors were found in gurami sago cultures in concrete ponds, floating net cages and earthen freshwater ponds. However, cultures of gurami sago in floating net cages had a smaller condition factor or had values close to 1.00. The variation in the condition factor (K) of gurami sago may be influenced by different factors, such as environmental conditions, feed intake and increased of body weight. The condition factor (K) of fish depends on many factors, including species diversity, growth, physiological performance, age, and gonadal maturity14,56,58–60.\n\n\nConclusion\n\nIn conclusion, our study showed that the gurami sago strain can be efficiently reared in floating net cages, concrete ponds and earthen freshwater ponds. However, gurami sago showed better growth performance and feed conversion efficiency when reared in floating net cages than when reared in concrete and earthen freshwater ponds. Nevertheless, concrete ponds are technically feasible alternatives for the optimal production of gurami sago based on the specified size of the pond. Additional research should focus on determining the duration of culture of gurami sago in floating cages and concrete ponds to increase the production of fish with the desired size to market demand.\n\n\nData availability\n\nFigshare: Row data growth performance of gurami sago in different aquaculture systems.doc, https://doi.org/10.6084/m9.figshare.11719542.v161.\n\nThis project contains the following underlying data:\n\n– Table 1. Sample size of weight and length of the gurami sago strain (0 days, 30 days, 60 days and 90 days) in the concrete pond culture (N=24)\n\n– Table 2. Sample size of weight and length of the gurami sago strain (0 days, 30 days, 60 days and 90 days) in the floating net cage culture (N=24)\n\n– Table 3. Sample size of weight and length of the gurami sago strain (0 days, 30 days, 60 days and 90 days) in the earthen freshwater pond culture (N=24)\n\n– Table 4. Sample size means of initial weight, final body weight and weight gain of gurami sago (N=24)\n\n– Table 5. Sample size means of initial length, final total length and length increase of gurami sago (N=24)\n\n– Table 6. Data on mean initial biomass, final biomass and gross yield of gurami sago (N=24)\n\n– Table 7. Data on mean SGR, feed intake and apparent feed conversion rate of gurami sago (N=24)\n\n– Table 8. Data on mean economic food conversion, feed conversion efficiency and waste load of feed (N=24)\n\n– Table 9. Data on mean growth (g) of gurami sago at 0 days, 30 days, 60 days, and 90 days (N=24)\n\n– Table 10. Data on mean size distribution (g) of gurami sago in the different aquaculture systems in the 90-day trial (N=72).\n\n– Table 11. Row data for water quality parameters of reared gurami sago in different aquaculture systems for each month.\n\nFigshare: Row Data_survival (fish) of gurami sago_12 Feb 2020.doc, https://doi.org/10.6084/m9.figshare.11845560.v162\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Acknowledgements\n\nThe authors thank the Director of Indonesia Endowment Fund for Education, and the Ministry of Finance, Republic of Indonesia for supporting this study through the competitive grants scheme Productive Innovative Research (Policy/Governance) 2019. We appreciate all of the students, fish farmers and partners who helped the author during data collection in the field.\n\n\nReferences\n\nAhmad N, Thompson S: The blue dimensions of aquaculture: A global synthesis. Sci Total Environ. 2019; 652: 851–861. PubMed Abstract | Publisher Full Text\n\nZhe Z, Di W, Wenxia T, et al.: Extracting aquaculture ponds from natural water surfaces around inland lakes on medium resolution multispectral images. Int J Appl Earth Obs Geoinformation. 2019; 80: 13–25. Publisher Full Text\n\nDauda AD, Ajadi A, Tola-Fabunmi AS, et al.: Waste production in aquaculture: Sources, components and managements in different culture systems. Aquaculture and Fisheries. 2019; 4(3): 81–88. Publisher Full Text\n\nJia B, St-Hilaire S, Singh K, et al.: Farm-level returns and costs of yellow catfish (Pelteobagrus fulvidrago) aquaculture in Guangdong and Zhejiang provinces China. Aquaculture Reports. 2016; 4: 48–56. Publisher Full Text\n\nIliyasu A, Mohamed ZA, Terano R: Comparative analysis of technical efficiency for different production culture systems and species freshwater aquaculture in Peninsular Malaysia. Aquaculture Reports. 2016; 3: 51–57. Publisher Full Text\n\nOpiyo MA, Marijani E, Muendo P, et al.: A review of aquaculture production and health management practices of farmed fish in Kenya. Int J Vet Sci Med. 2018; 6(2): 141–148. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCDSI (Central Data System Information): Ministry of Marine and Fisheries Republic of Indonesia. (In Indonesian). 2018.Reference Source\n\nAryani N, Azrita, Mardiah A, et al.: Influence of feeding rate on the growth, feed efficiency and carcass composition of the Giant gourami (Osphronemus goramy). Pak J Zool. 2017; 49(5): 1775–1781. Publisher Full Text\n\nPouil S, Samsudin R, Slembrouck J, et al.: Nutrient budgets in a small-scale freshwater fish pond system in Indonesia. Aquaculture. 2019; 504: 267–274. Publisher Full Text\n\nThe Ministry of Marine and Fisheries the Republic of Indonesia: Decision of the Ministry of Marine and Fisheries the Republic of Indonesia Number 56 / KEPMEN-KP/ 2018 on Release of Gurami Sago. (In Indonesian). 2018.Reference Source\n\nAzrita, Syandri H: Effects of salinity on survival and growth of Gurami Sago (Osphronemus goramy) juveniles. Pak J Biol Sci. 2018; 21(4): 171–178. PubMed Abstract | Publisher Full Text\n\nSyandri H, Mardiah A, Azrita, et al.: Effect of stocking density on the performance of juvenile Gurami sago (Osphronemus goramy) in the synthetic sheet pond. Pakistan Journal of Zoology. 2020; 52(2): 717–726. Publisher Full Text\n\nLiu W, Tan H, Chen W, et al.: Pilot study on water quality regulation in a recirculating aquaculture system with suspended growth bioreactors. Aquaculture. 2019; 504: 396–403. Publisher Full Text\n\nYu ZL, Hun N, Yang MJ, et al.: Environmental water flow can boost foraging success of the juvenile rapa whelk Rapana venosa.(Muricidae) in aquaculture tanks with still or flowing water: Indication of chemosensory foraging. Aquaculture. 2019; 513: 734392. Publisher Full Text\n\nPedrosa RU, de Mattos BO, Coasta DSP, et al.: Effects of feeding strategies on growth, biochemical parameters and waste excretion of juvenile arapaima (Arapaima gigas) raised in recirculating aquaculture systems (RAS). Aquaculture. 2019; 500: 562–568. Publisher Full Text\n\nWik TEI, Lindén BT, Wramner PI: Integrated dynamic aquaculture and wastewater treatment modelling for recirculating aquaculture systems. Aquaculture. 2009; 287(3–4): 361–370. Publisher Full Text\n\nBallester-Molto M, Sanchez-Jerez P, Cerezo-Valverde J, et al.: Particulate waste outflow from fish-farming cages. How much is uneaten feed? Mar Pollut Bull. 2017; 119(1): 23–30. PubMed Abstract | Publisher Full Text\n\nSyandri H, Azrita, Mardiah A: Nitrogen and phosphorus waste production from different fish species cultured at floating net cages in Lake Maninjau, Indonesia. Asian J Sci Res. 2018; 11(2): 287–294. Publisher Full Text\n\nMusinguzi L, Lugya J, Rwezarula P, et al.: The extent of cage aquaculture, adherence to best practices and reflections for sustainable aquaculture on African inland waters. J Great Lakes Res. 2019; 45(6): 1340–1347.Publisher Full Text\n\nHalide H, Stigebrandt A, Rehbein M, et al.: Developing a decision support system for sustainable cage aquaculture. Environmental Modelling & Software. 2009; 24(6): 694–702. Publisher Full Text\n\nLindim C, Becker A, Grüneberg B, et al.: Modelling the effects of nutrient loads reduction and testing the N and P control paradigm in a German shallow lake. Ecol Eng. 2015; 82: 415–427. Publisher Full Text\n\nGondwe MJ, Guildford SJ, Hecky RE: Carbon, nitrogen and phosphorus loadings from tilapia fish cages in Lake Malawi and factors influencing their magnitude. J Great Lakes Res. 2011; 37: 93–101. Publisher Full Text\n\nSyandri H, Azrita Junaidi, et al.: Levels of available nitrogen-phosphorus before and after fish mass mortality in Maninjau Lake of Indonesia. J Fish Aquat Sci. 2017; 12(4): 191–196. Publisher Full Text\n\nOpiyo MA, Marijani E, Muendo P, et al.: A review of aquaculture production and health management practices of farmed fish in Kenya. Int J Vet Sci Med. 2018; 6(2): 141–148. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNi Z, Wu X, Li L, et al.: Pollution control and in situ bioremediation for lake aquaculture using an ecological dam. J Clean Prod. 2017; 172: 2256–2265. Publisher Full Text\n\nVarol M: Impacts of cage fish farms in a large reservoir on water and sediment chemistry. Environ Pollut. 2019; 252(Pt B): 1448–1454. PubMed Abstract | Publisher Full Text\n\nKibria AS, Haque MM: Potentials of integrated multi-trophic aquaculture (IMTA) in freshwater ponds in Bangladesh. Aquacult Rep. 2018; 11: 8–16. Publisher Full Text\n\nHenriksson PJ, Tran N, Mohan C, et al.: Indonesian aquaculture futures-evaluating environmental and socioeconomic potentials and limitations. J Cleaner Prod. 2017; 162: 1482–1490. Publisher Full Text\n\nQi Z, Shi R, Yu Z, et al.: Nutrient release from fish cage aquaculture and mitigation strategies in Daya Bay, southern China. Mar Pollut Bull. 2019; 146: 399–407. PubMed Abstract | Publisher Full Text\n\nAryani N, Suharman I, Syandri H: Reproductive performance of asian catfish (Hemibagrus wyckii Bleeker, 1858), a candidate species for aquaculture [version 2; peer review: 2 approved]. F1000Res. 2018; 7: 683. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAPHA: Standard Methods for Examination of Water and Wastewater. 19th Edn. American Public Health Association, Washington DC., USA. 1995. Publisher Full Text\n\nMokoro A, Oyoo-Okoth E, Nguji CC, et al.: Effects of stocking density and feeding duration in cage-cum-pond-integrated system on growth performance, water quality and economic benefits of Labeo victorianus (Boulenger 1901) culture. Aquac Res. 2014; 45(10): 1672–1684. Publisher Full Text\n\nRicker WE: Computation and interpretation of biological statistics of fish population. Fisheries Research Board of Canada Bulletin. 1975; 191. Reference Source\n\nDuncan DB: Multiple range and multiple F tests. Biometrics. 1955; 11(1): 1–42. Publisher Full Text\n\nChiu A, Li L, Gou S, et al.: Feed and fish meal use in the production of carp and tilapia in China. Aquaculture. 2013; 414–415: 127–134. Publisher Full Text\n\nHernández C, Ibarra-Castro L, Hernández CH, et al.: Growth performance of Spotted Rose Snapper in floating cages and continuous water-flow tank Systems. N Am J Aquac. 2015; 77(4): 423–428. Publisher Full Text\n\nMungkung R, Joël A, Tri HP, et al.: Life Cycle Assessment for environmentally sustainable aquaculture management: a case study of combined aquaculture systems for carp and tilapia. J Clean Prod. 2013; 47: 249–256. Publisher Full Text\n\nSamuel-Fitwi B, Wuertz S, Schroeder JP, et al.: Sustainability assessment tools to support aquaculture development. J Clean Prod. 2012; 32: 183–192. Publisher Full Text\n\nValenti WC, Kimpara JM, Preto B de L, et al.: Indicators of sustainability to assess aquaculture systems. Ecol Indic. 2018; 88: 402–413. Publisher Full Text\n\nJayanthi M, Thirumurthy S, Muralidhar M, et al.: Impact of shrimp aquaculture development on important ecosystems in India. Glob Environ Change. 2018; 52: 10–21. Publisher Full Text\n\nBudi DS, Alimuddin, Suprayudi MA: Growth response and feed utilization of giant gourami (Osphronemus goramy) juvenile feeding different protein levels of the diets supplemented with recombinant growth hormone. HAYATI Journal of Biosciences. 2015; 22(1): 12–19. Publisher Full Text\n\nShrestha MK, Bhandari MP, Diana JS, et al.: Positive impacts of Nile tilapia and predatory sahar on carp polyculture production and profits. Aquaculture and Fisheries. 2018; 3(5): 204–208. Publisher Full Text\n\nMoniruzzaman M, Uddin KB, Basak S, et al.: Effects of stocking density on growth body composition, yield and economic return of monosex tilapia (Oreochromis niloticus L.) under cage culture system in Kaptai Lake in Bangladesh. Journal Aquaculture Research & Development. 2015; 6: 357. Publisher Full Text\n\nOké V, JurgensGoosen N: The effect of stocking density on profitability of African catfish (Clarias gariepinus) culture in extensive pond systems. Aquaculture. 2019; 507: 385–392. Publisher Full Text\n\nBoyd CE, Turker CS, Somridhivej B: Alkalinity and hardness: Critical but Elusive concepts in aquaculture. J World Aquac Soc. 2016; 47(1): 6–41. Publisher Full Text\n\nDesai AS, Singh RK: The effects of water temperature and ration size on growth and body composition of fry of Common carp, Cyprinus carpio. J Therm Biol. 2009; 34(6): 276–280. Publisher Full Text\n\nYuen JW, Dempster T, Oppedal F, et al.: Physiological performance of ballan wrasse (Labrus bergylta) at different temperatures and its implication for cleaner fish usage in salmon aquaculture. Biological Control. 2019; 135: 117–123. Publisher Full Text\n\nSun M, Hassan SG, Li D: Models for estimating feed intake in aquaculture: A review. Computers and Electronics in Agriculture. 2016; 127: 425–438. Publisher Full Text\n\nDe Oliveira EG, Pinheiro AB, de Olivera VQ, et al.: Effects of stocking density on the performance of juvenile pirarucu (Arapaima gigas) in cages. Aquaculture. 2012; 370–371: 96–101. Publisher Full Text\n\nRowland SJ, Mifsud C, Nixon M, et al.: Effects of stocking density on the performance of the Australian freshwater silver perch (Bidyanus bidyanus) in cages. Aquaculture. 2006; 253(1–4): 301–308. Publisher Full Text\n\nWijayanto D, Faik K, Ristiawan AN: Model of profit maximization of the giant gourami (Osphronemus goramy) culture. Omni-Akuatika. 2017; 13(1): 54–59. Publisher Full Text\n\nDe Oliveira Continho JJ, Neira LM, de Sandre LCG, et al.: Carbohydrate-to-lipid ratio in extruded diets for NILE tilapia farmed in net cages. Aquaculture. 2018; 497: 520–525. Publisher Full Text\n\nAzevedo PA, Podemski CL, Hesslein RH, et al.: Estimation of waste outputs by a rainbow trout cage farm using a nutritional approach and monitoring of lake water quality. Aquaculture. 2011; 311(1–4): 175–186. Publisher Full Text\n\nBosma RH, Verdegen MCJ: Sustainable aquaculture in ponds: Principles, practices and limits. Livest Sci. 2011; 139(1–2): 58–68. Publisher Full Text\n\nFroese R: Cube law, condition factor and weight-length relationships: history, meta-analysis and recommendations. J Appl Ichthyol. 2006; 22(4): 241–253. Publisher Full Text\n\nNehemia AD, Maganira JD, Rumisha C: Length-Weight relationship and condition factor of tilapia species grown in marine and fresh water ponds. Agri Biol J N Am. 2012; 3(3): 117–124. Publisher Full Text\n\nOkomoda VT, Koh ICC, Hassan A, et al.: Length-weight relationship and condition factor of the progenies of pure and reciprocal crosses of Pangasianodon hypophthalmus and Clarias gariepinus. AACL Bioflux. 2018; 11(4): 980–987. Reference Source\n\nGebremedhim S, Mingist M: Length-weight relationship, gonado somatic index and Fulton condition factor of the dominant fishes at Aveya river, Blue Nile Basin Ethiopia. J Fish Aquat Sci. 2014; 9(1): 1–13. Publisher Full Text\n\nMortuza MG, Al-Misned FA: Length-weight relationships, condition factor and sex-ratio of Nile Tilapia, Oreochromis niloticus in Wadi Hanifah, Riyadh, Saudi Arabia. World J Zool. 2013; 8(1): 106–109. Reference Source\n\nAryani N, Suharman I, Sabrina H: Length-weight relationship and condition factor of the critically endangered fish of Geso, Hemibagrus wyckii (Bleeker, 1858) Bagridae from Kampar Kanan River, Indonesia. J Entomol Zool Stud. 2016; 4(2): 119–122. Reference Source\n\nSyandri H, Azrita A, Aryani N, et al.: Row Data of Growth performance of Gurami Sago.doc. figshare. Dataset. 2020. http://www.doi.org/10.6084/m9.figshare.11719542.v1\n\nSyandri H, Azrita A, Aryani N, et al.: Row Data_survival(fish) of gurami sago_ 12 Feb 2020.doc. figshare. Dataset. 2020. http://www.doi.org/10.6084/m9.figshare.11845560.v1"
}
|
[
{
"id": "60880",
"date": "19 Mar 2020",
"name": "Simon Pouil",
"expertise": [
"Reviewer Expertise aquaculture",
"aquatic ecotoxicology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe present manuscript entitled \"Growth, production and feed conversion performance of the gurami sago (Osphronemus goramy Lacepède, 1801) strain in different aquaculture systems\" examines the zootechnical performances of a giant gourami Osphronemus goramy strain called “gurami sago” in three different rearing structures: concrete ponds, floating net cages and earthen freshwater ponds. Overall, the study presented is quite simple but nevertheless interesting especially because some important gaps of knowledge remain in giant gourami aquaculture. Thus, results presented, showing better growth performance and feed conversion efficiency when giant gourami is reared in floating net cages, are relevant to the field. Having said that, I think that the present version of the manuscript may be improved in some ways.\n\nMy main concerns are related to some methodological aspects as well as the presentation of the data.\n\nI addressed below some general comments regarding on the different sections of the manuscript. I hope the following comments help the authors in revising the manuscript.\n\nIntroduction: The description of the rationale of the study could be improved. I suggest to go straight to the point with a first paragraph explaining why giant gourami is important in Indonesian aquaculture and avoiding too general information. Authors may provide some production figures and explain, based on relevant references, that giant gourami is an emblematic local species with high practical and market value, omnivorous with a strong vegetarian component as thus, a candidate species for improving sustainability in aquaculture.\n\nAuthors should clearly state that, although giant gourami has been reared for decades in Indonesia, there are still important gaps of knowledge in its aquaculture.\n\nAnother point that is true throughout the manuscript: be careful in the use of “strain” and “species”. These two terms seem to be used as synonyms in the manuscript although they refer to different concepts. I am not sure that the focus done on the strain used is so important in the Introduction. I think that the results provided here are useful for the species itself and not only this specific strain.\n\nMethods: Overall, I found the Methods well-presented and informative enough. Nevertheless, I have one important concern regarding the statistics. Indeed, water parameters were recorded monthly, meaning that only 3 values per rearing structures are available to perform the canonical discriminant functions (CDF). Considering the variations of most of the measured parameters that can be occur in rearing structures such as shallow earthen ponds sometimes on the same day, I think such analysis is not appropriate.\n\nFurthermore, Authors should state why only 10% of the fish were sampled every month. I guess is because giant gourami is sensitive to handling but this information may be interesting to add.\n\nResults: As I already mentioned, I have some doubts regarding the validity of the CDF using water quality data.\n\nI believe that the presentation of the results can be improved. Authors should provide visible standard deviations values and statistical differences in the Figures when it is appropriate and better axis scale in order to improve data readability. For allometric relationships, p-values for model estimates should be provided.\nTable 1: “Final food conversion rate (Fulton’s K)”, I guess it should be changed by “Condition factor (Fulton’s K)”\nDiscussion: Although water quality is likely a key parameter to explain some of the observed differences in zootechnical performances among the rearing structures, unfortunately, since water parameter values were recorded only once a month, I think that there is not enough information provided to use these results.\n\nConclusion: In the concluding paragraph, I expected clear recommendations for giant gourami aquaculture based on the findings from this study.\n\nReferences: I believe that some references relating to the aquaculture of the giant gourami are missing. I suggest to consider the following references which can be useful in the Introduction and Discussion:\nFAO (2019) Cultured Aquatic Species Information Programme. Osphronemus goramy. Cultured Aquatic Species Information Programme. Text by Caruso, D., Arifin, Z.O., Subagja, J., Jacques Slembrouck, J. and New, M. In: FAO Fisheries and Aquaculture Department [online]. Rome. Updated 26 September 2019.\n\nArifin O.Z., Slembrouck J., Subagja J., Pouil S., Yani A., Asependi A., Kristanto A.H., Legendre M. (2020). New insights into giant gourami (Osphronemus goramy) reproductive biology and egg production control. Aquaculture 519: 734743.1\n\nKristanto A.H., Slembrouck J., Subagja J.,Pouil S., Arifin O.Z., Prakoso V.A., Legendre M. (2020). Egg and fry production of giant gourami (Osphronemus goramy): Rearing practices and recommendations for future research. Journal of the World Aquaculture Society 51: 119-138.2\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "5420",
"date": "16 Apr 2020",
"name": "Hafrijal Syandri",
"role": "Author Response",
"response": "Introduction:The introduction has been improving. The paragraph go straight to the point of why giant gourami important in Indonesian Aquaculture.Methods:Overall, the method has been revised based on your comment. We also add water quality data which come from our daily logbook. Actually water quality parameters were recorded weekly, but at the first we show in the Table 3 only per month for results Principal Component Analysis (PCA). Now we have changed it into weekly recorded (Table 3). Results:The results have been revised based on your comment. The allometric relationships (p-values) have been added in Figures 3, 4 and 5. Furthermore, Table 1 has been revised. Discussion:The water quality parameters have been revised based on your comment. We have changed it into weekly recorded. We have data from our daily logbook which already recorded during the research period. The complete raw data shows in Figshare (Table 11 revised). ConclusionWe have been revised the conclusion. References:We have been added some references based on your suggestion."
}
]
},
{
"id": "60878",
"date": "01 Apr 2020",
"name": "Peter Vilhelm Skov",
"expertise": [
"Reviewer Expertise Aquaculture nutrition and bioenergetics"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIntroduction:\n“Aquaculture activities are responsible for the supply of fish for human consumption. To meet the demand for food from aquaculture production, competition uses natural resources, such as land and water1–3. Many studies state that aquaculture production depends on many factors, including its species, aquaculture system, technical efficiency, production inputs, and infrastructure” - This is a rather convoluted paragraph. Please rephrase to improve coherence.\n\n“In 2018, the total aquaculture production in Indonesia was 16,032,122 metric tons” - That is probably correct, but then you go on to state that it all comes from fish culture. I think (supported by FAO 2018 yearbook) that around 12 million tons is seaweed. Please check and correct.\n\n”synthetic sheet ponds..” - I would prefer “artificial ponds lined with membranes” or something similar\n\nIt is not clear to me what is meant by ”middle-scale commercial culture”. Is this in relation to intensity? Please clarify.\n\nM&M:\n“Feed conversion efficiency (FCE) = [fish weight gain (g)/total feed ingested (g)] Apparent feed conversion rate (AFCR) = supplied feed/increase fish weight” - It is interesting that you have two indicators of feed performance where one is listed as apparent. I presume that feed waste was not collected in any of your rearing systems, and therefore all of your feed intake are apparent and based on “supplied feed”, also the FCE.\n\n“Waste load of feed = [feed intake (kg)] – [final biomass (kg)]” - I am not familiar with this variable, nor do I completely understand what it signifies, but presumably, it should be biomass gain, and not just final biomass?\n\nResults:\nOne of the things that can explain the observed differences in growth performance is likely to be your water quality parameters. While it is fine with the PCA plot, I would really like to see the water quality measurements in a table. Once these are available, perhaps it would be possible to discuss which water quality parameters would be essential to control to successfully produce gourami in land-based systems.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "5421",
"date": "16 Apr 2020",
"name": "Hafrijal Syandri",
"role": "Author Response",
"response": "Introduction: Introduction in first paragraph have been rephrase to improve coherences in the paragraph. We have omitted the data of Indonesian fisheries production in the introduction. “Synthetic sheet ponds” have been change to “artificial ponds lined with membranes”. “middle-scale commercial culture” meant “middle-scale farms” M & M We used the “Feed Conversion Efficiency (FCE)” Formula based on Chatvijitkul et al. 2017. Reference: Chatvijitkul, S., Boyd C.E., Davis, D.A., McNevin, A.A. (2017). Pollution potential indicators for feed-based fish and shrimp culture. Aquaculture 477: 43-49. We have been revised the formula of “Waste load of feed”. Please see in sub-bab “Measurement Parameters”. Results: In Table 1, the parameters of waste load /kg of feed were added The water quality parameters have been revised. We have data of water quality parameter in Table form, and it is has been state in raw data which have uploaded in Figshare (see Table 11 revised)."
}
]
}
] | 1
|
https://f1000research.com/articles/9-161
|
https://f1000research.com/articles/8-1874/v1
|
07 Nov 19
|
{
"type": "Software Tool Article",
"title": "Falco: high-speed FastQC emulation for quality control of sequencing data",
"authors": [
"Guilherme de Sena Brandine",
"Andrew D. Smith",
"Andrew D. Smith"
],
"abstract": "Quality control is an essential first step in sequencing data analysis, and software tools for quality control are deeply entrenched in standard pipelines at most sequencing centers. Although the associated computations are straightforward, in many settings the total computing effort required for quality control is appreciable and warrants optimization. We present falco, an emulation of the popular FastQC tool that runs on average three times faster while generating equivalent results. Compared to FastQC, falco also provides greater scalability for datasets with longer reads and more flexible visualization of HTML reports.",
"keywords": [
"FastQC",
"high-throughput sequencing",
"quality control"
],
"content": "Introduction\n\nHigh-throughput sequencing is routinely used to profile copy number variations in cancers1, assemble genomes of microbial organisms2,3, quantify gene expression4, identify cell populations from single-cell transcriptomes in a variety of tissues5 and track epigenetic changes in developing organisms and diseases6, among numerous other applications. New sequencing protocols are constantly being introduced7,8, and as the cost of sequencing per base decreases, sequencing data is growing in abundance, dataset size and read length9.\n\nWhen high-throughput sequencing data is generated it often undergoes common upstream analysis steps involving quality control (QC), adapter trimming, filtering contaminants and low-quality reads, and mapping reads to a reference genome or transcriptome. Excluding sequence assembly applications, read mapping should be the most computationally expensive step early in analysis pipelines. In comparison, the time and computation required for QC should be negligible. However, the efficiency of mapping algorithms has improving substantially over the past decade, while software for QC has received far less attention. As a consequence, the computation required for QC is appreciable, and can no longer be ignored when considering the total cost of sequencing.\n\nThe most commonly used tool for quality control of sequencing data is FastQC10, which, since 2011, has incorporated a wide range of QC checks covering multiple use cases. FastQC has been cited over 3,000 times, with citations increasing steadily since its introduction. Its analysis reports have become the standard for several QC tools, and automated analysis pipelines often rely on its evaluation as a safety criteria to proceed with downstream steps or, alternatively, to filter, trim or ultimately discard the data11,12.\n\nFastQC is implemented in a modular design, where multiple independent analysis procedures are run sequentially after an input record is read. This design allows new modules to be incorporated easily, but it implies that each analysis module is applied independently to each read, so the time required to process each read is the sum of the processing times for each module. If multiple modules use similar measurements, such as nucleotide content or average sequence quality, the same measurement will be calculated multiple times, causing the total analysis run time to increase.\n\nSeveral QC software tools have been introduced since FastQC, many focusing on speed improvements, more flexible module visualization, incorporation of paired-end reads and filtering sequences that failed QC checks. Despite proposing different alternatives to calculate and present QC results, the modules available in these tools are largely similar to FastQC’s (Table 1).\n\nAt the same time, FastQC’s analysis results are already part of many standard initial analysis pipelines. If a new QC software tool were to be incorporated in these pipelines, it is desirable that its results, and its output formats, remain consistent with those of FastQC.\n\nTo improve the speed of quality control while retaining the behaviour of FastQC, we developed FastQC Alternative Code (falco)13, an emulation of FastQC’s current analysis modules. We show that falco generates the same results as FastQC across a wide variety of datasets of different read lengths, sizes, file formats and library preparation protocols at significantly shorter running times. We also present example datasets from the public domain where FastQC fails to generate reports even when run on high-performance computing hardware, demonstrating that falco expands the range of possible cases in which these quality control metrics can be applied.\n\n\nMethods\n\nWe designed falco13 to faithfully emulate FastQC’s calculations, results and text reports. Our goal was to minimize the effort required to replace FastQC with falco in the context of larger automated analysis pipelines. We use the same set of command line arguments, configuration file names and formats. We also produce the same plain text format output, and the same report structure as FastQC, allowing users to take advantage of improved speed without adjusting to different program behaviors.\n\nThere are major differences between the implementations of falco and FastQC. While FastQC’s code emphasizes modularity in a way that allows for additional types of QC information to be added easily and uniformly,\n\nfalco’s design centralizes the function to read sequences from the input file and collects the minimum data necessary to subsequently create all modules after file processing. To ensure consistency with FastQC, we wrote each module’s source code based on FastQC’s implementation, adapting the portions that relate to sequence processing and maintaining the postprocessing functions that define how the collected data is used to generate summaries and reports.\n\nCompilation of falco requires a GNU GCC compiler version 5.0.0 (July 16, 2015; full support for the C++11 standard) or greater. Once installed, falco can be run on uncompressed files (FASTQ and SAM) without any additional dependencies. In order to process files in gzip compressed FASTQ and BAM formats, falco must be compiled with the ZLib14 and HTSLib15 libraries. The full documentation on how to compile, install dependencies and run the program is available in the README file in the falco repository.\n\n\nUse cases\n\nLike FastQC, falco13 can be applied to any sequencing data file (i.e. a file of sequenced reads) in the accepted formats. The only required command line argument is the path to the input file. Also like FastQC, a wide range of options can be provided if users only require a given subset of its analysis modules or outputs. The letters and symbols used for command line arguments were chosen to maintain consistency with FastQC’s options. As mentioned above, this choice is to facilitate integration with larger pipelines that already employ FastQC and depend on its behaviours.\n\nFalco can be run on a FASTQ format file named example.fastq with the following simple command:\n\n$ falco example.fastq\n\nThis will generate three files:\n\n1. fastqc_data.txt: The complete numerical values generated in each module’s individual analysis.\n\n2. fastqc_report.html: A visual page display of the text report’s data and plots generated in modules.\n\n3. summary.txt: A short summary indicating whether the input file passed or failed each module, and whether any warnings were raised.\n\nDefault configuration files are contained in a Configuration directory that is included with the program, but falco also allows users to manually define the thresholds for statistics to be considered a pass, warning or fail, the list of adapters to search for in reads and the list of contaminants to check overrepresented sequences by using configuration files in the same format used by FastQC.\n\n\nResults\n\nWe compared the output of falco13 to its FastQC counterpart using 11 datasets (Table 2). The tests consist of Illumina files originating from a range of different library preparation protocols for DNA, RNA and epigenetic experiments, as well as reads from the nanopore16 technology. For simplicity, Illumina paired-end datasets were only tested on their first read.\n\nFASTQ files available in the Sequencing Read Archive (SRA) were downloaded using the fastq-dump command from the SRA toolkit. We used the following flags when running fastq-dump: -skip-technical, -readids, -read-filter pass, -dumpbase, -split-3 and -clip. One dataset was downloaded from the Whole Human Genome Sequencing Project17.\n\nWe directly compared the text summary for each output of falco to FastQC’s output summary files, obtaining the same outputs (pass/warn/fail) for all tested criteria in all datasets.\n\nTo assess if falco’s output is consistent with FastQC’s format, we used the fastqcr18 R package version 0.1.2 and MultiQC11 version 1.7. Both tools can successfully parse the text reports generated by falco for the tested files. Differences in the fastq_data.txt files between the two programs result from choices for numerical precision output, or as a result of falco calculating certain averages based on more of the data within each file.\n\nSome alternative software tools exist for quality control of sequencing data, and users may opt for them due to their efficiency in cases where not all FastQC analysis modules are necessary. Among these, fastp19 has gained popularity for its speed and versatile set of options for trimming. fastp has demonstrated superior runtime to FastQC even when generating FASTQ format output files corrected by trimming adapters and filtering (which requires both input and output). HTQC20 is another tool that was developed with the intent to both improve speed performance and incorporate trimming functions after quality control. The two programs were used as benchmarks to compare with falco’s performance.\n\nAlthough most fastp modules are both calculated and displayed equivalently to FastQC, one major difference between these tools is how overrepresented sequences are estimated. fastp counts the sequences at every P reads (which users may specify), whereas FastQC stores the first 100,000 reads encountered for the first time, and subsequently checks if the following sequences match any of the stored candidates. This choice of implementation causes fastp’s runtime to greatly differ when over-representation is enabled. Conversely, FastQC’s runtime does not seem to be affected by disabling the overrepresented sequences module. For a comprehensive comparison between programs, we have measured the run times for our test datasets both with and without the overrepresented sequences module enabled. Programs were compared both in compressed (gzipped FASTQ) and uncompressed (plain FASTQ) file formats.\n\nFiles used to assess falco’s output comparison to FastQC (Table 2) were also used for speed benchmarking. Tests were executed in an Intel Xeon CPU E5-2640 v3 2.60GHz processor with a CentOS Linux 7 operating system. All file I/O was done using local disk to reduce variability in execution runtime. Programs were instructed to run using a single thread.\n\nFastQC version 0.11.8 was run with default parameters and the configuration limits, adapters and contaminants provided with the software. fastp version 0.20.0 was run with the -A, -G, -Q and -L flags to disable adapter trimming, poly-G trimming, quality filtering and length filtering, thus requiring the program to only perform QC checks without generating a new FASTQ file. When testing for overrepresented sequences, we set the -p flag to enable this module, and set the frequency of counts to the program’s default value of P = 20. We ran the ht-stat program on the tested files using the -S flag for single-ended reads. HTQC was not tested on gzip\n\ncompressed files as this file format is not accepted by the program. We used the time command (using the BASH shell keyword) to measure the total running times for each program, using the real time (total wall clock from program start to finish) as measurement. The benchmarking results (Table 3 and Table 4) show that falco performs faster than fastp and FastQC in all datasets, with an average 3x faster runtime than FastQC, both with the overrepresented sequences module on and off. Despite HTQC failing to process most test datasets due to unaccepted header formats, the two tests that ran to completion demonstrate that falco’s analysis times are also significantly smaller in comparison.\n\nAsterisks (*) indicate tests in which tools did not run to completion.\n\nNanopore sequencing is gaining popularity in genome assembly applications and as a low-cost protocol to quantify short reads26. Nanopore sequencers can generate reads of up to millions of bases, and assessing quality metrics for these datasets is fundamental to test for potential problems in quality or bias in specific regions of such long reads. While FastQC is capable of making summaries for protocols such as 45427 PacBio28, which generate sequences with around 10,000 bases per read, we have observed that it does not run to completion when given files with larger reads of over 100,000 bases. Files for which FastQC’s analysis does not finish are marked with an asterisk in Table 3 and Table 4. Falco successfully completes its analysis on these datasets, demonstrating that it can equally be used as a QC tool for longer reads.\n\nDespite FastQC’s clarity in its HTML reports, graphs are displayed as static images and have limited visualization flexibility, such as tile heatmaps not displaying raw deviations from average Phred scores in base positions or raw values in line plots not being visible. We have opted to display falco’s analysis results using the Plotly JavaScript library29, which allows interactive changes of axis labels, hovering on data points to visualize raw values and screenshots from specific position on the plot (Figure 1). This choice of presentation provides greater options to explore and interpret QC results while maintaining the visualization standards set by FastQC.\n\nLayout and plots are based on FastQC’s HTML report.\n\n\nConclusions\n\nFalco13 is a faster alternative to calculate the wide range of QC metrics generated by FastQC. It is entirely based on emulating the analysis modules FastQC provides while running faster than popular QC tools and generating dynamic visual summaries of analysis results. Both falco’s text and HTML outputs provide the same information generated by FastQC’s report, so tools that parse these files for custom visualization and downstream analysis can seamlessly incorporate falco into their pipeline.\n\n\nData availability\n\nDatasets used to compare Falco and FastQC are shown in Table 2. Guidance for how to accept accession wgs-FAB49164 is available from the Benchmark directory of the falco GitHub page.\n\n\nSoftware availability\n\nSource code for falco available at: https://github.com/smithlabcode/falco.\n\nThe scripts used to download files and reproduce the benchmarking steps described are also available in the same repository within the “benchmark” directory.\n\nArchived source code at time of publication: http://doi.org/10.5281/zenodo.352093313.\n\nLicense: GNU General Public License version 3.0.",
"appendix": "References\n\nAlkan C, Kidd JM, Marques-Bonet T, et al.: Personalized copy number and segmental duplication maps using next-generation sequencing. Nat Genet. 2009; 41(10): 1061–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLoman NJ, Quick J, Simpson JT: A complete bacterial genome assembled de novo using only nanopore sequencing data. Nat Methods. 2015; 12(8): 733–5. PubMed Abstract | Publisher Full Text\n\nMasella AP, Bartram AK, Truszkowski JM, et al.: PANDAseq: paired-end assembler for illumina sequences. BMC Bioinformatics. 2012; 13(1): 31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOzsolak F, Milos PM: RNA sequencing: advances, challenges and opportunities. Nat Rev Genet. 2011; 12(2): 87–98. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHan X, Wang R, Zhou Y, et al.: Mapping the Mouse Cell Atlas by Microwell-Seq. Cell. 2018; 172(5): 1091–1107.e17. PubMed Abstract | Publisher Full Text\n\nBuenrostro JD, Wu B, Chang HY, et al.: ATAC-seq: A Method for Assaying Chromatin Accessibility Genome-Wide. Curr Protoc Mol Biol. 2015; 109(1): 21.29.1–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDatlinger P, Rendeiro AF, Schmidl C, et al.: Pooled CRISPR screening with single-cell transcriptome readout. Nat Methods. 2017; 14(3): 297–301. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSpanjaard B, Hu B, Mitic N, et al.: Simultaneous lineage tracing and cell-type identification using CRISPR–Cas9-induced genetic scars. Nat Biotechnol. 2018; 36(5): 469–473. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSvensson V, Vento-Tormo R, Teichmann SA: Exponential scaling of single-cell RNA-seq in the past decade. Nat Protoc. 2018; 13(4): 599–604. PubMed Abstract | Publisher Full Text\n\nAndrews S, et al.: FastQC: a quality control tool for high throughput sequence data. 2010. Reference Source\n\nEwels P, Magnusson M, Lundin S, et al.: MultiQC: summarize analysis results for multiple tools and samples in a single report. Bioinformatics. 2016; 32(19): 3047–3048. Publisher Full Text\n\nBrown J, Pirrung M, McCue LA: FQC Dashboard: integrates FastQC results into a web-based, interactive, and extensible FASTQ quality control tool. Bioinformatics. 2017; 33(19): 3137–3139. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSena G, Smith A: smithlabcode/falco: 0.1 - 2019/10/28. 2019. http://www.doi.org/10.5281/zenodo.3520933\n\nDeutsch P, Gailly JL: Zlib compressed data format specification version 3.3. 1996. Publisher Full Text\n\nLi H, Handsaker B, Wysoker A, et al.: The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009; 25(16): 2078–2079. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJain M, Olsen HE, Paten B, et al.: The Oxford Nanopore MinION: delivery of nanopore sequencing to the genomics community. Genome Biol. 2016; 17(1): 239. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJain M, Koren S, Miga KH, et al.: Nanopore sequencing and assembly of a human genome with ultra-long reads. Nat Biotechnol. 2018; 36(4): 338–345. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKassambara A: fastqcr: Quality control of sequencing data. R package version 0.1.2. 2019. Reference Source\n\nChen S, Zhou Y, Chen Y, et al.: fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics. 2018; 34(17): i884–i890. Publisher Full Text\n\nYang X, Liu D, Liu F, et al.: HTQC: a fast quality control toolkit for Illumina sequencing data. BMC Bioinformatics. 2013; 14(1): 33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDecato BE, Lopez-Tello J, Sferruzzi-Perri AN, et al.: DNA Methylation Divergence and Tissue Specialization in the Developing Mouse Placenta. Mol Biol Evol. 2017; 34(7): 1702–1712. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang J, Zhang L, Jiang Z, et al.: TCF12 promotes the tumorigenesis and metastasis of hepatocellular carcinoma via upregulation of CXCR4 expression. Theranostics. 2019; 9(20): 5810–5827. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMacosko EZ, Basu A, Satija R, et al.: Highly Parallel Genome-wide Expression Profiling of Individual Cells Using Nanoliter Droplets. Cell. 2015; 161(5): 1202–1214. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNusse YM, Savage AK, Marangoni P, et al.: Parasitic helminths induce fetal-like reversion in the intestinal stem cell niche. Nature. 2018; 559(7712): 109–113. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBuenrostro JD, Giresi PG, Zaba LC, et al.: Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA-binding proteins and nucleosome position. Nat Methods. 2013; 10(12): 1213–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPrabakar RK, Xu L, Hicks J, et al.: SMURF-seq: efficient copy number profiling on long-read sequencers. Genome Biol. 2019; 20(1): 134. Publisher Full Text\n\nMargulies M, Egholm M, Altman WE, et al.: Genome sequencing in microfabricated high-density picolitre reactors. Nature. 2005; 437(7057): 376–80. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRhoads A, Au KF: PacBio Sequencing and Its Applications. Genomics Proteomics Bioinformatics. 2015; 13(5): 278–289. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSievert C, Parmer C, Hocking T, et al.: plotly: Create Interactive Web Graphics via ‘plotly. js’. 2017. Reference Source"
}
|
[
{
"id": "66327",
"date": "07 Jul 2020",
"name": "R Henrik Nilsson",
"expertise": [
"Reviewer Expertise Metabarcoding",
"molecular ecology",
"systematics",
"mycology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present a welcome addition to the flora of FastQC-style read processing packages. The fact that it is a drop-in replacement for FastQC is particularly nice.\n\nThe manuscript is a bit too short in my opinion. I miss some background information and some performance-related data. If the authors want to address a wide audience, they should probably work a bit more on the installation instructions and documentation too. The authors should probably also consider defining who their target audience is.\nIntroduction:\nAs a discretionary comment: the authors sometimes, but not always, use the Oxford comma (see example below). I wonder if this is something that should be streamlined. “sequencing data is growing in abundance, dataset size and read length9.” vs. “…adapter trimming, filtering contaminants and low-quality reads, and mapping reads to a reference genome or transcriptome.”\n\n“as a safety criteria” should probably be “as a safety criterion”.\n\nMethods:\nGood thinking behind the “We designed falco13 to faithfully…” paragraph. Nobody would be helped by yet another set of new file formats. A drop-in replacement is the way to go, if you ask me. And that is indeed what the authors deliver.\n\nI must make the observation, though, that the name “falco” may not be available (?): https://www.falcoseed.com/ca/article/cibus-registers-new-falco-brand-79k-canola-hybrid/ (some other more or less commercial uses of “Falco” can be found on https://en.wikipedia.org/wiki/Falco). I’m not sure how North American trademark/proprietary laws operate, but I suggest that the authors consult with the lawyers of their university. Better safe than sorry, right. (I consulted with our lawyers at one point, and they had me change the name of a software package we were working on at the time.)\n\nTwo unintended line breaks:\n“uniformly, falco’s design centralizes”\n\nResults:\nPlease cite the Sequence Read Archive formally; see Kodama et al. (20121).\n\nI like the reproducible nature of the “Results” section.\n\nTwo superfluous line breaks:\n\n“gzip compressed files.”\n\nConclusions:\nBoth “Falco” and “falco” are used in the manuscript. You’d think that the name would be fixed as either “Falco” or “falco”.\n\nSoftware availability:\nWhy not use the term “open source” at least once in the manuscript?\n\nMiscellaneous questions and observations:\nOut of curiosity: how does falco compare to Liu et al. (20192)?\n\nI’ve seen one software tool for quality-score-based trimming of sequences that actually reads the entire query file into memory, and then started to process it. This works less well as data files continues to grow, obviously. Is it worth pointing out that falco does not do this? What is, in fact, the maximum file size allowed by falco? Or is this dictated solely by the operating system?\n\nYou can produce some pretty funny behavior in some other tools for sequence QC/trimming by feeding them a file with a single FastQC entry in it, speaking of nothing. Is there, then, a minimum file size or number of query sequences for falco?\n\nUnless I’m mistaken, there are no fastq files available on https://github.com/smithlabcode/falco. I think the authors should make one or two available, so that it will be smooth and easy for the reader to try the software out. “The scripts used to download files and reproduce the bench-marking steps described are also available in the same repository within the “benchmark” directory.” comes across as somewhat indirect to me.\n\nWhat, exactly, are the hardware and software requirements for falco? This may be worth pointing out. Between the lines I read “any computer you can install the GNU GCC compiler on” – but not all readers will probably read it this way. Instead you’ll get the question: “Does it run on Windows 10?”.\n\nAlso, how much memory is needed to run falco? And how much memory is used up by falco when it processes a large file?\n\nBetween the lines (“Programs were instructed to run using a single thread.”), I take it that falco can use multiple threads. Is that correct? And if so, why not point it out more explicitly? And out of curiosity: when you run falco on 4 cores, do you see a 4x speedup? Or is the bottleneck something else (disk IO?) than raw computational power? How does it scale with the number of cores, in other words?\n\nSuppose my dataset is 10 Gb, and that I have 15 Gb left of free disk space. Do I dare to run falco on that dataset? How large is the output compared to the input? In a “minimum” (no extra features) as well as “maximum” (all extra features) mode?\n\nSuppose a user finds a bug, or wants to put forward a feature request. How can the user do that? Should this be mentioned in the manuscript?\n\nThe first question I always get from users of my software tools is: “where can I download the Windows binaries?”. Is it, actually, not a good idea to be explicit about the fact that this is a command-line tool that you compile on your own computer? Would probably save the authors some time to be upfront with this.\n\nThe list of references comes across as somewhat untidy. Some examples follow below. The authors should probably go through all references to make sure they comply with journal specifications.\n\n() Journal names are sometimes abbreviated, sometimes not. Ref 2 is abbreviated, whereas ref 3 should be abbreviated “BMC Bioinform.”\n\n() Article titles: should verbs and key nouns in article titles have a leading uppercase letter, as in, e.g., ref 5, or should they not, as in ref 1?\n\n() Should page ranges be written out in full (ref 23, “1202–1214.”) or should they be abbreviated (ref 25, “1213–8.”)?\n\nAre the installation instructions a bit too thin? I’d say yes, at least if the intention of the authors is to address a diverse audience and not just readers with Linux-style experience. To simulate a less experienced user, I used my son’s MacBook Pro and tried to install falco on it following the manual:\n\n“Upon downloading, inflating and moving to the source directory, installation can be done through the following commands: … $ ./configure CXXFLAGS=\"-O3 -Wall\" $ make all $ make install”\n\nSo I did:\n\n$ cd src $ ./configure CXXFLAGS=\"-O3 -Wall\" -bash: ./configure: No such file or directory\n\nAnd then\n\nconda install -c bioconda falco -bash: conda: command not found\n\nAnd that was it. No further clues or assistance to be found in the instructions.\n\nIf the authors are happy with this behavior, then they should make it clear in the manuscript that falco is not for everyone, but rather only for those with significant Linux-style experience.\n\n“Source code for falco available at: https://github.com/smithlabcode/falco.” – the trailing “.” should be removed, I’d say. The link won’t work for users who copy-and-paste it into their browser. The same thing goes for\n\n“Archived source code at time of publication: http://doi.org/10.5281/zenodo.352093313.” where both the reference and the “.” cause problems.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "6259",
"date": "27 Jan 2021",
"name": "Guilherme de Sena Brandine",
"role": "Author Response",
"response": "The reviewer has presented a thorough feedback to the description of the Falco software tool, as well as its implementation, description and documentation. We truly appreciate the very helpful comments, and provide our responses to improvement suggestions below. Comments were divided and numbered to allow us to refer to them in other places in our response if certain modifications to the manuscript pertain to multiple comments. (1) The authors present a welcome addition to the flora of FastQC-style read processing packages. The fact that it is a drop-in replacement for FastQC is particularly nice. The manuscript is a bit too short in my opinion. I miss some background information and some performance-related data. If the authors want to address a wide audience, they should probably work a bit more on the installation instructions and documentation too. The authors should probably also consider defining who their target audience is. We appreciate the comments on the introduction. We have expanded on the target audience on further comments (we address these in more detail on comment 16). We also fully agree that more background information can be provided. We expanded the second paragraph of the “introduction” section to add a brief description of some common quality control tests applied to most next-generation sequencing datasets. (2) As a discretionary comment: the authors sometimes, but not always, use the Oxford comma (see example below). I wonder if this is something that should be streamlined. “sequencing data is growing in abundance, dataset size and read length9.” vs. “…adapter trimming, filtering contaminants and low-quality reads, and mapping reads to a reference genome or transcriptome.” We thank the reviewer for this observation. We have revised the manuscript and ensured that the Oxford comma is adopted across the entire manuscript. (3) “as a safety criteria” should probably be “as a safety criterion”. We fully agree with the reviewer. This correction was made on the manuscript. (4) Good thinking behind the “We designed falco13 to faithfully…” paragraph. Nobody would be helped by yet another set of new file formats. A drop-in replacement is the way to go, if you ask me. And that is indeed what the authors deliver. We appreciate the comments, thank you! (5) I must make the observation, though, that the name “falco” may not be available (?): https://www.falcoseed.com/ca/article/cibus-registers-new-falco-brand-79k-canola-hybrid/ (some other more or less commercial uses of “Falco” can be found on https://en.wikipedia.org/wiki/Falco). I’m not sure how North American trademark/proprietary laws operate, but I suggest that the authors consult with the lawyers of their university. Better safe than sorry, right. (I consulted with our lawyers at one point, and they had me change the name of a software package we were working on at the time.) We really appreciate the reviewer raising the possible legal issue that may arise from the program name. We have researched the matter and believe that the software name should not raise legal issues, both due to the program not having any commercial or profitable goals and the name “Falco” being a common proper noun in many Latin languages. (6) Two unintended line breaks: “uniformly, falco’s design centralizes” We thank the reviewer for this observation. This line was removed from the manuscript. (7) Please cite the Sequence Read Archive formally; see Kodama et al. (20121). We followed the reviewer’s suggestion and added the appropriate citation when referring to the Sequence Read Archive. We also corrected the meaning of the acronym from \"Sequencing Read Archive\", as it was previously written. (8) I like the reproducible nature of the “Results” section. We appreciate the comment, thank you! (9) Two superfluous line breaks: “gzip compressed files.” We have removed the line break from the manuscript. (10) Both “Falco” and “falco” are used in the manuscript. You’d think that the name would be fixed as either “Falco” or “falco”. We thank the reviewer for this observation. The notation used in the manuscript used “Falco” at the start of sentences and “falco” everywhere else, to resemble the name of the binary program used in the command-line interface. We have modified the manuscript to use “Falco” everywhere except in the “use cases” section, where an example command-line call for the program is shown. (11) Why not use the term “open source” at least once in the manuscript? We thank the reviewer for this observation. To address this, we started the “Implementation choices” subsections with the following sentence: Falco is an Open Source C++ implementation of the FastQC software tool built for UNIX-based operating systems. (12) Out of curiosity: how does falco compare to Liu et al. (20192)? We have downloaded the software from the URL provided in the manuscript (github.com/megagenomics/fastprongs) and performed comparisons on identical hardware to what was used in the manuscript. We tested on two datasets: Dataset 1 consisted of 76 million 150 base reads from arabidopsis (SRR12075121 in SRA), and dataset 2 contained 139 million 100 base reads from chicken (SRR5015166 in SRA). On dataset 1, Falco ran in 9:40 and FastProNGS ran in 5:16 with 3 threads (the default program configuration) and 10:42 on a single thread. On dataset 2, Falco ran in 14:30 and FastProNGS ran in 7:55 with 3 threads and 16:13 with a single thread. We noticed that FastProNGS only reports the following modules in their output: Basic statistics, adapter content, per base sequence quality, per base sequence content and sequence length distribution. We also tried to run Falco by enabling only these modules. Under these settings Falco ran in 4:35 for dataset 1 and 7:12 for dataset 2. In all tests, Falco ran with 92 MB of RAM, whereas FastProNGS used 1.26 GB. In a long-read dataset (test 12 in the manuscript), FastProNGS did not run successfully unless we configured it to only consider the first 200 bases of each read, in which case FastProNGS ran in 2 seconds, but only reported summaries for the first 200 bases of all reads. Falco ran in 12 seconds for this dataset. A very meaningful conclusion of this comparison is the potential advantage of multithreading in QC, as evidenced by the steep decrease in processing time from FastProNGS when multithreading is enabled. We noticed, upon inspecting its source code, that this performance improvement can be explained by FastProNGS reading multiple reads in batch and allowing a new set of reads to be loaded while the previous batch of reads is processed. In contrast, Falco loads and processes each read sequentially, which reduces RAM usage but makes multithreading difficult in its current implementation. The performance of FastProNGS suggests that switching to a “batch processing” paradigm may have significant speed advantages when multiple cores are used and enough RAM is available to load reads in batch, and this is something we will incorporate in future versions of Falco, especially in order to address (18). We thank the reviewer for bringing this tool to our attention. (13) I’ve seen one software tool for quality-score-based trimming of sequences that actually reads the entire query file into memory, and then started to process it. This works less well as data files continues to grow, obviously. Is it worth pointing out that falco does not do this? What is, in fact, the maximum file size allowed by falco? Or is this dictated solely by the operating system? Both disk and memory requirements will depend on the length of the largest read in the dataset, as Falco processes the input FASTQ one read at a time. To address the computational requirements necessary to run Falco in more detail in the manuscript, we created an additional subsection named “system requirements” under the “Methods” section, where the computational resources (memory and disk) required to run Falco successfully are discussed. Furthermore, we have added a paragraph in the “results” section summarizing the memory and disk usage for the tests used for comparison across programs. (14) You can produce some pretty funny behavior in some other tools for sequence QC/trimming by feeding them a file with a single FastQC entry in it, speaking of nothing. Is there, then, a minimum file size or number of query sequences for falco? There are no minimum or maximum file sizes required by Falco. We have tested (although not disclosed in the manuscript) that Falco successfully runs on empty files and single-read files. We thank the reviewer for having raised this issue, and have addressed that there are no constraints in file size or number of reads in the “system requirements” section stated in (13). (15) Unless I’m mistaken, there are no fastq files available on https://github.com/smithlabcode/falco. I think the authors should make one or two available, so that it will be smooth and easy for the reader to try the software out. “The scripts used to download files and reproduce the benchmarking steps described are also available in the same repository within the “benchmark” directory.” comes across as somewhat indirect to me. We thank the reviewer for this observation. While we cannot provide the full FASTQ files used to perform our comparisons in the GitHub repository, we do agree that the documentation of our tests can be made simpler for users who wish to test the program. We made modifications in our repository to simplify both testing in an example file and testing in the FASTQ files used in the manuscript for comparison. Specifically, we added (1) direct links to the SRA files under the “benchmark” directory and (2) an “example.fq” file, consisting of a FASTQ file of 1000 reads, which is used as input for the example commands provided in the README. (16) What, exactly, are the hardware and software requirements for falco? This may be worth pointing out. Between the lines I read “any computer you can install the GNU GCC compiler on” – but not all readers will probably read it this way. Instead, you’ll get the question: “Does it run on Windows 10?”. We agree that constraints should be disclosed in more detail, and that the limited support for usage of Falco on Windows should be more explicit. We have rephrased the first paragraph in the “implementation choices”. The last sentences disclose more explicitly that Falco, by design, is a UNIX-centric program made to be run on a command line and that, unlike FastQC, it cannot be run in a graphical user interface. (17) Also, how much memory is needed to run falco? And how much memory is used up by falco when it processes a large file? Falco requires under 1 GB of memory for any short or long read file generated by the current sequencing technologies. More memory will be required when technologies expand read lengths to the order of millions or billions of bases per read. We have added a discussion of disk and memory requirements under the “systems requirement” section, and also discussed the memory usage of the programs compared in the manuscript under the section “Falco is faster than popular QC tools”. (18) Between the lines (“Programs were instructed to run using a single thread.”), I take it that falco can use multiple threads. Is that correct? And if so, why not point it out more explicitly? And out of curiosity: when you run falco on 4 cores, do you see a 4x speedup? Or is the bottleneck something else (disk IO?) than raw computational power? How does it scale with the number of cores, in other words? Falco, like FastQC currently does not use multiple threads to process a single file, and no significant speed difference was observed when running fastp with multiple threads, which is why we focused our comparison on single-thread across the software tools. Despite QC computations being fast relative to IO, our comparison with FastProNGS described in (12) suggests that multithreading can lead to speed improvements if reading and processing are done in parallel, and we certainly plan on exploring this paradigm in the next release of Falco. We have also rephrased the third paragraph of the subsection “Falco is faster than popular QC tools” to say “Both fastp and fastqc were instructed to run on a single thread” to avoid ambiguities regarding Falco’s multithread option. (19) Suppose my dataset is 10 Gb, and that I have 15 Gb left of free disk space. Do I dare to run falco on that dataset? How large is the output compared to the input? In a “minimum” (no extra features) as well as “maximum” (all extra features) mode? We have addressed the disk requirement on the “system requirements” section disclosed in (13) and (14), specifically adding the sentence “The total disk space required to store the three output files generated by Falco is under 1 MB”. Like FastQC, Falco’s output is a set of reports whose size scales with the maximum read length of the input but are never under 1 MB in total. We fully agree that the fact that disk space is not crucial to run Falco should be made more explicit. (20) Suppose a user finds a bug, or wants to put forward a feature request. How can the user do that? Should this be mentioned in the manuscript? We thank the reviewer for this observation. All issues and bug reports can be done through our GitHub page, the same one provided for the source code. To make this clearer for users, we have added a sentence at the “software availability” section, stating that errors, installation problems and bugs can be reported in the “Issues” section in the same URL provided to download the source code. (21) The first question I always get from users of my software tools is: “where can I download the Windows binaries?”. Is it, actually, not a good idea to be explicit about the fact that this is a command-line tool that you compile on your own computer? Would probably save the authors some time to be upfront with this. We fully agree with the reviewer, and have added the statements of a more specific target audience as discussed in (16). (22) The list of references comes across as somewhat untidy. Some examples follow below. The authors should probably go through all references to make sure they comply with journal specifications. Journal names are sometimes abbreviated, sometimes not. Ref 2 is abbreviated, whereas ref 3 should be abbreviated “BMC Bioinform.” Article titles: should verbs and key nouns in article titles have a leading uppercase letter, as in, e.g., ref 5, or should they not, as in ref 1? Should page ranges be written out in full (ref 23, “1202–1214.”) or should they be abbreviated (ref 25, “1213–8.”)? We really appreciate the keen observations on the reference standards. We have reviewed our citations and standardized journals to their non-abbreviated names, uppercase letters only in leading words, and pages written in full. (26) Are the installation instructions a bit too thin? I’d say yes, at least if the intention of the authors is to address a diverse audience and not just readers with Linux-style experience. To simulate a less experienced user, I used my son’s MacBook Pro and tried to install falco on it following the manual: “Upon downloading, inflating and moving to the source directory, installation can be done through the following commands: … $ ./configure CXXFLAGS=\"-O3 -Wall\" $ make all $ make install” So I did: $ cd src $ ./configure CXXFLAGS=\"-O3 -Wall\" -bash: ./configure: No such file or directory And then conda install -c bioconda falco -bash: conda: command not found And that was it. No further clues or assistance to be found in the instructions. If the authors are happy with this behavior, then they should make it clear in the manuscript that falco is not for everyone, but rather only for those with significant Linux-style experience. We really appreciate the reviewer bringing up this observation about our documentation. We agree that the wording of “moving to the source directory” was misleading and may cause users to try to run the commands on the “src” directory. We have updated our README with clearer command line instructions that show the user how to clone the repository or download a release file, as well as which directory to move to in order to run the commands. We are striving to make the documentation as clear and simple as possible, and truly appreciate these suggestions on how these can be improved. (27) “Source code for falco available at: https://github.com/smithlabcode/falco.” – the trailing “.” should be removed, I’d say. The link won’t work for users who copy-and-paste it into their browser. The same thing goes for “Archived source code at time of publication: http://doi.org/10.5281/zenodo.352093313.” where both the reference and the “.” cause problems. We appreciate this observation and the potential problems punctuation near links may cause. We have ensured that the trailing dots and citations are clearly separated from the links and that they will not cause problems when copying URLs directly from the manuscript."
}
]
},
{
"id": "72941",
"date": "30 Oct 2020",
"name": "Weihong Qi",
"expertise": [
"Reviewer Expertise Genome informatics."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors developed falco, an emulation of the popular FastQC tool, which is faster and can handle very long Nanopore reads. It is a useful development, especially for core facilities and research labs that produce high volumes of sequencing data regularly, where generating read QC reports in a timely fashion is indeed helpful. I only have a few questions and one minor comment:\nMain questions:\nThe implementation session could be expanded with more details. From my understanding, the major improvement was identified duplicated analysis in FastQC analysis modules, and implemented a single analysis workflow that was sufficient to generate the same modularized results. But it is not clear to me which changes make falco to handle long ONT reads successfully, while the original FastQC failed.\n\nThe original FastQC is portable (Unix. Mac and Windows). It also has a GUI version for less experienced users. These features are not important for experienced users and automated workflows where analyzing large amounts of data in a short time is the focus. But they can be important for other type of end users. The authors should at least point out these differences.\n\nIn results, run times of multiple QC tools analyzing different datasets were compared, how about the RAM usages?\n\nMinor comment: The sentence “While FastQC is capable of making summaries for protocols such as 45427 PacBio28, which generate sequences with around 10,000 bases per read ” should be updated to \"While FastQC is capable of making summaries for protocols such as 45427 PacBio28, which generate sequences with around 10,000-20,000 bases per read\".\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "6260",
"date": "27 Jan 2021",
"name": "Guilherme de Sena Brandine",
"role": "Author Response",
"response": "The reviewer has raised some important questions about points not addressed by the manuscript. We provide our responses below, and highlight changes made to the manuscript to address the reviewer’s comments. The authors developed falco, an emulation of the popular FastQC tool, which is faster and can handle very long Nanopore reads. It is a useful development, especially for core facilities and research labs that produce high volumes of sequencing data regularly, where generating read QC reports in a timely fashion is indeed helpful. I only have a few questions and one minor comment: Main questions: 1) The implementation session could be expanded with more details. From my understanding, the major improvement was identified duplicated analysis in FastQC analysis modules, and implemented a single analysis workflow that was sufficient to generate the same modularized results. But it is not clear to me which changes make falco to handle long ONT reads successfully, while the original FastQC failed. We really appreciate the reviewer highlighting the missing details regarding FastQC’s behavior. Upon trying to address this comment, we have further explored the FastQC code to understand why it failed for long reads, and we have learned that the perl script that wraps the FastQC call imposes a maximum memory limit of 250 MB per thread, which we were not aware of at the time we wrote the manuscript, and was prohibitive for the long read samples we have selected. Changing this configuration internally allowed us to run FastQC in the samples that we were previously unable to, thus allowing us to report a more comprehensive comparison of both time and memory for every test we have gathered. For this reason, we have removed the section named “Falco scales for larger nanopore reads”, instead replacing it with a paragraph at the end of the subsection named “Falco is faster than popular QC tools”, where the memory usage of each mapper is discussed. We have also filled Tables 3 and 4 with the results of running FastQC in the three long-read samples in our tests under the same hardware settings used for other tests. 2) The original FastQC is portable (Unix. Mac and Windows). It also has a GUI version for less experienced users. These features are not important for experienced users and automated workflows where analyzing large amounts of data in a short time is the focus. But they can be important for other type of end users. The authors should at least point out these differences. We thank the reviewer for this observation. We have made modifications to the first paragraph of the “implementation choices” subsection under “methods” to clarify that falco was designed for UNIX systems and does not include a graphical user interface. 3) In results, run times of multiple QC tools analyzing different datasets were compared, how about the RAM usages? We appreciate the observation about RAM comparison. We have added two paragraphs in the manuscript that address memory requirement in more detail. A “system requirement” subsection under “methods” was added to emphasize that falco requires about 100 MB for short read samples and under 1 GB for samples with read lengths of at most 1 million. As stated in question (1), we also reported the RAM usage for the software tools compared in the tests, both for short-read and long-read tests in the section “Falco is faster than popular QC tools”. Minor comment: The sentence “While FastQC is capable of making summaries for protocols such as 45427 PacBio28, which generate sequences with around 10,000 bases per read ” should be updated to \"While FastQC is capable of making summaries for protocols such as 45427 PacBio28, which generate sequences with around 10,000-20,000 bases per read\". We thank the reviewer for the observation. The section containing this sentence was removed given the shift in the focus of the manuscript to memory comparison, as addressed in questions (1) and (3)."
}
]
}
] | 1
|
https://f1000research.com/articles/8-1874
|
https://f1000research.com/articles/10-49/v1
|
27 Jan 21
|
{
"type": "Method Article",
"title": "Impact of diagnostic accuracy on the estimation of excess mortality from incidence and prevalence: simulation study and application to diabetes in German men",
"authors": [
"Ralph Brinks",
"Thaddäus Tönnies",
"Annika Hoyer",
"Thaddäus Tönnies",
"Annika Hoyer"
],
"abstract": "Aggregated data about the prevalence and incidence of chronic conditions is becoming more and more available. We recently proposed a method to estimate the age-specific excess mortality in chronic conditions from aggregated age-specific prevalence and incidence data. Previous works showed that in age groups below 50 years, estimates from this method were unstable or implausible. In this article, we examine how limited diagnostic accuracy in terms of sensitivity and specificity affects the estimates. We use a simulation study with two settings, a low and a high prevalence setting, and assess the relative importance of sensitivity and specificity. It turns out that in both settings, specificity, especially in the younger age groups, dominates the quality of the estimated excess mortality. The findings are applied to aggregated claims data comprising the diagnoses of diabetes from about 35 million men in the German Statutory Health Insurance. Key finding is that specificity in the lower age groups (<50 years) can be derived without knowing the sensitivity. The false-positive ratio in the claims data increases linearly from 0.5 per mil at age 25 to 2 per mil at age 50. As a conclusion, our findings stress the importance of considering diagnostic accuracy when estimating excess mortality from aggregated data using the method to estimate excess mortality. Especially the specificity in the younger age-groups should be carefully taken into account.",
"keywords": [
"Illness-death model",
"chronic conditions",
"diabetes",
"lupus",
"partial differential equations",
"epidemiology"
],
"content": "Introduction\n\nFor research purposes, aggregated data about the prevalence and incidence of chronic conditions become more and more available. Examples range from data of huge public health surveys, such as the National Health Interview Study (NHIS) in the US [CDC 2020] or the Global Health Data Exchange (GHDx) catalog [GHD 2020], which covers up to three decades of international health data, to claims data from health service providers [CMS 2020].\n\nRecently, we proposed a new method to estimate the age-specific excess mortality in chronic conditions from aggregated age-specific prevalence and incidence data based on a differential equation [Tönnies et al., 2018; Brinks et al., 2019]. The idea, in brief, is to relate the temporal change of the prevalence with the incidence and the excess mortality. If the incidence and prevalence are given, the excess mortality can be estimated. In age groups below 50 years of age, estimates from this method have been proven to be unstable or implausible [Brinks et al., 2020]. For example, we obtained estimates of the mortality rate ratio in type 2 diabetes with values greater than 100 in ages below 40 years [Brinks et al., 2020]. The typical range for type 2 diabetes in this age group is between 3 and 10 [Carstensen et al., 2020]. In [Brinks et al., 2020] it was hypothesized “that the diagnostic accuracy of the claims data plays a crucial role for the proposed methods of estimating excess mortality.”\n\nSimilar to diagnostic accuracy studies, we are interested in the sensitivity and specificity of the available diagnoses in the claims data. As “gold standard” we consider the presence or absence of the chronic condition in real life (as judged by an expert from the associated medical domain). Within the claims data, two types of error may occur: People with the condition in real life might not have the diagnosis coded in the claims data (false negative) or vice versa, people without the condition in real life might have a corresponding diagnosis (false positive). Finally, this leads to the concept of sensitivities and specificities of the aggregated prevalence and incidence data.\n\nThe aim of this article is twofold: First, we want to examine and quantify the impact of diagnostic accuracy on the estimates of excess mortality. For this, we use a simulation study comprising two settings, a low and high prevalence setting. Second, as a real-world application of the findings in the first part, we estimate the age-specific diagnostic accuracy of claims data about diabetes from about 35 million German men in the Statutory Health Insurance [Goffrier et al., 2017].\n\n\nMethods\n\nBefore we start with the simulation and the real-world application, we briefly sketch the theoretical background. Detailed derivations are given in Extended Data [Brinks et al., 2021].\n\nBased on the illness-death model for chronic diseases (Figure 1), it can be shown that the temporal change, ∂p=(∂t+∂a) p, of the age-specific prevalence p is related to the incidence rate i, and the mortality rates m0 and m1 of the people with and without the chronic condition (disease), respectively. Instead of the rates m0 and m1, the general mortality m = pm1 + (1 − p) m0 and the mortality rate ratio R = m1/m0 can be used according to the following equations [Brinks et al., 2014; Brinks et al., 2016]:\n\nPeople aged a at time t in the population are in one of the three states: Healthy, Diseased, or Dead. Transitions between these states are described by the rates i, m0, and m1, which in general depend on t and a.\n\nGiven the age-specific prevalence p, the age-specific incidence rate i and the general mortality rate m, Equation (1) provides an estimator for the mortality rate ratio R:\n\nAssuming that the sensitivity (se) and specificity (sp) in the age-specific prevalence and incidence are known, the prevalence p and incidence i in Equations (1) and (2) can be obtained from the observed (and possibly imperfect) prevalence p(obs) and incidence i(obs) by\n\nThe derivations of these equations are shown in Extended Data Appendix 2 [Brinks et al., 2021]. The observed values p(obs) and i(obs) may have been prone to error by incomplete case-detection (i.e., se < 1) and/or false positive findings (sp < 1). If all sensitivities and specificities equal 1, we find p = p(obs) and i = i(obs). Note that in Equations (3a) and (3b) we distinguish between sensitivities and specificities in prevalence and incidence (indicated by the sub-indices p and i, respectively). To examine potential age effects, se and sp may depend on age a. Age dependency is taken into account, because diagnostic accuracy in many diseases is known to depend on age. For example, sensitivity of diagnosing type 2 diabetes in 80 years old people is higher than in 40 year old people, which is, for instance, reflected by the higher percentage of undiagnosed diabetes in younger age groups [Gregg et al., 2004].\n\nThe steps for running the simulation studies in the low and high prevalence setting are as follows: We first solve Equation (1) with known i, m and R to obtain prevalence data p. Second, imperfect diagnostic accuracy is mimicked by using Equations (3a) and (3b) such that the quantities p(obs) and i(obs) are observed instead of the (true) quantities p and i. In the third step, Equation (2) is applied to p(obs) and i(obs) in order to obtain an estimate for the mortality rate ratio (R(obs)). Finally, R(obs) is compared to the true R underlying the simulation. This is done for a wide range of age-groups (Table 1).\n\nWe use two figures for the comparisons: 1) The age-specific difference between R and R(obs) and 2) the summed absolute relative errors (where the sum is taken over the whole considered age range). The later figure is used to assess the relative importance of the sensitivities and specificities in the form of a tornado plot. A tornado plot displays the change of the considered outcome compared to a base-case scenario, if exactly one input variable, say the sensitivity of the incidence in an age group, is changed while all the other input values (i.e., the remaining sensitivities and specificities) are kept fixed. This is done for all input variables. The changes in the output are presented as vertical bars, which are then ordered descendingly to indicate the importance of the associated input variables on the output. The descending order leads to the largest bar being presented on top and the smallest bar at the bottom, which visually appears as a half of a tornado (see Figure 3).\n\nTable 1 shows the parameters for the two simulation settings in the low and the high prevalence scenarios. The low and the high prevalence scenarios are motivated by systemic lupus erythematosus (SLE) in women and type 2 diabetes in men, respectively. As SLE is more relevant in younger ages, we consider the age range from 20 to 70 years in this setting. Type 2 diabetes is especially important for ages greater then 40, which lead us to the choice of considering the range 40 to 80 years of age. Although the values for the sensitivity and specificity in Table 1 are the same in the younger and older ages, they are treated independently to allow exploration of the relative importance in the tornado plots. In any case, sensitivities and specificities are interpolated affine-linearly between the younger and the older age.\n\nThe source code for use with the free, open-source statistical software R (The R Foundation For Statistical Computing) can be found in [Brinks et al., 2020].\n\nBased on claims data of German men in the Statutory Health Insurance (SHI), Goffrier and colleagues report the age-specific prevalence p(obs) of type 2 diabetes in the years 2009 and 2015 [Goffrier et al., 2017]. Furthermore, the age- and sex-specific incidence rate i(obs) in middle of the period, i.e., in the year 2012, is given in the same report. In addition to the prevalence and incidence, the mortality rate ratios R of men with and without diabetes in the German SHI in the year 2014 have been reported in [Scheidt-Nave 2019]. Strictly speaking, the estimates of R from [Scheidt-Nave 2019] might have undergone diagnostic inaccuracies as well. However, the estimates are based on individual data (ID) and potential biases in ID analyses (e.g., by missing disease status at death [Binder et al., 2017]), are beyond the scope of this article. Thus, for simplicity we assume R = R(obs).\n\nWe use these data about p(obs), i(obs) and R to obtain estimates about the age-specific sensitivity and specificity of the prevalence and incidence via Equations (3a) and (3b). For this, we make the following approach: for each age group (denoted ak, k = 1, …, K) we assume that the sensitivity and specificity of prevalence and incidence are the same, i.e., sep(ak) = sei(ak) and spp(ak) = spi(ak), for all k = 1, …, K. The assumption of same sensitivity and specificity with respect to prevalence and incidence is justified because prevalent and incident cases are derived from reported diagnoses of all physicians treating the men in the SHI. If prevalence data suffer from incomplete case-detection or false positive findings, incidence data will suffer in the same way.\n\nIf we assume for the moment that the sensitivity se = sep = sei is known, we can combine Equations (3a) and (3b) with Equation (1) to estimate the specificity sp = spp = spi. This is possible, because with given general mortality m from the Federal Statistical Office of Germany [FSG 2020], all measures p(obs), i(obs), and R in Equation (1) are known from [Goffrier et al., 2017] and [Scheidt-Nave 2019] after applying the corrections in Equations (3a) and (3b). Hence for known sensitivity se, we can calculate sp from these data and the analytical findings in the previous section by a functional relation Φ\n\nThe exact formula for the functional relation Φ between sp on the left hand side and se, p(obs), i(obs), m, and R on the right hand side of Equation (4), is lengthy and presented together with its derivation and an algorithm in Extended Data Appendix 3 [Brinks et al., 2021]. An implementation of the algorithm in the statistical software R can be found in [Brinks et al., 2020]. For now, it is sufficient to notice that the relation in Equation (4) follows from Equations (1), (3a) and (3b).\n\nUnfortunately, we do not know the sensitivity of the diagnoses in the claims data. To overcome this problem, we use a probabilistic approach and randomly sample se from epidemiologically reasonable ranges between 70% and 99%. Then, we examine how the estimated specificity sp changes. For easier interpretation, we present the false positive ratio (FPR), FPR = 1 − sp.\n\nThe data and the source code for use with the free statistical software R (The R Foundation For Statistical Computing) can be found in [Brinks et al., 2020] (DOI: 10.5281/zenodo.4300684).\n\n\nResults\n\nFigure 2 shows the estimated age-specific mortality rate ratios R in the simulation studies. The left and right panel in Figure 2 refers to the low and high prevalence settings, respectively. While in case of perfect diagnostic accuracy, i.e. sp = se = 100%, the input values of the simulation (blue lines) and the estimates by Equation (2) (solid black dots) do not (visually) differ. Imperfect sensitivity and specificity lead to estimates biased upwards (open circles). It becomes visible that with increasing age the difference between the true and estimated values decreases.\n\nThe low prevalence and high prevalence setting are shown in the left and right panels, respectively. The input values are shown as blue lines. Mortality rate ratios R are estimated without any (visual) difference in case of perfect sensitivity se = 100% and perfect specificity sp = 100% (solid dots). In case of imperfect sensitivity and specificity, the estimates of R are biased upward (open circles).\n\nIn the assessment of the relative importance of the sensitivity and specificity in prevalence and incidence, we obtain the tornado plots as shown in Figure 3. Irrespective of the low (left panel in Figure 3) and high (right panel) prevalence setting, the specificity of the incidence (spi) in the lower age group has the greatest impact on the estimated mortality rate ratios. Specificity spi in the higher age group has the second strongest effect, followed by the specificities in prevalence (spp). The impact of the sensitivities is far weaker compared to the specificities. Note that the relative importance (abscissa) is given on the log scale.\n\nIn both settings, low (left panel) and high prevalence (right), the specificities (prefix sp) are the four dominant error factors in estimating the mortality rate ratio R. Compared to specificities, sensitivities (prefix se) have a low impact on the error in R.\n\nBy comparing the horizontal bars in the low and high prevalence settings, we see that the four specificities in the low prevalence settings have a greater effect than those in the high prevalence setting. The opposite is true in the sensitivities: in the high prevalence setting sensitivities have a larger impact than in the low prevalence setting.\n\nFrom Equation (4) we infer FPR = 1 - Φ(se, p(obs), i(obs), m, R). After uniformly sampling se(ak), where ak = 25, 32.5, 40, …, 85, represents the K = 9 age groups [ak - 7.5/2, ak + 7.5/2) of width 7.5 years, k = 1, …, 9, from the range 0.7 to 0.99 with N = 10000 samples, and calculating the associated FPR, we obtain the graph presented in Figure 4. Each dot in the grey area represents an FPRn(ak) based on a random sen(ak), n = 1, …, N. We see that irrespective of the randomly sampled values sen(ak) for ak < 50, the FPR increases from 0.5 to 2 per mil. For example, at age 40 the FPR is about 1.5 per mil, which means that roughly 3 in 2000 diagnoses of type 2 diabetes at that age are false positive findings. For age groups > 50, we can see an upper bound for the FPR that continues linearly, while the lower bound can reach 0 at ages between 60 and 70 years. For higher ages, the lower bound of the FPR increases again.\n\nEach dot in the grey area represents the FPR generated by one of the scenarios about the age-specific sensitivities.\n\n\nDiscussion\n\nIn this work we have described the impact of diagnostic accuracy on the estimates of the excess mortality of a chronic condition from aggregated age-specific prevalence and incidence data. It turned out in simulation studies that the specificity in lower age groups had the greatest impact on the estimated mortality rate ratio. Compared to sensitivity, specificity has a greater impact across all age groups. The reason may be seen in the fact that the specificity has a direct additive effect on the true prevalence and incidence, while the sensitivity has an multiplicative impact only, cf. Equations (3a) and (3b).\n\nIn the simulation studies it turned out that estimation of the mortality rate ratio is accurately possible if the underlying sensitivity and specificities are known. In principle, these quantities are estimable in surveys. For example, in the claims data a cross-sectional comparison of the diagnoses with the gold standard (expert examination) could be conducted. These findings could be used to apply the corrections as in Equations (3a) and (3b) before using Equation (1) to estimate the mortality rate ratio.\n\nBy application of the theory to the claims data from 35 million German men, we were able to estimate the false positive ratio (FPR) in diabetes diagnoses. The most striking conclusion is the linearly increasing FPR in age groups between 20 and 50 years. In age groups older than 50 years of age, we could estimate upper and lower bounds for the FPR, which allows an assessment of diagnostic quality in the claims data.\n\nAlthough most of our findings can be seen in the general theory of using the method of estimating excess mortality described in [Tönnies et al., 2018] and [Brinks et al., 2019], the application to real world data has two limitations that are important to mention. First, we assumed that the age-specific sensitivity and specificity are the same in both years 2009 and 2015. This might be an oversimplification, because it could, at least in principle, be that the diagnostic accuracy during this period of six years changed, for example, by implementation of screening programs, change of diagnostic criteria or by changes of reimbursement policies for diagnosing diabetes. However, we are not aware of such changes and refer studies about temporal changes in diagnostic accuracy to future analysis.\n\nThe second limitation lies in the assumption that the observed mortality rate ratio R(obs) in 2014 as reported in [Scheidt-Nave 2019] equals the true rate ratio R in 2012. Since the mortality rate ratio is relatively stable [p. 59 in Breslow et al., 1980], the mismatch between the two years is unlikely to impose a problem. However, we cannot assess the difference between the observed and true rate ratio. The main reason is the brief and vague description of the methods to estimate R in [Scheidt-Nave 2019]. For example, it remains unclear how the possible problem of competing risks (contracting diabetes versus dying without diabetes) has been addressed. However, the findings in [Scheidt-Nave 2019] are consistent with epidemiological surveys in Germany [Röckl et al., 2017] and with observations from the Danish diabetes register [Carstensen et al., 2020]. Thus, we think that the assumption R(obs) = R is justified.\n\nApart from these limitations, our findings stress the importance of considering diagnostic accuracy when estimating excess mortality from aggregated data using the method described in Equation (1). In particular the specificity in the younger age-groups should be taken care about.\n\n\nData Availability\n\nZenodo: Simulation to study impact of diagnostic accuracy on estimation of excess mortality, http://doi.org/10.5281/zenodo.4300684 [Brinks et al., 2020].\n\nZenodo: Estimation of excess mortality from incidence and prevalence: impact of the diagnostic accuracy, http://doi.org/10.5281/zenodo.4302183 [Brinks et al., 2020].\n\nZenodo: Extended Data: Impact of diagnostic accuracy on the estimation of excess mortality from incidence and prevalence - simulation study and application to diabetes in German men, http://doi.org/10.5281/zenodo.4434806 [Brinks et al., 2021].\n\nThis project contains the following extended data:\n\n- Detailed derivations of the Equations (1) to (4).\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nReferences\n\nBernatsky S, Boivin JF, Joseph L, et al.: Mortality in systemic lupus erythematosus. Arthritis Rheum. 2006; 54: 2550–7. PubMed Abstract | Publisher Full Text\n\nBinder N, Herrnböck AS, Schumacher M: Estimating hazard ratios in cohort data with missing disease information due to death. Biom J 2017; 59(2): 251–269. PubMed Abstract | Publisher Full Text\n\nBreslow NE, Day NE: Statistical Methods in Cancer Research. In: The Analysis of Case-Control studies Lyon: IARC; 1980; Vol 1. .\n\nBrinks R, Landwehr S: Age- and time-dependent model of the prevalence of non-communicable diseases and application to dementia in Germany. Theor Popul Biol. 2014; 92: 62–8. PubMed Abstract | Publisher Full Text\n\nBrinks R, Hoyer A, Landwehr S: Surveillance of the incidence of non-communicable diseases (NCDs) with sparse resources: a simulation study using data from a National Diabetes Registry, Denmark, 1995-2004. PLoS One. 2016; 11(3): e0152046. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrinks R, Hoyer A, Weber S, et al.: Age-specific and sex-specific incidence of systemic lupus erythematosus: an estimate from cross-sectional claims data of 2.3 million people in the German statutory health insurance 2002. Lupus Sci Med. 2016; 25;3(1): e000181. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrinks R, Tönnies T, Hoyer A: New ways of estimating excess mortality of chronic diseases from aggregated data: insights from the illness-death model. BMC Public Health. 2019; 19(1): 844. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrinks R, Tönnies T, Hoyer A: Assessing two methods for estimating excess mortality of chronic diseases from aggregated data. BMC Res Notes. 2020; 13: 216. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrinks R, Tönnies T, Hoyer A: Simulation to study impact of diagnostic accuracy on estimation of excess mortality (Version 01).Zenodo; 2020. Publisher Full Text\n\nBrinks R, Tönnies T, Hoyer A: Estimation of excess mortality from incidence and prevalence: impact of diagnostic accuracy (Version 01).Zenodo; 2020. Publisher Full Text\n\nBrinks R, Tönnies T, Hoyer A: Estimation of excess mortality from incidence and prevalence: impact of diagnostic accuracy (Version 01).Zenodo; 2021. Publisher Full Text\n\nCarstensen B, Rønn PF, Jørgensen ME: Prevalence, incidence and mortality of type 1 and type 2 diabetes in Denmark 1996-2016. BMJ Open Diabetes Res Care. 2020; 8: e001071. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCenters for Disease Control and PreventionNational Health Interview Survey (NHIS).Reference Source Last access, 2020-12-07.\n\nCenters for Medicare & Medicaid ServicesChronic Conditions Public Use Files (PUF).Reference Source Last access, 2020-12-07\n\nFederal Statistical Office of GermanyPeriod life table for 2011/13. Table code 12621-0001, GENESIS-Online database, Last access 2020-12-06.\n\nGlobal Health Data Exchangedata catalog. by the Institute for Health Metrics and Evaluation (IHME), Seattle WAReference Source Last access, 2020-12-07.\n\nGoffrier B, Schulz M, Bätzing-Feigenbaum J: Administrative Prävalenzen und Inzidenzen des diabetes mellitus von 2009 bis 2015. Versorgungsatlas. 2017. Publisher Full Text\n\nGregg EW, Cadwell BL, Cheng YJ, et al.: Trends in the Prevalence and Ratio of Diagnosed to Undiagnosed Diabetes According to Obesity Levels in the U.S. Diabetes Care 2004; 27(12): 2806–2812. PubMed Abstract | Publisher Full Text\n\nRöckl S, Brinks R, Baumert J, et al.: All-cause mortality in adults with and without type 2 diabetes: findings from the national health monitoring in Germany. BMJ Open Diabetes Res Care. 2017; 5(1): e000451. PubMed Abstract | Publisher Full Text | Free Full Text\n\nScheidt-Nave C: Nationale Diabetes-Surveillance am Robert Koch-Institut Diabetes in Deutschland - Bericht der Nationalen Diabetes-Surveillance 2019.Robert-Koch-Institut Berlin2019. Publisher Full Text\n\nTönnies T, Hoyer A, Brinks R: Excess mortality for people diagnosed with type 2 diabetes in 2012 - estimates based on claims data from 70 million Germans. Nutr Metab Cardiovasc Dis. 2018; 28(9): 887–91. PubMed Abstract | Publisher Full Text\n\nTamayo T, Brinks R, Hoyer A, et al.: The Prevalence and Incidence of Diabetes in Germany. Dtsch Arztebl Int 2016; 113(11): 177–82. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "85845",
"date": "24 May 2021",
"name": "Andreas Wienke",
"expertise": [
"Reviewer Expertise biostatistics and epidemiology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nFirst I would like to congratulate the authors for this excellent paper which examines how limited diagnostic accuracy in terms of sensitivity and specificity affects estimates of excess mortality based on prevalence and incidence data. In the first part relevant formulas from previous work by the authors are given with respect to the relationship between prevalence and incidence on one side and on excess mortality on the other side. Then, based on assumptions about sensitivity and specificity of aggregated data the influence of sensitivity and specificity at different ages in a high and low prevalence situation are investigated by simulations. One key result is that specificity can be obtained without knowledge of the sensitivity in lower age groups. Furthermore, the false positive ratio is investigated and quantified. Finally, the methodology is applied to diabetes 2 data of 35 million men in the German Statutory Health Insurance.\n\nThe paper is written in a very clear and sound style, I have only very minor remarks:\nAt page 4 the authors state that sensitivity of diagnosing type 2 diabetes in 80 years old people is higher than in 40 years old people. Surprisingly, this is not taken into account in Table 1 where sensitivity is given as 95% for both age groups.\n\nIn Figure 2 there is no blue line to see because of the coincidence of the simulation and the perfect estimation. It is explained in the text, but should be solved for the figure.\n\nMaybe it makes the discussion in the second last paragraph more clear when the authors add (again) that the estimates of R(obs) considered there are based on individual data.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "85844",
"date": "01 Jun 2021",
"name": "Dianna J. Magliano",
"expertise": [
"Reviewer Expertise Diabetes epidemiology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a modelling analysis which aims to assess the impact of diagnostic accuracy on the estimation of excess mortality from incidence and prevalence using a simulation study. The stimulation study which is developed tests two scenarios: one with a high prevalence setting and the other with a low prevalence setting. The finding is then applied to real diabetes data from claims data from the German Statutory Health insurance. The modelling shows that when estimating excess mortality of diabetes, diagnostic accuracy is very important. Specificity is more important than sensitivity across all age groups, and in particular, specificity in younger people has the greatest impact on the estimated mortality rate ratios.\nOverall, this is a clear and well-presented piece of work. One thing which may be useful is to have some idea of the size of the impact of specificity on the estimation of mortality ratio rate, in comparison to the effect of sensitivity. The authors state that there is a difference between the effect of sensitivity and specificity, but it may be useful for the reader to understand how much of an impact it has.\nMy other points are minor and relate to language:\nThe last line of the abstract should be re-written. Starting that sentence with ‘especially’ means the sentence is unclear. You could start with: ‘In particular…’.\n\nThe first sentence of the introduction could be re written to say: “…chronic diseases are becoming more available.”\n\nThe heading in the first row of table 1 could be more descriptive. Expand on “setting”. In the actual table heading: insert the word “used” between “settings “and “in”.\n\nTable entries of “Lupus” should be written in full.\n\nSignificant figures in table 1 are not consistent. I do understand why though.\n\nFigure 3 should have the panels labelled on figure. “low prevalence” and “high prevalence” or A and B.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "85843",
"date": "07 Jun 2021",
"name": "Bruce Bartholow Duncan",
"expertise": [
"Reviewer Expertise diabetes epidemiology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an important evaluation of the use of secondary data from a claims database to estimate the sensitivity and specificity of the inclusion in the database of chronic diseases present in the covered population. Secondary data are increasingly being used for disease surveillance, in this case for diabetes mellitus. As they come with error, corrections must frequently be made in analyses. These corrections must be derived and evaluated, as is done here. My comments do not include a verification of the equations presented, which I have no reason to doubt. However, such verification is a task for someone other than myself.\nMajor comments:\n\"false positive ratio\", I believe, should be \"false positive rate\".\n\nThe authors have appropriately alerted that sensitivity and specificity, as used here, are not of a diagnostic test, but rather of the presence of a diagnosis in the claims data. As these terms are applied in a context different from the usual one, I believe that readers would benefit from a bit greater detail, noting that sensitivity is the capacity of the claims system to include in its database all cases of diabetes (whether detected or not) present in those covered by the system and specificity is the capacity to include only true cases of diabetes among those covered. Thus, for example, a covered individual who has diabetes but was never tested and thus never detected would be a false negative, detracting from sensitivity.\n\nThe horizontal axis of graphs in Figure 3 is the \"Relative importance\". Please define what this means.\n\nAt the end of the Results, the authors state: \"...which means that roughly 3 in 2000 diagnoses of type 2 diabetes at that age are false positive findings...\". As the FPR is being described, the denominator should not be diagnoses of type 2 diabetes, but rather covered individuals truly without diabetes.\n\nA major issue for the diabetes epidemiology community is the relative frequency of undiagnosed diabetes, i.e., for every 100 true cases, how many are unknown cases). Some discussion of how to use the approach presented to achieve estimates of the prevalence of undiagnosed diabetes (1-positive predicted value) could increase the relevance of this report (or a future one).\n\nFigure 4: Is it possible to trace not only the bounds of the estimated FPR, but also the FPR point estimate at each age?\n\nWhy are the base-case sensitivity and specificity so high? In terms of sensitivity, the IDF´s 2019 Diabetes Atlas (https://www.diabetesatlas.org/en/) estimates that 24% of those with diabetes in its European region are undiagnosed. A German investigation estimated that between 3 and 9% of adults had undiagnosed diabetes (Tamayo et al., 20141). In terms of specificity, the fact that several percent of those who report having diabetes, when tested, are found to have normoglycemia (1-positive predicted value), coupled with the known large within-individual (biologic) variability over time of available means of diagnosis, suggests that specificity is not 99.95%.\n\nIs not the greater impact of specificity mainly due to the fact that many more individuals in the population do not have diabetes than do, and thus the specificity is acting on a larger (at younger ages far larger) fraction of the population?\n\nThe mortality rate ratio of diabetes has declined considerably over recent decades (see: Tables 3 and 4 of Gregg et al. (20182). However, as you state, the impact of this decline over a 2 year period is likely to be sufficiently small as to not impose a problem.\n\nMinor comments:\n\nKeywords should be reviewed. My understanding is that they should be MeSH terms. Thus, for example, \"lupus\" should be \"systemic lupus erythematosus\".\n\n1st sentence Introduction, better: \"...of chronic conditions has become...\".\n\nPage 4, before \"Simulation studies\", better: \"...For example, the sensitivity of a code for type 2 diabetes in the claims database in 80 years old...\".\n\nLast sentence page 4, better: \"exemplified\" than \"motivated\".\n\nDiscussion, second paragraph: I don´t understand what \"accurately possible\" means.\n\nAdditional comments related to specific review questions;\nAs I am not fluent in R, I cannot verify that the additional materials include the source data. I imagine not, as the source data must be huge, and initially with personal identifiers.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-49
|
https://f1000research.com/articles/9-295/v1
|
27 Apr 20
|
{
"type": "Opinion Article",
"title": "An environment for sustainable research software in Germany and beyond: current state, open challenges, and call for action",
"authors": [
"Hartwig Anzt",
"Felix Bach",
"Stephan Druskat",
"Frank Löffler",
"Axel Loewe",
"Bernhard Y. Renard",
"Gunnar Seemann",
"Alexander Struck",
"Elke Achhammer",
"Piush Aggarwal",
"Franziska Appel",
"Michael Bader",
"Lutz Brusch",
"Christian Busse",
"Gerasimos Chourdakis",
"Piotr Wojciech Dabrowski",
"Peter Ebert",
"Bernd Flemisch",
"Sven Friedl",
"Bernadette Fritzsch",
"Maximilian D. Funk",
"Volker Gast",
"Florian Goth",
"Jean-Noël Grad",
"Sibylle Hermann",
"Florian Hohmann",
"Stephan Janosch",
"Dominik Kutra",
"Jan Linxweiler",
"Thilo Muth",
"Wolfgang Peters-Kottig",
"Fabian Rack",
"Fabian H.C. Raters",
"Stephan Rave",
"Guido Reina",
"Malte Reißig",
"Timo Ropinski",
"Joerg Schaarschmidt",
"Heidi Seibold",
"Jan P. Thiele",
"Benjamin Uekermann",
"Stefan Unger",
"Rudolf Weeber",
"Hartwig Anzt",
"Felix Bach",
"Stephan Druskat",
"Frank Löffler",
"Bernhard Y. Renard",
"Alexander Struck",
"Elke Achhammer",
"Piush Aggarwal",
"Franziska Appel",
"Michael Bader",
"Lutz Brusch",
"Christian Busse",
"Gerasimos Chourdakis",
"Piotr Wojciech Dabrowski",
"Peter Ebert",
"Bernd Flemisch",
"Sven Friedl",
"Bernadette Fritzsch",
"Maximilian D. Funk",
"Volker Gast",
"Florian Goth",
"Jean-Noël Grad",
"Sibylle Hermann",
"Florian Hohmann",
"Stephan Janosch",
"Dominik Kutra",
"Jan Linxweiler",
"Thilo Muth",
"Wolfgang Peters-Kottig",
"Fabian Rack",
"Fabian H.C. Raters",
"Stephan Rave",
"Guido Reina",
"Malte Reißig",
"Timo Ropinski",
"Joerg Schaarschmidt",
"Heidi Seibold",
"Jan P. Thiele",
"Benjamin Uekermann",
"Stefan Unger",
"Rudolf Weeber"
],
"abstract": "Research software has become a central asset in academic research. It optimizes existing and enables new research methods, implements and embeds research knowledge, and constitutes an essential research product in itself. Research software must be sustainable in order to understand, replicate, reproduce, and build upon existing research or conduct new research effectively. In other words, software must be available, discoverable, usable, and adaptable to new needs, both now and in the future. Research software therefore requires an environment that supports sustainability.\nHence, a change is needed in the way research software development and maintenance are currently motivated, incentivized, funded, structurally and infrastructurally supported, and legally treated. Failing to do so will threaten the quality and validity of research. In this paper, we identify challenges for research software sustainability in Germany and beyond, in terms of motivation, selection, research software engineering personnel, funding, infrastructure, and legal aspects. Besides researchers, we specifically address political and academic decision-makers to increase awareness of the importance and needs of sustainable research software practices. In particular, we recommend strategies and measures to create an environment for sustainable research software, with the ultimate goal to ensure that software-driven research is valid, reproducible and sustainable, and that software is recognized as a first class citizen in research. This paper is the outcome of two workshops run in Germany in 2019, at deRSE19 - the first International Conference of Research Software Engineers in Germany - and a dedicated DFG-supported follow-up workshop in Berlin.",
"keywords": [
"Sustainable Software Development",
"Academic Software",
"Software Infrastructure",
"Software Training",
"Software Licensing",
"Research Software"
],
"content": "Background\n\nMeet Kim, who is currently a post-grad PhD student in researchonomy at the University of Arcadia (UofA). We will follow Kim’s fictional career in order to understand different aspects of research software sustainability. Note that in Kim’s world, many of the changes this paper calls for have already been implemented. (In our example, Kim is a female person. Of course, research software engineers (RSEs) can be of any gender.)\n\nComputational analysis of large data sets, computer-based simulations, and software technology in general play a central role for virtually all scientific breakthroughs of at least the 21st century. The first image of a black hole may be the most prominent recent example where astrophysical experiments and the collection and processing of data had to be complemented with sophisticated algorithms and software to enable research excellence1,2. Similarly, it is research software that allows us to get a glimpse of the consequences our actions today have on the climate of tomorrow. However, an implication of computer-based research is that findings and data can only be reproduced, understood, and validated if the software that was used in the research process is sustained and their functionality maintained.\n\nAt the same time, sustaining research software, and in particular open research software, comes with a number of challenges. Commercial research software often has revenue flows that can facilitate sustainable software development, maintenance, and documentation as well as the operation of adequate infrastructure. However, a large share of researchers base their research on software that was developed in-house or as a community effort. Many of these software stacks can not be sustained – often because research software was not a first class deliverable in a research project and hence remained in a prototype state, or because of missing incentives and resources to maintain the software after project funding ended. Another fundamental difference to industrial software development is that most developers of academic research software (often doctoral students or postdoctoral researchers) never receive training in sustainable software development3. In particular, as they see themselves usually as the primary user of a software product, there are virtually no incentives to invest in sustainability measures such as code documentation or portability. In combination with the predominance of temporary positions in research, this results in a highly inefficient system where millions of lines of code are generated every year that will not be re-used after the termination of the developer’s position. Part of the problem is the reluctance to accept research software engineering as an academic profession that results in a lack of incentives to produce high-quality software: producing high software quality needs sufficient resources, and although the San Francisco Declaration on Research Assessment (DORA) demands a change in the academic credit system, many institutions base promotion and appointments on traditional metrics like the Hirsch index4. It is obvious that an extraordinary amount of idealism is required to write sustainable code, including documentation and installation routines, as well as running infrastructure and giving support to others when resources can be used more profitably in writing scientific publications based on fragile prototype software5,6.\n\nThus, one main factor for the poor sustainability of research software is the lack of long-term funding for research software engineers (RSEs)7 who take care of the appropriate architecture, organization, implementation, documentation, and community interaction for the software, paired with the implementation of measures towards making the software sustainable during and beyond the development process8.\n\nIn this paper, we describe the state of the practice and current challenges for research software sustainability and suggest measures towards improvements that can solve these challenges. The paper is the result of a community effort, with work undertaken during two workshops and subsequent collaborative work across the larger RSE community in Germany. It has been initiated during a half-day workshop at first International Conference for Research Software Engineers in Germany (deRSE19) in Potsdam, Germany on June 5th, 2019, and continued during a dedicated two-day workshop in Berlin, Germany on November 7th and 8th, 2019, which was funded by the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG). Subsequently, the draft produced during the latter event was opened up for collaborative discussion by the German RSE community through de-RSE e.V. - Society for Research Software.\n\nWe mainly focus on the situation of research software and RSEs in Germany, where funding bodies increasingly acknowledge the importance and value of sustainable research software and related infrastructures. The DFG, the largest funding body for fundamental research in Germany, for example, opened a call for sustainable research software development at the end of 2016 and a second call for quality management in research software in June 2019. The first call was oversubscribed by a factor of 10-15, a strong indicator of unmet demand. As another example, the 2019 “Guidelines for Safeguarding Good Research Practice” codex of the DFG now explicitly lists software side-by-side with other research results and data. The FAIR principles for research data9 provide guidelines for data archiving, but enabling full reproducibility and traceability of research software requires additional steps10. In consequence, there are ongoing discussions on whether software should be considered as a specific kind of research data or as a separate entity11.\n\nThese positive developments notwithstanding, guidelines and policies for sustainable research software development in Germany are unfortunately still lacking, and longterm funding strategies are missing. This all leads to unmet requirements and unsolved challenges that we want to highlight in this paper by elaborating on (1) why research software engineering needs to be considered an integral part of academic research; (2) how to decide which software to sustain; (3) who sustains research software; (4) how software can be funded sustainably; (5) what infrastructure is needed for sustainable software development; and (6) legal aspects of research software development in academia. While we specifically focus on the research software landscape in Germany, we are convinced that many of the analyses, findings, and recommendations may carry beyond. We want to address RSEs who are experiencing similar challenges and newcomers to the field of research software development, but first and foremost political and academic decision makers to raise awareness of the importance of and requirements for sustainable software development. As a community, we work hard on overcoming the challenges of software development in an academic setting, but we need support – and reliable funding options and institutional recognition in particular – for the sake of better research.\n\n\nWhy sustainable research software in the first place?\n\nAfter graduation, Kim joins a fixed-term researchonomical research project. For her PhD thesis, she wants to crunch some data. Her colleague recommends learning some Boa, which is an all-purpose programming language often used in researchonomy. Luckily, the UofA runs regular Software Plumbery courses for researchers, including a Boa course. Kim takes the course and gains a solid understanding of the basics of the Hash shell, version control with Tig, and the basics of Boa. She starts writing scripts, which help her a lot with the data processing. Unfortunately, Kim’s scripts are quite slow and actually break after she installs a newer version of Boa. She visits the weekly Code Café organized by her university’s central RSE team. The RSEs not only help her update her scripts but also suggest some changes which speed up the computation by a factor of 25.\n\nDuring the next meeting with her PhD supervisor, Kim presents her collection of scripts. The supervisor encourages Kim to create a Boa library from them, as they will be very useful to other researchonomists. Thankfully, Kim’s project PI had applied for three RSE person months in their grant, so the project enlists an RSE from the central team. Over the next three months, Kim and the RSE work together to build the library, document it, test it, license it under the permissive Comanche license, update the TigLab repository to let others contribute, introduce automated builds for every code change via a continuous integration platform, and make the library citable. Finally, they release the first major version of the library, named hal9k and publish it through the university library’s software portal, where they get a DOI (Digital Object Identifier) for the version as well as a concept DOI for any future versions of the library. Working with the RSE, Kim has gained a good understanding of some methods in software engineering, and she’s thrilled because this also means she’ll be able to get a job with a local tech company once her fixed-term contract has run out.\n\nKim passes her PhD - of which hal9k is an important part - with flying colors, and soon citations to her library start appearing in the researchonomic literature. To Kim’s surprise, she also reads a blog post about a citizen science maker project which has used hal9k to process researchonomic data measured in a neighborhood of her hometown. She is invited to give a talk at the local office of Siren, a global tech company, which look to adopt hal9k, and pay Kim a generous speaker honorarium. So generous in fact, that Kim can pay a student assistant for a full year from the money.\n\nOur credibility as researchers in society hinges on the notion of proper research conduct, also known as “good research practice”. The digitalization of research has introduced complex digital research outputs, such as software and data sets. Although first recommendations12 and policies13 exist, they are far from being widely adopted. It is still somewhat unclear how to translate good research practice into good research software practice, for example in terms of validity and reproducibility, but also pertaining to the responsible use of resources. The damage that failing to do so is causing both to the progress of the research community and to the credibility of academic research in society is becoming increasingly clear with the growth of the replication crisis - while the lack of universally agreed-upon and supported good research software practice is not the main reason for that crisis, it clearly is a contributing factor.\n\nWhile it is obvious that software qualifies as a potentially re-usable digital artifact, the additional benefit of not just reproducing a given scenario, but transferring software use to new problems, domains, and/or applications, justifies developing research software with a long-term perspective as sustainable research software.\n\nIn order to support research, a sustainable software must be correct14–16, validatable, understandable, documented, publicly released, adequately published (i.e. in persistently identifiable form as software source code17, and potentially in an additional paper which describes the software concept, design decisions, and development rationale), actively maintained, and (re-)usable18–20. We also argue that truly sustainable research software must ideally be published under a Free/Libre Open Source Software (FLOSS) license, and follow an open development model, to (1) enable the validation of research results that have been produced using the software, (2) enable the reproducibility of software-based research, (3) enable improvement and (re-) use of the software to support more and better research, and reduce resources to be spent on software development, (4) reduce legal issues (see section below), (5) meet ethical obligations from public funding, and (6) open research software to the general public, i.e., the stakeholder group with arguably the greatest interest in furthering research knowledge and improving research for the benefit of all.\n\nTo make software-based research (and with that almost any research) reproducible, the used software must continue to exist. Furthermore, it must continue to be usable, understandable, and return consistent results (or potential changes to results and bug fixes must be clearly documented) in the evolving software and hardware environment. Moreover, the software should support reuse scenarios to avoid duplication of efforts and drain of resources. Therefore, if research software is publicly funded, it should be freely available under a FLOSS license.\n\nCurrently, creating and using sustainable research software is not sufficiently incentivized. To evaluate in which area this shortcoming should be addressed, we have identified the following challenges:\n\nLack of benefit for the individual: Currently, the primary motivation for sustainable research software is the common benefit, rather than the individual benefit. It is clearly beneficial for the research community as a whole to direct resources towards sustainable research software, as it enables better and more research by freeing funds for domain research rather than (repetitive) software development. But the developers are often even at a disadvantage (e.g., they publish fewer papers5,6), which in turn prevents sustainable research software.\n\nLack of suitable incentive systems: Contributions to research that are not traditional text-based products (i.e., papers or monographs) are still not sufficiently rewarded, or not rewarded at all, due to the missing implementation of mandatory software citation17,21–29, among other reasons. Interestingly, one third of research software repositories have a lifespan (defined as the time from the first time any code was uploaded to the last contribution) of less than one day (median: 15 days10), indicating that many codes are only made available publicly for the publication in a journal (as increasingly encouraged or required by journals30 and associated with higher impact31) but are not maintained thereafter.\n\nLack of awareness: Research software sustainability and its importance is lacking visibility as well as acceptance32–35, and research software engineering in its implementation as sustainable software development and software maintenance is not sufficiently supported, both in Germany and beyond8,36,37.\n\nLack of expertise: Knowledge about how to create, maintain, and support sustainable research software is emerging38–40 but has not yet permeated related activities within organizations - specifically teaching, mentoring, and consultancy. This lack of expertise can also lead to divergence between software design and community uptake, e.g., if the software fails to meet the needs of the target group, or is insufficiently usable. RSEs combine sustainable software engineering expertise with experience in one or more research domains.\n\nHeterogeneous research community: There are significant differences with respect to how software is developed, published, used, and valued in the different academic disciplines. Additionally, there is even heterogeneity within a community in terms of application and approach. This also makes it hard to train researchers for sustainable software development, as beyond basic training in computational research such as provided by The Carpentries, advanced courses for research software engineering are not widely available (with the notable exception of the CodeRefinery project). Targeted curricula must be developed and updated regularly, and specialized instructors need to be trained.\n\nLack of impact measures: It is unclear how to measure the impact of research software with respect to its quality, reusability, and benefit for the research community. This exceeds the implementation of research software citation (which is work in progress17,28,29,41), and pertains to sustainability and policy studies.\n\nInfrastructure issues: Due to a lack of knowledge about how sustainability features impact the application of research software, there is not yet enough evidence for whether centralized or decentralized facilities should be favored to further research software sustainability42–44. This in turn leads to a lack of infrastructure as a whole.\n\nLegal issues: Many obstacles for research software pertain to legal issues, such as applicable licensing and compatibility of licenses45, and decisions about license types.\n\nFunding issues: Despite some individual initiatives46–49, funding for the creation, maintenance, and support of sustainable research software is still scarce.\n\nSlow adoption of research software engineering as a profession: Career options for research software work are not fully determined, although career paths are emerging in some regions. Initially, the RSE initiative in the UK has made progress in this area, and RSE groups have been installed in many institutions. In Germany, the US, and the Netherlands, this is still work in progress. It is also not yet determined how to match research software engineering roles in public institutions with industry roles50.\n\nIn summary, the necessary but resource-intensive practice of creating, maintaining, supporting, and funding sustainable research software is not yet sufficiently incentivized and enabled by research institutions and funding agencies, nor does it align well with the publish-or-perish culture that is still prominent in most fields.\n\nTherefore, it is necessary to comprehensively motivate sustainable research software practice. In the following, we identify stakeholders of research software51–53, and explicate their particular motivations for sustainable research software. Subsequently, we specify challenges towards satisfying the demands of the individual stakeholders.\n\nWhile a wide range of stakeholders share interest in sustainable software, we argue that their individual motivation can differ quite significantly:\n\nThe general public benefits from research which supports the common good, in other terms: creates a better world, faster. Taxpayers have an interest in economical use of their tax money, to which duplicated or flawed efforts to create research software – in contrast to software reuse – is contrary. A subset of this group may be interested in sustainable, i.e., re-usable and understandable, software as part of citizen science.\n\nDomain researchers benefit from better software to do more, better, and faster research. Sustainable research software supports this through validated functionality (e.g., correct algorithms), the potential for reuse, and general availability. Sustainable software also potentially simplifies building upon previous research results by reusing the involved software to produce additional data or by extending the software’s functionality. In light of recent updates to definitions of good research practice, sustainable research software also allows domain researchers to comply with guidelines and best practices. Additionally, using a software that is sustainable enough to establish itself as a standard tool in a field signifies inclusion in a research community. Less directly, researchers may benefit from the existence of sustainable standard tools as they yield standard formats, which in themselves facilitate reuse of research data.\n\nResearch software engineers (RSEs) have an intrinsic interest in sustainable research software. They create better software for research, which enables more and better research. RSEs have an inherent interest in developing and working with high quality software, as part of professional ethics as well as good research practice. RSEs build their reputation on high quality software and software citation17,28, which will open up new career paths. Finally, for RSEs, creating sustainable research software is part of an attractive, intellectually challenging, and satisfying work environment.\n\nResearch leaders as well as research performing organizations mainly focus on the economic aspects and management of research, i.e., available funds, people, and time employed to optimize research output. Both need to make sure that their employees continually improve their qualification and generate impact to improve their standing in the various research communities and ensure continued funding. Overseeing and enabling the creation of sustainable research software advances their visibility in the field and makes their research endeavors both more future-proof and more easily traceable, reproducible, and verifiable and thus more likely to attract additional resources (including human resources). Research performing organizations can additionally benefit from sustainable research software if it can be reused in other areas, creating synergies between different research disciplines. These synergies typically free resources that can then be used in areas other than software development and maintenance. Finally, organizations can gain highly competitive positions in terms of funding and hiring opportunities, as well as a reputation for being on the cutting edge of research, through early adoption of research software engineering units, and the implementation of sustainable research software policy and practice.\n\nResearch funding organizations have inherent interest in – and directly benefit from – the existence of sustainable research software as it allows them to direct more resources towards actual research (rather than recreation of software) and increase return on investment. At the same time, funding organizations can create incentives for sustainable software by imposing policies that reflect the necessity of research software sustainability and creating respective funding opportunities.\n\nGeopolitical units have a strategic interest to be independent of other geopolitical units to ensure that research can continue seamlessly regardless of geopolitical developments and ensuing embargoes on information flow. Reuse of sustainable software additionally frees up funding for uses other than software development. Well-established, sustainable software systems can also attract researchers and companies in the research and technology sector.\n\nLibraries (also registries, indices) benefit from sustainable research software, as it will undergo a formal publishing process and be properly described in its metadata. Libraries can extend their portfolio beyond text-based research objects and stake claims as organizations harnessing the digitalization of research. In turn, they help to increase visibility and discoverability for research software through their services and advance the competitiveness of their organization or geopolitical unit. In addition, libraries also use research software and would thus benefit directly from a more sustainable research software landscape. Last but not least, by using FLOSS research software, libraries could avoid expensive licenses and often insufficiently adapted commercial software.\n\nInfrastructure units, such as supercomputing facilities and university computing centers, benefit from sustainable software as it makes their daily work in terms of software installation and user support easier. Additionally, they can position themselves at the forefront of research by bundling expertise on the creation and maintenance of sustainable research software and installing research software engineering teams.\n\nIndustry benefits from sustainable research software, as the process of creating and maintaining research software produces a highly-skilled workforce. Depending on the employed licensing model, sustainable research software can also be adopted by industry partners to reduce cost in corporate research and development. Helping to sustain research software may also enable positive outreach for companies across industry and into society.\n\nIndependent (open source) developers can get involved in research software, even if they are not employed by a research institution. This can help them get in contact with other developers in the field and may potentially lead to collaborations or job opportunities in research based on this extended experience.\n\n\nHow to decide which software to sustain?\n\nKim’s PI is happy because Kim writes a longer section on hal9k for the final project report and provides a software management plan alongside it, which ticks off a box in the template that the PI had previously worried about. The PI does not want to let Kim go and instead offers her to be co-PI on a follow-up project to test new methods on the data, and integrate them into hal9k as well. They are positive that such a project proposal has a good chance to be funded, as they can show impact of their first project via their university’s current research information system (CRIS) and through the number of citations of hal9k and the publications for which it was used. While they write the proposal, the faculty dean approaches the two to tell them that based on Kim’s work, they will now negotiate about two new RSEs for the central RSE team with the university’s provost for research and plan to consider candidates with a background in researchonomics.\n\nWhen they get the decision letter from the research funding organization, Kim and her co-PI are happy to learn that their new project has won the grant. The reviewers specifically point out the value of extending Kim’s Boa library to include the proposed new methods, as well as the significant reuse potential of hal9k for the researchonomic community as a direct effect of its well-engineered architecture and modularity. Additionally, they stress that it was really easy to evaluate the software due to the comprehensive test suite, documentation, and example data. In fact, during the first month of the new project, three other researchonomic research projects approach them to ask whether they can contribute to Kim’s library and offer to fund six months of RSE work for this. Kim uses this money to also parallelize hal9k together with the RSEs and works with her university’s computing center to offer it as a standard tool for researchonomic supercomputing.\n\nThe sustained funding of all existing software efforts is not only impossible but would risk overly splintering the community and eventually become counterproductive to the efficiency of the research community. Therefore, it is important to agree on a list of transparent criteria that qualify a software product for sustained funding. We recognize that defining research software engineering criteria for software evaluation will also lead to activities aiming at optimizing scores to achieve these criteria. Hence, the criteria have to be designed such that all score-pushing effort truly advances the value of the software. Criteria that can be manipulated without effectively adding value, i.e., wasting resources, should be excluded. The list of criteria presented in this section could be the basis for a structured review process that facilitates an unbiased evaluation of software tools from various fields. Therefore, this list must be general enough to be applied to research software from various research disciplines while also respecting differences between fields (e.g. citation rates between humanities and life sciences). The challenge to do justice to a wide spectrum is e.g. reflected by suggesting criteria comprising different levels54. One of the major challenges in the endeavor to define a selection scheme for sustainable funding of research software is to organize a fair and transparent review process. We believe that it is important that the review process is conducted by experts, or teams of experts, that have a strong background both on software engineering as well as on the domain-specific aspects, the latter because certain criteria often exist on a spectrum that is most likely shaped by the specific demands of the respective research community.\n\nWhile an assessment based purely on quantitative metrics would allow for seemingly objective comparisons between programs, the definition of valid and robust quantitative metrics that can be evaluated with reasonable effort is a major challenge. On the other hand, a structured qualitative assessment with scores for groups of criteria can provide a middle ground. It is clear that both preparing an application for a review against these criteria from the applicant side as well as the evaluation by the reviewers requires significant effort. We believe that the added value significantly outweighs the investment but appropriate resources need to be factored in. Sustainability of research software should be considered from the beginning for new projects. The criteria listed below, or a subset such as the “good enough” practices proposed by Wilson et al.40, are valuable throughout the development process (including early phases) for almost all types of research software applications. “Classical” research funding schemes should acknowledge the need to follow best practices during the development of new software and allow factoring in appropriate resources to design and implement for sustainability. In this section, we focus on the question which software to support in dedicated sustainability funding schemes. For such sustained funding, only software in application class 2 or 3 as defined by Schlauch et al.55, i.e., with significant use beyond personal or institutional purposes, would likely be considered. Excellence as reflected in funded projects, publications, and software adoption, i.e., backing by a community, should be considered during selection. Nevertheless, we believe a good scheme should strike a balance between consolidating the field to few well-established software packages on one side and stimulating innovation and cooperation promoting diversity in terms of more than one monopolistic package on the other side. Last but not least, there is an inherent conflict between the long-term goals of sustainability funding a software and the necessary reevaluation to monitor the state of the software over time.\n\nSeveral evaluation schemes for research software have been proposed before and led to the formulation of first recommendations12,13. Gomez-Diaz & Recio suggested the CDUR scheme based on Citation, Dissemination (including aspects like license, web site, contact point), Use, and Research (output)56. Lamprecht et al. rephrased the FAIR data principles9 for research software11. Hasselbring et al. found that the adoption of FAIR principles is different between fields with an emphasis on reuse in computer science as opposed to a reproducibility focus in computational science10. Fehr et al. collected a set of best practices for the setup and publication of numerical experiments57. Jiménez et al. boiled it down to four best practices58: public source code, community registry, license, and governance. Hsu et al.59 proposed a framework of seven sustainability influences (outputs modified, code repository used, champion present, workforce stability, support from other organizations, collaboration/partnership, and integration with policy). They found that the various outputs are widely accessible but not necessarily sustained or maintained. Projects with most sustainability influences often became institutionalized and met required needs of the community59. In the field of open source software, the CHAOSS (Community Health Analytics Open Source Software) project has developed metrics to evaluate sustainability. One objective of CHAOSS is to automatically generate project health reports based on software that evaluates the metrics, with most of the metrics already covered. The UK Software Sustainability Institute (SSI) suggested both a subjective tutorial-based and a more objective criteria-based software evaluation scheme60, the latter being available as an online form. ROpenSci61 provides software reviews for R developers, which have been very successful in the community. The review criteria of the Journal of Open Source Software (JOSS) focus on the aspects license, documentation, functionality, and tests. This list of essential items should be fulfilled by all research software that wants to be considered not only for publication but also for sustained funding.\n\nWe drew inspiration from all these works and suggest a set of criteria on which to base reviews for sustainable funding. This set comprises mandatory, hard criteria that we think have to be fulfilled across domains (highlighted in italics) and additional desirable, soft criteria that can be implemented to different degrees depending on the use case and domain-specific software development requirements. The soft criteria should be evaluated in a structured way by the reviewers with a specific response for each section rather than one running text. The fact that most of these criteria will be considered in any software management plan (SMP) highlights its importance for sustainable research software.\n\nUsage and impact. Requirements qualifying software for sustained funding are (1) its use beyond a single research group, (2) the scientific relevance and validity of the software documented in at least one peer-reviewed scientific publication. Ideally a paper also describes the scope, performance, and design of the software. (3) The use of the software in publications is a measure of impact but quantitative assessment brings about additional challenges24. Therefore, other, potentially domain-specific, impact measures, such as influence on policy and practice as well as use in other software and products should be considered as well to evaluate relevance for academia and society. Considerable attendance at training and networking events can be considered as a proof of use as well. (4) A market analysis needs to show that the software is important to a user base of relevant size and either unique or one of the main players in a field with several existing solutions. Geographical or political aspects can be considered as well, e.g. to support the maintenance of a European solution. A convergence process of (parts of) a research community towards a specific software stack, i.e., documented transition of several research groups to a common software, would be a strong indicator of impact. (5) As community uptake and benefits are a central goal of sustained software funding, outreach and appropriate training material for new users of the software are essential.\n\nSoftware quality. As mandatory criteria of software quality that have to be fulfilled, we consider (6) the public availability of the source code in both a code repository and an archive (for long term availability), developed using (7) version control with meaningful commit messages and linked to an issue tracker (ideally maintained, but at least mirrored on a public platform). (8) Documentation of the software needs to be publicly available comprising both user documentation (requirements, installation, getting started, user manual, release notes) and developer documentation (with a development guide and API documentation within the code, e.g. using Doxygen)62. (9) The license under which the software is distributed must be defined. Publicly funded software should be published under a Free/Libre Open Source Software (FLOSS) license by default, although exceptions to this might apply (e.g. excluding commercial use). (10) Dependencies on libraries and technologies must be defined.\n\nWe acknowledge that some additional criteria have to be evaluated under consideration of the research domain. These comprise (11) the availability of examples (comprising input data and reference results), (12) mechanisms for extensibility (software modularity) as one aspect of software architecture63 and (13) interoperability (APIs / common and open data formats for input and output), (14) a test suite (including at least some of the following: unit tests, regression tests, integration tests, end-to-end tests, performance tests; ideally run in an automated fashion in a continuous integration environment), (15) tagged releases (considering their frequency, and availability for end users in terms of binary packages for major operating systems, or availability via package managers or containers), (16) no large-scale re-implementations for functionality for which good solutions already exist. Many of these aspects require appropriate infrastructure (see page 12).\n\nMaturity. The research software applying for sustained funding must have already reached a certain level of maturity (typically class 2 or 3 as defined by Schlauch et al.55). A mandatory requirement is (17) a comprehensive and up--to-date software management plan64. The software should (18) be maintainable with an appropriate amount of resources as detailed in a sustainability section of the software management plan. The software has (19) a well maintained website with a clearly defined point of contact and a communication channel to inform users about news regarding the software such as new releases. Besides an active user community, sustainable software requires (20) a group of developers (i.e., definitely more than 1 developer) documented, e.g. by contributions to the code base or participation in documented, public discussions or issue tracking. Another criterion is (21) whether potential contributors are invited to participate in a clearly defined process (e.g., a CONTRIBUTING document). The group of developers should have defined a governance model for their project and easy ways for users to provide input regarding their needs.\n\nGiven the diversity in the software technology landscape, and the domain-specific software development cultures65, some of the above-mentioned criteria have to be evaluated against domain-specific requirements. Therefore, we highly recommend to base the selection process on a combination of (1) a software quality-based review and (2) a domain-specific scientific review. In particular, the former should be ideally performed by a central institution (e.g. at funding bodies or other independent agencies such as a software sustainability institute). Only criteria for which improvement truly advances the value of the software should be considered in evaluation schemes, i.e. no criteria that can be gamed. After rejecting software not fulfilling the mandatory criteria in a first stage of the review process, the second stage of the selection process should be realized as a transparent procedure ideally allowing the reviewers to interact with the PIs of the software (e.g. remote meetings, forum-like discussions) and put the software quality and development efforts into the domain-specific context. The outcome of this second stage should be a structured review assessing each criterion explicitly and a rating for each of the dimensions Usage and impact, Software quality, and Maturity. For sustained software funding, it is important to audit the performance, relevance, impact, progress, and level of sustainability of funded software after reasonable time frames. Such a reevaluation should revisit the criteria under consideration of evolving software technology and scientific standards, without requiring a completely new proposal being submitted. We envision funding periods of 5 years to provide sufficient security for funded software projects, while allowing for adaptation of the portfolio of funded software to novel research directions and community needs. Failure to meet the reevaluation criteria should lead to the decision to phase-out sustainable funding. The phase-out process may come with a 1-year funding program based on a consolidation plan with clear goals regarding the archiving and preservation of the software, documentation, and all existing resources.\n\n\nWho sustains research software?\n\nKim wants to broaden her research portfolio within researchonomics and applies for postdoctoral positions at other institutions. Her library hal9k is growing in popularity within researchonomics, and she wants to continue working on it. As her university has adopted an open science policy, hal9k is free software under a Free/Libre Open Source Software (FLOSS) license, and Kim is free to continue her work on the library even after moving away from UofA. Due to her involvement in the creation of hal9k as well as her previous success in attracting funding, Kim has the choice between multiple, attractive positions and decides to move to the researchonomics group at Eden University (EdU). She has already extended hal9k in multiple directions in the past and plans to continue this work at EdU. Her group leader at EdU would like to continue funding her but due to a law called the Fixed-term Research Contract Bill, EdU is not allowed to extend her contract, and neither third-party funding for her own position nor a permanent position are available. After having developed a now widely-used research tool, several publications in software and paper form, as well as having attracted funding, Kim finds herself looking for a job again.\n\nResearch relies on software and software relies on the people developing and maintaining it. Sustainable research requires sustainable software, and this in turn requires continuity for those who develop and maintain it.\n\nPossibly the most important demand is the need for an increase in recognition and awareness of research software as a first class citizen in research13,66,67. For sustainability of research software, long-term commitments of the respective software leads are crucial, but very few professional RSE profiles currently exist. In consequence, it is essential to create career paths for RSEs that are attractive and include permanency perspectives. While creating permanent positions in the German academic system below the faculty level is an actively discussed topic overall68, we specifically focus on the needs originating from the development and maintenance of research software here.\n\nAs already mentioned, research software development not only requires domain expertise, but also software development education, skills, and competence. Currently, most of the domain researchers developing and maintaining domain-specific software technology have not received professional training on software development3,38. To enhance the productivity and sustainability of computer-based research, it is essential to integrate software development training into the education of domain researchers.\n\nCurrently, a significant portion of the existing research software is developed by individuals or in small groups, primarily to serve their own requirements. This situation is unsatisfying in terms of collaboration and inefficient in terms of several groups spending resources on generating similar or even the same functionality. To enable and promote synergies, it is important to allocate resources for research software development and to build communities, as described in 69.\n\nWe are currently facing a lack of awareness for the importance of research software as discussed above. Moreover, there is little recognition for the efforts put into software development and maintenance. In consequence, software development in academic settings is mostly considered as a means to an end and sustainability is often not considered in project planning and grant proposals and contributes little to progressing research careers70. The main challenge here is the continued use of metrics that primarily leverage traditionally published articles and article citation numbers.\n\nIn academia, developers of research software are typically domain researchers, and in particular if new areas are explored, the software development process itself has research character. Obviously, developing research software requires not only domain knowledge but also software development skills, and the researchers leading the software development process are often domain experts with substantial software development experience, making them extremely valuable members of the research community. However, the current academic system in Germany does not provide a defined RSE role. Fixed-term positions are, at least currently within the German academic system, often effectively the end of a Research Software Engineer's career path, sometimes even a dead end. The challenge here is the lack of available permanent positions within the non-professorial academic faculty (“Mittelbau”) in Germany, compounded by a lack of access to these few permanent positions for RSEs. This in turn is due to the already mentioned lack of recognition for efforts concerning research software for faculty appointments within domain sciences.\n\nIn order to develop sustainable software, researchers need to have the skills and expertise to build software that is easy to maintain and extend71. However, most of the researchers are self-taught developers3,38. Ideally, these skills have to be built into the domain science curricula, which could generally be done in two different ways (or a combination of them). One obvious solution attempt are additional courses that focus on these topics. The main challenge here is to decide which other topic(s) to possibly drop due to the limited volume of any given curriculum. A different approach is to incorporate software-related topics into existing domain science courses. While this would provide the benefit of show-casing the usage of specific software skills directly within the domain science, the challenge here is the amount of work necessary to change existing lecture material, let alone the need of the lecturers to acquire those skills themselves in the first place.\n\nAs long as the necessary software skills within domain sciences are not yet wide-spread, building a network from those that have acquired relevant skills is difficult. Community efforts, that concentrate on questions regarding research software, can help to fill this gap. Examples of such efforts include the Software Carpentries, national and international RSE societies (e.g., within Germany deRSE e.V.). However, since research software is such an interdisciplinary topic, it is hard to get recognition and find funding within any specific discipline. As a result, existing communities often have to rely heavily on volunteers. This is challenging because despite benefits to domain science, volunteers hardly receive recognition for their work “back home”, i.e., within their domain, underlining again the importance of our first demand.\n\nIncreasing recognition and awareness is a challenge that calls for both immediate action and perseverance. Nevertheless, some measures will likely show positive effects comparatively soon.\n\nSimilarly to plans for research data management, funding agencies should request that applicants include considerations about how software developed in a project can be sustained beyond the end of the funded project. A follow up on these plans during and after the project lifetime, i.e., a dedicated software management plan, is crucial.\n\nAnother recommendation is aimed at decision makers concerning recruitment for academic positions: broaden the definition of research impact beyond traditional scientific publications to also include other impactful results. Not all researchers that think of themselves as RSEs pursue a faculty position as their main career goal. However, permanent academic non-faculty positions are rare within the German academic system, also due to the lack of a defined RSE role. We recommend research institutions to leverage the benefit of dedicated RSEs by establishing attractive long-term career options in the academic environment. The long-term solution in order to gain sufficient software development skills should be education that is included early in the career path, ideally already at the Bachelor level. For the time being however, efforts involving workshops and seminars that provide easy access to hands-on training on software-related questions should be promoted and supported as much as possible.\n\nIt is important to provide an environment where communities can form and flourish by allocating resources for research software development and for building communities around it58,69,72. The identification with a community of like-minded people and personal action73 can lead to a permanent establishment of sustainable research software as a valuable research output. Thus, research institutions as well as funding agencies should not only be open-minded regarding existing volunteer organizations, but should actively promote the creation of such groups.\n\n\nHow can research software be sustainably funded?\n\nHal9k has grown into a widely used software in researchonomics, and Kim is proactively asked to apply for - and is subsequently awarded - a permanent RSE position at the institute for researchonomy at UofA, based on her work on the library. She works closely with the central RSE team, but mostly due to bureaucracy and the high demand for her library, Kim does not have enough time to maintain and further develop hal9k alone anymore. Together with the dean she develops a course for the researchonomics curriculum which teaches data processing with hal9k. As a lesson from her own career, she starts the course with sessions on the Hash shell, version control with Tig, Boa, and two whole sessions on basics of sustainable software development. This is very fruitful, and due to the implementation of a new research software funding scheme at UofA, Kim is able to hire one of the course students, who has shown great RSE skills, straight into a long-term position at her institute, where they focus on the maintenance and development of hal9k, work with the computing center to support hal9k-based supercomputing on a new, dedicated FGPA cluster, develop training materials for external users, and organize the yearly hal9k users and developers conference. Kim gets to travel the world to visit researchonomics groups who are using hal9k.\n\nSustainable funding for research software boils down to funding the four main pillars enabling sustainable software development: (1) Personnel with expertise in research software development; (2) Infrastructure for developing, testing, validating, and benchmarking research software; (3) Training in software design and sustainable software development; and (4) Community management and events for creating synergies between research groups and software efforts.\n\nShort-term engagement of (early career) researchers raises the question of how to maintain a constant level of expertise within a developer team and prevent knowledge drain concerning domain knowledge and software engineering skills. Conversely, the permanent engagement of qualified personnel requires to offer career perspectives, especially due to the fact that academia competes with industry for the same people. A challenge specific to Germany is posed by the shortage of permanent positions and by the restrictions for temporary positions due to the German Wissenschaftszeitvertragsgesetz74.\n\nSustainable software development requires hardware technology to develop, test, validate, and benchmark features in a continuous integration cycle. The challenge in this context is the persistent evolution of the hardware landscape. Hence, for creating an environment promoting sustainable software development, it is important to provide access to a wide hardware portfolio and to support a development cycle based on continuous integration.\n\nExpertise in sustainable research software development is a scarce resource, and training is heavily needed as one way of building up more expertise. However, while integrating interdisciplinary software engineering courses into the education curriculum can build up basic skills, some expertise is domain-specific and requires interinstitutional training activities. Furthermore, there exist no financial incentives for creating software-specific documentation and tutorials nor to provide other forms of support.\n\nWhile the creation of research software communities is one major asset in sustaining research software technology, promoting this process requires the installation of new funding instruments. Traditionally, research grants are limited to rather short time frames and support personnel, material, hardware, and to a limited degree also travel and research visits. Creating a research software community however requires funding for community and training events as well as “virtual hardware” such as webspace, versioning systems, task-managing systems, and compute cycles. These demands can hardly be met without third-party funding42,75–77.\n\nFunding is a crucial factor for sustaining research software. Currently available sources and instruments are not adequately shaped for the challenges and solutions outlined above. We recommend actions on the individual, organizational, and national level.\n\nExisting project-focused funding instruments on the local, national, and international level need to be complemented with funding instruments specifically designed for research software development and sustained research software maintenance to make research software a first class citizen in the research landscape. For example, software projects enhancing research and fulfilling the sustainability criteria detailed in section How to decide which software to sustain? may be entitled for sustained funding as long as they live up to the standards and remain a central component of the research landscape.\n\nComputing centers and supercomputing facilities for research need to receive earmarked resources for the support of sustainable software development. This funding is necessary to provide continuous integration services, a hardware portfolio for development, testing and benchmarking software, as well as personnel for training domain researchers in software design and the proper usage of the services.\n\nThe creation and maintenance of training materials for general research software engineering education and the software-specific documentation and tutorial creation needs to be reflected in funding opportunities. This can either happen by dedicating modules of research or software grants to providing support and the generation of training material, or by opening funding schemes focusing on interdisciplinary software development education. The latter may include research that looks at research software development as a process to analyze which measures, interactions, and team compositions make research software successful. Additionally, funding instruments fostering the formation of research software communities have to be established.\n\n\nWhich infrastructure is needed to sustain research software?\n\nAs the hal9k community grows, so does the need for infrastructure. Kim and her team collaborate with the National RSE Consortium to set up hal9k on the Consortium’s distributed TigHub instance and organize world-wide access to it via the NRSEC-AAI federation. Going forward, the Consortium’s Research Software Hub - a registry and Software Heritage Archive-based long-term repository for research software on a national level - ingests hal9k releases with complete metadata: citation information, the hal9k provenance graph and computational environment information, ORCID iDs, etc. and provides its own DOIs for versions under a concept (umbrella) DOI. The community reviews all code and documentation changes that are contributed to hal9k via the central TigHub instance. The Hub’s CI system Alfred builds, tests, and pushes new releases automatically to the registered supercomputing clusters. Community efforts become better and more streamlined by the day, as research software development training is now offered as part of most curricula, and skilled RSEs are now much easier to find and hire by research institutions.\n\nResearch software is developed by individual researchers, in small teams within a single institution, or in larger teams distributed across multiple institutions. In particular if software development is distributed across institutions, there exists an urgent need for frameworks and tools enabling collaborative code development, software feature planning, and software management. As research software development typically includes bleeding-edge research and development that the researchers do not want to disclose for a certain time to preserve intellectual property, distributed research software development also needs a global Authentication and Authorization Infrastructure (AAI). We recommend the development and/or deployment of tools for distributed software development and software management as central research infrastructure. An important aspect in this context is the cataloging of research software to reduce the duplication of development efforts. This can efficiently be realized by promoting the registration of all research software with a unique identifier and developing a tool that allows to explore the research software landscape. Research software contributors should have an ORCID iD to be uniquely identifiable and referable. While some funding for such tools and software repositories is emerging (e.g. the bio.tools catalogue of bioinformatics tools funded as part of the European ELIXIR project78), a standardized extension of such efforts to the RSE community as a whole is necessary. However, as the experiences from ELIXIR demonstrate, this is a non-trivial effort that requires significant dedicated and long-term funding.\n\nAs elaborated, training in sustainable software development is key to achieve sustainability in research software. At the same time, it is not clear how such training should be facilitated and institutionalized. Furthermore, for deriving software quality standards, evaluating the quality of software, and providing a code review service, central resources are necessary that individuals and groups in the research software landscape can draw from.\n\nWe consider Software Carpentry and similar efforts like the creation of the Data Science Academy HIDA in the Helmholtz Association of German Research Centers helpful solutions to exchange and distribute knowledge. Local chapters of RSE groups and (inter-)national conferences will further foster networking and community building. We strongly recommend the creation of a national Software Sustainability Institute (involving funded positions to establish web platforms and training material) similar to the UK Software Sustainability Institute (SSI), which serves as a national contact for all aspects related to research software. The UK SSI also publishes best practice guidelines for research software engineering.\n\nProper software publication and possibilities for the community to find existing software solutions for a given problem are a prerequisite to optimally exploit synergies and avoid redundant development. However, we observe that today, many funding proposals lack a thorough state-of-the-art report of software that could possibly be reused. This is most often caused by insufficient information retrieval strategies, lack of knowledge about relevant repositories, and an abundance of locations where software is collaboratively developed and stored79. Discovery requires publication in a globally accessible location with appropriate metadata, e.g. Citation File Format (CFF)80 and CodeMeta. Comprehensive metadata (e.g. contributors, contact, keywords, linked publications, etc.) and publishing platforms have to enable persistent citing, which in turn benefits research evaluation. Selection and curation of software (probably by a data/software librarian) for publication and discovery are certainly challenging.\n\nWe consider GitLab or GitHub as collaborative working environments and repositories like Zenodo appropriate publication platforms, because the latter mint DOIs, allow versioning and are publicly funded for long-term access. GitHub, Figshare, and Mendeley Data are examples of commercial enterprises with business cases in the background, which leverage research results. Besides the aforementioned metadata standards, it is advisable to document source code, e.g. using MarkDown (with Doxygen tooling). Metadata and citations play a role in beneficial tools like PIDgraph, DataCite.org, CrossRef, which utilize Persistent Identifiers (PIDs) like DOIs. Another solution to discovery are (mostly) disciplinary software indices like swMATH or the Astronomy Source Code Library as well as language focused systems like CRAN for R. Most of them started as national endeavors and became platforms of global importance. For Germany, we assume that the Nationale Forschungsdateninfrastruktur (NFDI) will put effort into creating or supporting discovery platforms at a central point that ease information retrieval. At the same time, all stakeholders should be aware of and counteract potential institutional “fear” of losing “their” data, software, and intellectual property.\n\nEspecially in interdisciplinary environments, it would be helpful to have access to a meta software repository index, similar to what re3data81 does for research data repositories. We recommend the creation of such a meta index covering important (disciplinary) software indexes in order to ease discovery of relevant software locations. Evaluation of discovered software is an unsolved problem. Here, anonymous telemetry of usage may provide information for the selection of relevant software. Publishing software, their dependencies, and environment in containers may also ease evaluation and further reuse. These suggestions require significant investment in longterm infrastructure. When publishing research software it is recommended to make use of integration schemes like GitHub with Zenodo or local GitLab instances with publication platforms. Such indices and publication outlets may benefit national federated research indexing & archiving systems, similar to the hierarchy of library catalogs82.\n\nSoftware preservation aims to extend the lifetime of software that is no longer actively maintained. There are different approaches, which vary in the effort required and the likelihood of success. Software archiving is one important aspect of software preservation: the process of storing a copy of a software so that it may be referred to in the future. The publication of a certain software version for reference in research articles requires simple ways to archive research software on a long-term basis. Furthermore, its integration with collaborative software development environments such as GitLab or GitHub and with publication repositories is needed to facilitate archiving of referenced software versions based on sustainable frameworks (e.g. Invenio for GitHub to Zenodo integration).\n\nA challenge for software archiving is the need to (ideally) preserve the runtime environment and all dependencies of the software. This could improve reproducibility, especially when running the software in its original state. If research data are needed to reproduce results, they should also be archived with the software or the publication. Specialized and unique hardware like high performance computing resources can be part of the runtime environment, which may not be accessible in the future. To overcome this, an emulation of hardware may be a (challenging) solution. Emulation involves the encapsulation and distribution of the complete hardware and software stacks, including the operating system and driver interdependencies. This can result in intellectual property issues when offered as a service.\n\nThere are both local and global approaches to software conservation. One solution to keep the software in an executable state by preserving its context and runtime environment is to use containers such as Docker. However, to archive the Docker containers, additional metadata should be added and stored with the software in an archive container format that allows exchange between repositories and exit strategies, such as the BagIt container format83. Application or platform conservation is also achieved by conservational efforts where unmaintainable (virtual) machines are sandboxed to keep the platform in a secure but running state. Another threat is losing project repositories on global platforms like Github or BitBucket. Here, global platforms like Software Heritage harvest those repositories and prevent loss by long-term archiving.\n\n\nLegal aspects\n\nMore and more industrial partners enter the hal9k community, and they bring their lawyers. Together with UofA’s research software task force, the RSE team, the researchonomy institute, the corporate lawyers, and community representatives, Kim decides to create a foundation to govern hal9k and its environment: the Fullest Possible Use Foundation for Open Researchonomy, funded by the Ministry of Research and Education and a consortium of corporate partners. As a first step, they re-license hal9k under the OSI-approved MIT license.\n\nA common situation in research software creation is that the developer has no knowledge or awareness of legal aspects and therefore did not consider them early enough. Thus, we think the main legal demands for research software development are raising awareness and empowering all levels of responsible persons in academia (from researcher and RSEs over PIs to research performing organizations and research funding organizations) in legal aspects. This will hopefully lead to a general legal certainty before, during, and after the research software development process and thus enable better options for collaborations between universities, non-commercial research institutions, and other national or international partners. Legal aspects always have to be considered regarding the relevant jurisdiction. Though similar issues arise in all jurisdictions, the following will focus on the European and specifically German legal framework.\n\nSoftware development is a creative activity. The main relevant law governing legal aspects is therefore the copyright law. It regulates the rights and obligations of the parties involved. Chapter 8 of the German Act on Copyright and Related Rights (UrhG) contains specific provisions applicable to computer programs and is based on the EU computer programs directive. Copyright law protecting the creator of software in similar ways exist in nearly all legal systems. It is important for the identification of rights that software, in the sense of (German) law, includes not only the source code but also the design materials84. The challenge in the use, distribution, and commercialization of software is to determine the chain of rights and to identify all right holders. The owner of the copyright is not necessarily the owner of the right of use. For Germany, the Copyright Act regulates the rights for employment relationships85. In such cases, the right of use is automatically transferred to the employer. This means that in most cases of employed software developers and research staff, the institution holds the rights of use for the software work. This is not automatically the case for students, freelancers, and individual external cooperation partners. Employment and service contracts with contributors could contain regulations regarding the transfer of rights of use. For researchers who conduct free research not subject to directives, in Germany the constitution guarantees freedom of research so that the rights of use for their work remains initially with the natural person. In addition to the rights of the people directly involved, other rights of third parties may also be relevant. Existing source code (e.g., other Free/Libre Open Source Software (FLOSS)), external libraries, and contributions from institutional cooperation partners are published and provided under certain licenses and their conditions must be observed (which, due to incompatibilities even among FLOSS licenses, may well mean that individually reusable pieces of software cannot be reused together or in a new context). The nature of research careers often brings additional complications to the chain of rights. It happens that researchers take their software with them when they change institutions and develop it further during their career. Here, the former employer may be entitled to some rights of use. In third-party funded projects, in particular with industry but also with public funding, rules regarding rights of use are often defined. Last but not least, the software can also be affected by other (intellectual) property rights such as patents or trademarks. Software itself is usually not patentable but it may implement a technical invention covered by patents. When using or distributing such software, an additional matching patent license may be necessary. Licenses exist (for example: GNU GPL v3) which automatically grant related patent licenses while using the software license. That should be considered when exploitation of the patent is planned.\n\nIssues of warranty and liability for faulty software must be taken into account. We consider the possibilities of contractual limitation of liability in licenses. Full exclusions of liability are generally invalid in the German law. Limitations of liability usually depend on the form of distribution: The limitation options are larger if the rights of use are granted free of charge, e.g. provision “as is” as defined in the BSD 3-clause license.\n\nIn order to meet the legal challenges mentioned, it is absolutely necessary for the software developer (team) to document the rights chain comprehensively during the software development (see e.g. Figure 3). Contributions of individual persons must be traceable and their (labor law) status must be named. At best, contracts with rules on the transfer of rights of use should be concluded before work begins. Declarations of assignment of rights can be made for existing works. License conditions for external contributions must be evaluated with regard to further rights of use and possible sub-licensing. Contracts and funding conditions must be conscientiously documented and analyzed with regard to rules on rights of use. In case that different parts of the software are based on different conditions and rights of third parties, individual modules of the new software could be published under different licenses and merged accordingly.\n\nA national research software sustainability institute could be established. This institute supports local research software task forces and thereby respective researchers and research teams in the licensing of research software and related legal issues. For this purpose, a legal help desk will be set up, to which all members of their respective research performing organization can apply. If researchers want to publish the research software under a Free/Libre Open Source Software (FLOSS) license, the organization could bundle the necessary rights beforehand. This is particularly useful when teams of researchers, often international, write software. In addition, the sustainability institute may serve as a one-stop-shop for the licensing of research software.\n\nWe see it as an essential part of the sustainability of research to enable the free distribution of research software. There are a variety of open source software licensing models (ranging from permissive to copyleft; for further information, see tldrlegal, the ifrOSS Lizenz-Center, or Morin et al., 201245). The use of an FSF- or OSI-approved FLOSS license for example would enable a truly free model and also reduce legal issues. We recommend that research funding organizations such as the DFG discuss if they expect publishing all funded software under these licenses, following the paradigm of “public money, public code”.\n\nAlso for legal aspects, we believe it is important that all (German) research performing organizations install a research software task force, especially in light of the new DFG Code of Conduct. Besides organization and bundling of technical and infrastructural support for local RSEs and researchers (see previous sections), this group should organize a local legal help desk, organize educational offers e.g. for the legal topics presented, and (if not implemented yet) develop the software policy of the research performing organization. As an example, with the help of on-boarding processes performed by the research software task force, RSEs should be able to keep the clearance of rights as simple as possible right from the start. One possibility how local legal help desks could structure their work is shown in the decision trees in Figure 1–Figure 4. We suggest that the local task forces build a network with the other research performing organizations for exchange of ideas but also for generating a bottom-up strategy to organize RSE standards for Germany and beyond and possibly be the origin of the aforementioned software sustainability institute.\n\nThis tree helps to figure out whether the academic institution where the software is developed owns the intellectual property (copyright).\n\nThis legal decision tree recommends to check closely any policies implemented in the software developers organization.\n\nThe code history decision tree points out tasks for projects that incorporate existing code.\n\nDepending on the distribution model, open access (OA) or open source software can be selected.\n\n\nConclusions\n\nWe find that the research software ecosystem is notoriously lacking resources despite its strategic importance. If funding and support does not improve, the success story of science based on academic research software may be at stake. We recommend the installation of infrastructure that enables sustainable software development including platforms for collaboration, continuous integration, testing, discovery, and long-term preservation. We suggest the establishment of a nationwide institution similar to the Software Sustainability Institute (SSI) to provide project consulting and code review services as well as sustainable software development training. We think that sustainable software development should become an integral component of the universities’ teaching curriculum. We encourage the research funding bodies to reflect the licensing models for academic software development, and to decide whether the “public money, public code” paradigm justifies the requirement that all publicly funded software has to be publicly available under a Free/Libre Open Source Software (FLOSS) license. Ultimately, we strongly advise the implementation of funding schemes for sustainably supporting the development and maintenance of research software based on clear and transparent criteria, for creating incentives to produce high quality community software, and for enabling career paths as research software engineer (RSE).\n\n\nGlossary\n\ndomain researchers The people doing the research to advance knowledge in a field.\n\ngeneral public Lay people that do not necessarily have specific insight regarding a research domain.\n\ngeopolitical units Governed public units, ranging from cities and councils, over federal states and countries, up to political unions such as the EU. In the context of this paper, the discussion usually focuses on the larger units (countries and political unions).\n\nindependent (open source) developers Project-external software developers who are not employed by the institution(s) carrying out the project.\n\nindustry Companies conducting research or profit from available academic research software which they can directly or indirectly apply to their field.\n\ninfrastructure units Computing centers of research bodies such as universities and other research centers, as well as high-performance computing facilities.\n\nlibraries (also registries, indices) Infrastructure units of research bodies such as universities, or independent organizations, which gather research outputs and their structured metadata, and provide indices, search, etc.\n\nresearch funding organizations Public research funding bodies but potentially also companies, foundations, associations, etc.\n\nresearch leaders Heads of research groups, such as professors and other people with staff responsibility.\n\nresearch performing organizations Research groups, departments, faculties, research institutions (universities, research institutions, cross-institutional research groups, etc.), umbrella organizations, such as Helmholtz-Gemeinschaft Deutscher Forschungszentren, Max-Planck-Gesellschaft zur Förderung der Wissenschaften, Leibniz-Gemeinschaft, etc.\n\nresearch software engineers (RSEs) People creating and maintaining research software; this group ranges from research-focused software developers, to software engineers with a focus on research; other definitions include other roles, such as research software managers.\n\n\nData availability\n\nNo data are associated with this article.\n\n\nAuthor information\n\nWe are a group of software-providing researchers, RSEs, and infrastructural as well as legal supporters. Initially, a group of representatives of funded projects of funded projects of the first DFG sustainability call met during the first German RSE conference (deRSE19) in June 2019 in a grass-roots workshop on sustainable research software addressing the software-based research community. During this workshop, we realized that a lot of valuable experience and good ideas are present in the group, and we decided to start working on this paper together with other interested practitioners. We followed the generous invitation of the DFG for the above-mentioned two-day meeting at the Robert Koch Institute in Berlin in November 2019 to sharpen the focus of this paper.",
"appendix": "References\n\nThe Event Horizon Telescope Collaboration, Akiyama K, Alberdi A, et al.: First M87 Event Horizon Telescope Results. IV. Imaging the Central Supermassive Black Hole. Astrophys J. 2019; 875(1): L4. Publisher Full Text\n\nNowogrodzki A: How to support open-source software and stay sane. Nature. 2019; 571(7763): 133–134. PubMed Abstract | Publisher Full Text\n\nPhilippe O, Hammitzsch M, Janosch S, et al.: softwaresaved/international-survey: Public release for 2018 results. 2019. Publisher Full Text\n\nHirsch JE: An index to quantify an individual's scientific research output. Proc Natl Acad Sci U S A. 2005; 102(46): 16569–16572. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBangerth W, Heister T: Quo Vadis, Scientific Software? SIAM News. 2014; 47(1): 8. Reference Source\n\nPrins P, de Ligt J, Tarasov A, et al.: Toward effective software solutions for big biology. Nat Biotechnol. 2015; 33(7): 686–687. PubMed Abstract | Publisher Full Text\n\nRichardson C, Croucher M: Research Software Engineer: A New Career Track? 2018. Reference Source\n\nBrett A, Croucher M, Haines R, et al.: Research Software Engineers: State of the Nation Report 2017. 2017. Publisher Full Text\n\nWilkinson MD, Dumontier M, Aalbersberg IJ, et al.: The FAIR Guiding Principles for scientific data management and stewardship. Sci Data. 2016; 3(1): 160018. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHasselbring W, Carr L, Hettrick S, et al.: FAIR and Open Computer Science Research Software. arXiv: 1908.05986. 2019. Reference Source\n\nLamprecht AL, Garcia L, Kuzak M, et al.: Towards FAIR principles for research software. Data Sci. 2019. Publisher Full Text\n\nKaterbow M, Feulner G: Recommendations on the development,use and provision of Research Software. 2018. Publisher Full Text\n\nScheliga K, Pampel H, Konrad U, et al.: Dealing with research software:Recommendations for best practices. Helmholtz Open Science Coordination Office. 2019. Publisher Full Text\n\nHatton L: The Chimera of Software Quality. Computer. 2007; 40(8): 104–103. Publisher Full Text\n\nChang G, Roth CB, Reyes CL, et al.: Retraction. Science. 2006; 314(5807): 1875–1875. PubMed Abstract | Publisher Full Text\n\nMatthews BW: Five retracted structure reports: inverted or incorrect? Protein Sci. 2007; 16(6): 1013–1016. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmith AM, Katz DS, Niemeyer KE, et al.: Software Citation Principles. PeerJ Comput Sci. 2016; 2: e86. Publisher Full Text\n\nMerali Z: Computational science: ...Error. Nature. 2010; 467(7317): 775–777. PubMed Abstract | Publisher Full Text\n\nBarnes N: Publish your computer code: it is good enough. Nature. 2010; 467(7317): 753. PubMed Abstract | Publisher Full Text\n\nTse H: Computer code: more credit needed. Nature. 2010; 468(7320): 37. PubMed Abstract | Publisher Full Text\n\nHafer L, Kirkpatrick AE: Assessing Open Source Software As a Scholarly Contribution. Commun ACM. 2009; 52(12): 126–129. Publisher Full Text\n\nHowison J, Bullard J: Software in the Scientific Literature: Problems with Seeing, Finding, and Using Software Mentioned in the Biology Literature. J Assoc Inf Sci Technol. 2016; 67(9): 137–2155. Publisher Full Text\n\nLi K, Yan E, Feng Y: How Is R Cited in Research Outputs? Structure, Impacts, and Citation Standard. J Informetr. 2017; 11(4): 989–1002. Publisher Full Text\n\nLi K, Chen PY, Yan E: Challenges of measuring software impact through citations: An examination of the lme4 R package. J Informetr. 2019; 13(1): 449–461. Reference Source\n\nPark H, Wolfram D: Research software citation in the Data Citation Index: Current practices and implications for research software sharing and reuse. J Informetr. 2019; 13(2): 574–582. Publisher Full Text\n\nPan X, Yan E, Cui M, et al.: How Important Is Software to Library and Information Science Research? A Content Analysis of Full-Text Publications. J Informetr. 2019; 13(1): 397–406. Publisher Full Text\n\nDoerr A, Rusk N, Vogt N, et al.: Giving Software Its Due. Nat Methods. 2019; 16(3): 207–207. PubMed Abstract | Publisher Full Text\n\nDruskat S: Software and Dependencies in Research Citation Graphs. Comput Sci Eng. 2020; 22(2): 8–21. Publisher Full Text |Free Full Text\n\nKatz DS, Bouquin D, Hong NP, et al.: Software Citation Implementation Challenges. arXiv: 1905.08674. 2019. Reference Source\n\nResnik DB, Morales M, Landrum R, et al.: Effect of impact factor and discipline on journal data sharing policies. Account Res. 2019; 26(3): 139–156. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVandewalle P: Code Sharing Is Associated with Research Impact in Image Processing. Comput Sci Eng. 2012; 14(4): 42–47. Publisher Full Text\n\nVenters CC, Jay C, Lau LMS, et al.: Software Sustainability: The Modern Tower of Babel. In: Proceedings of the Third International Workshop on Requirements Engineering for Sustainable Systems Co-Located with 22nd International Conference on Requirements Engineering (RE 2014) vol. 1216 Karlskrona, Sweden: CEUR-WS. 2014; 7–12. Reference Source\n\nGoble C: Better Software, Better Research. IEEE Internet Comput. 2014; 18(5): 4–8. Publisher Full Text\n\nDruskat S: A Proposal for the Measurement and Documentation of Research Software Sustainability in Interactive Metadata Repositories. In Proceedings of the Fourth Workshop on Sustainable Software for Science: Practice and Experiences (WSSSPE4), vol. 1686 Manchester, UK: CEUR-WS, 2016. Reference Source\n\nKatz DS: Fundamentals of Software Sustainability. 2018. Reference Source\n\nAkhmerov A, Cruz M, Drost N, et al.: Raising the Profile of Research Software: Recommendations for Funding Agencies and Research Institutions. NWO (The Netherlands Organisation for Scientific Research). 2019. Reference Source\n\nCasties R, Czmiel A, Damerow J, et al.: DH Research Software Engineers - For We Are Many. 2019. Reference Source\n\nWilson G, Aruliah DA, Brown CT, et al.: Best practices for scientific computing. PLoS Biol. 2014; 12(1): e1001745. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStodden V, Miguez S: Best Practices for Computational Science: Software Infrastructure and Environments for Reproducible and Extensible Research. J Open Res Softw. 2014; 2(1): e21. Publisher Full Text\n\nWilson G, Bryan J, Cranston K, et al.: Good enough practices in scientific computing. PLoS Comput Biol. 2017; 13(6): e1005510. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi K, Lin X, Greenberg J: Software Citation, Reuse and Metadata Considerations: An Exploratory Study Examining LAMMPS. Proc Assoc Infor Sci Tech. 2016; 53(1): 1–10. Publisher Full Text\n\nKuchinke W, Ohmann C, Stenzhorn H, et al.: Ensuring sustainability of software tools and services by cooperation with a research infrastructure. Per Med. 2016; 13(1): 43–55. PubMed Abstract | Publisher Full Text\n\nLoewe A, Seemann G, Wülfers EM, et al.: SuLMaSS - Sustainable Lifecycle Management for Scientific Software. E-Science-Tage 2019: Data to Knowledge. 2019. Publisher Full Text\n\nDruskat S, Krause T, Lüdeling A, et al.: Infrastrukturstrategien für nachhaltige Forschungssoftware in befristeten Projekten. deRSE19 - Conference for Research Software Engineers in Germany. Potsdam, Germany. 2019. Publisher Full Text\n\nMorin A, Urban J, Sliz P: A Quick Guide to Software Licensing for the Scientist-Programmer. PLoS Comput Biol. 2012; 8(7): e1002598. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKatz DS, Ramnath R: Looking at Software Sustainability and Productivity Challenges from NSF. 2015. arXiv: 1508.03348.\n\nDFG: Nachhaltigkeit von Forschungssoftware. 2016. Reference Source\n\nDFG: Qualitätssicherung von Forschungssoftware durch ihre nachhaltige Nutzbarmachung. 2019. Reference Source\n\nChan Zuckerberg Initiative: Essential Open Source Software for Science. Reference Source\n\nRodríguez-Sánchez F, Marwick B, Lazowska E, et al.: Academia's failure to retain data scientists. Science. 2017; 355(6323): 357–358. PubMed Abstract | Publisher Full Text\n\nKatz DS, Druskat S, Haines R, et al.: The State of Sustainable Research Software: Learning from the Workshop on Sustainable Software for Science: Practice and Experiences (WSSSPE5.1). J Open Res Softw. 2019; 7(1): 11. Publisher Full Text\n\nDruskat S, Katz DS: Mapping the Research Software Sustainability Space. In: 2018 IEEE 14th International Conference on E-Science (e-Science). 2018; 25–30. Publisher Full Text\n\nYe Y, Boyce RD, Davis MK, et al.: Open Source Software Sustainability Models: Initial White Paper from the Informatics Technology for Cancer Research Sustainability and Industry Partnership Work Group. 2019. Reference Source\n\nHong NC: Minimal information for reusable scientific software. In: 2nd Workshop on Sustainable Software for Science: Practice and Experiences (WSSSPE2). 2014. Publisher Full Text\n\nSchlauch T, Meinel M, Haupt C: DLR Software Engineering Guidelines. Deutsches Zentrum für Luft- und Raumfahrt (DLR). 2018. Publisher Full Text\n\nGomez-Diaz T, Recio T: On the evaluation of research software: the CDUR procedure [version 2; peer review: 2 approved]. F1000Res. 2019; 8: 1353. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFehr J, Heiland J, Himpe C, et al.: Best practices for replicability, reproducibility and reusability of computer-based experiments exemplified by model reduction software. AIMS Mathematics. 2016; 1(3): 261–281. Publisher Full Text\n\nJiménez RC, Kuzak M, Alhamdoosh M, et al.: Four simple recommendations to encourage best practices in research software [version 1; peer review: 3 approved]. F1000Res. 2017; 6: pii: ELIXIR-876. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHsu L, Hutchison VB, Langseth ML: Measuring sustainability of seed-funded earth science informatics projects. PLoS One. 2019; 14(10): e0222807. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJackson M, Crouch S, Baxter R: Software Evaluation Guide. 2019. Reference Source\n\nrOpenSci; Anderson B, Chamberlain S, et al.: Software Peer Review, Why? What? In: rOpenSci Packages: Development, Maintenance, and Peer Review Zenodo. 2019. Publisher Full Text\n\nLee BD: Ten simple rules for documenting scientific software. PLoS Comput Biol. 2018; 14(12): e1006561. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVenters CC, Capilla R, Betz S, et al.: Software sustainability: Research and practice from a software architecture viewpoint. J Syst Software. 2018; 138: 174–188. Publisher Full Text\n\nSSI: Writing and using a software management plan. 2019. Reference Source\n\nJohanson A, Hasselbring W: Software engineering for computational science: Past, present, future. Comput Sci Eng. 2018; 20(2): 90–109. Publisher Full Text\n\nAkhmerov A, Cruz M, Drost N, et al.: Making Research Software a First-Class Citizen in Research. 2019. Publisher Full Text\n\nChue Hong N: Making Software A First-Class Citizen. 2019. Publisher Full Text\n\nVereinigung der Kanzlerinnen und Kanzler der Universitäten Deutschlands: Bayreuther Erklärung zu befristeten Beschäftigungsverhältnissen mit wissenschaftlichem und künstlerischem Personal in Universitäten. 2019. Reference Source\n\nKatz DS, McInnes LC, Bernholdt DE, et al.: Community Organizations: Changing the Culture in Which Research Software Is Developed and Sustained. Comput Sci Eng. 2019; 21(2): 8–24. Publisher Full Text\n\nScience Guide: Room for everyone’s talent. 2019. Reference Source\n\nCarver JC, Hong NPC, Thiruvathukal GK: Software engineering for science. CRC Press, 2016; 274. Reference Source\n\nIaffaldano G, Steinmacher I, Calefato F, et al.: Why do developers take breaks from contributing to OSS projects? A preliminary analysis. arXiv: 1903.09528. 2019. Reference Source\n\nAllen A, Aragon C, Becker C, et al.: Engineering Academic Software (Dagstuhl Perspectives Workshop 16252). Dagstuhl Manifestos 2017; 6(1): 1–20. Reference Source\n\nBundesministerium der Justiz und für Verbraucherschutz: Gesetz über befristete Arbeitsverträge in der Wissenschaft. 2017. Reference Source\n\nChang V, Mills H, Newhouse S: From Open Source to long-term sustainability: Review of Business Models and Case studies. In: Proceedings of the UK e-Science All Hands Meeting 2007 University of Edinburgh/University of Glasgow (acting through the NeSC) 2007. Reference Source\n\nAartsen W, Peeters P, Wagers S, et al.: Getting Digital Assets from Public-Private Partnership Research Projects through “The Valley of Death,” and Making Them Sustainable. Front Med (Lausanne). 2018; 5: 65. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGabella C, Durinx C, Appel R: Funding knowledgebases: Towards a sustainable funding model for the UniProt use case [version 2; peer review: 3 approved]. F1000Res. 2018; 6: pii: ELIXIR-2051. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIson J, Rapacki K, Ménager H, et al.: Tools and data services registry: a community effort to document bioinformatics resources. Nucleic Acids Res. 2016; 44(D1): D38–D47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStruck A: Research Software Discovery: An Overview. In: 2018 IEEE 14th International Conference on e-Science IEEE 2018. Publisher Full Text\n\nDruskat S, Spaaks JH, Chue Hong N, et al.: Citation File Format (CFF) - Specifications. 2019. Publisher Full Text\n\nre3data.org – Registry of Research Data Repositories. Publisher Full Text\n\nMönnich MW: KVK - a meta catalog of libraries. LIBER Quarterly. 2001; 11(2): 121–127. Publisher Full Text\n\nKunze J, Scancella J, Adams C, et al.: The bagIt file packaging format (v1. 0). RFC Editor. 2018; 8493.\n\nBundesministerium der Justiz und für Verbraucherschutz, § 69a subsection (1) UrhG. 2014. Reference Source\n\nBundesministerium der Justiz und für Verbraucherschutz, § 69b UrhG. 2014. Reference Source"
}
|
[
{
"id": "62873",
"date": "13 May 2020",
"name": "Wilhelm Hasselbring",
"expertise": [
"Reviewer Expertise Software Engineering"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors describe the state of the practice and current challenges for research software sustainability and suggest measures towards improvements that can solve these challenges. In particular, they propose to fund a German Software Sustainability Institute. The paper is the result of a community effort, with work undertaken during two workshops and subsequent collaborative work across the larger RSE community in Germany.\n\nThe UK Software Sustainability Institute has already been established during a decade (https://www.software.ac.uk/blog/2020-05-05-impact-institute-10-years). Thus, the idea of such an institute is not new, but it makes sense to take a specific look at the German situation. Besides universities, the German states (local and in particular federal) fund significant large-scale research associations (Helmholtz/DLR, Max-Planck, Leibniz). This is not the case for most other European states, at least not with a similar scale. Another specialty is the lack of long-term funding for research software engineers, as discussed by the authors.\n\nThe paper is well-written and easy to read. I like the boxed story of Kim’s career path.\n\nHowever, I’ve some suggestions for improving the paper:\nConcerning the statement “In order to support research, a sustainable software must be correct”, I suggest to include a short discussion of the test oracle problem for scientific software (see for instance https://doi.org/10.1109/SECSE.2013.66150991).\n\nConcerning the discussion of “The list of criteria presented in this section could be the basis for a structured review process…” I suggest to include two additional initiatives for software review. The first is artifact evaluation in computer science conferences (the process is explained in https://doi.org/10.1515/itit-2019-00402). The second is the SPEC Research Group’s review process of tools for quantitative system evaluation and analysis (https://research.spec.org/tools/submission.html).\nThe authors write “We also argue that truly sustainable research software must ideally be published under a Free/Libre Open Source Software (FLOSS) license, and follow an open development model…” what I fully support (see for instance https://doi.org/10.1515/itit-2019-00402). However, later under the section heading “Legal aspects” this requirement is thwarted. I fully agree that legal aspects have to be considered, but the general bias of this section seems to be on commercial licensing of research software. For instance, the decision tree in Figure 1 starts with the question “Licensing planned?”. I assume that commercial licensing is meant, but this is not clear since the figures are not explained in the paper. Instead, the process should start with open sourcing the software. If licenses such as Apache or MIT are applied, the research institutions may later still commercialize the software if appropriate. Such open source licensing is also beneficial for start-ups, that intend to provide professional services for the software.\n\nMy experience with technology transfer units of German universities and research institutes is that they do not understand the ideas of open source business models (see for instance https://doi.org/10.1109/MC.2019.28981633). Their focus is on patents and commercializing licenses, sometimes also on start-ups. Conversely, in the software industry, one major motivation for open sourcing software is on improving the quality of software. I cite from https://doi.org/10.1109/ICSAW.2017.114 : “the open-source approach has some psychological effects: Developers show a tendency to apply higher quality standards if they know that the code will be publicly available.” For sustainability, quality is an important property of software.\n\nThe Figures 1-4 do more harm than good. They are daunting to researchers who intend to publish their code open source. These figures should be removed from the paper, they are useless without proper explanation.\nI suggest that the authors focus in the present paper on their main message (request for funding a German Software Sustainability Institute, which I fully support). Figures 1-4 could be moved to a separate paper, enriched with proper explanation.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly",
"responses": [
{
"c_id": "6209",
"date": "26 Jan 2021",
"name": "Axel Loewe",
"role": "Author Response",
"response": "Thank you for the thorough review and constructive feedback regarding our manuscript. Below, we address the issues raised by you point-by-point. Our responses are set in italics. Concerning the statement “In order to support research, a sustainable software must be correct”, I suggest to include a short discussion of the test oracle problem for scientific software (see for instance https://doi.org/10.1109/SECSE.2013.66150991). We thank the reviewer for this suggestion, and included a brief discussion of the test oracle problem as suggested, and additionally of further challenges to verification and validation, such as large configuration spaces and heterogeneous data (as discussed in e.g. https://doi.org/10.1109/SE4Science.2019.00010), and have suggested to implement the solutions mentioned in the literature. Concerning the discussion of “The list of criteria presented in this section could be the basis for a structured review process…” I suggest to include two additional initiatives for software review. The first is artifact evaluation in computer science conferences (the process is explained in https://doi.org/10.1515/itit-2019-00402). The second is the SPEC Research Group’s review process of tools for quantitative system evaluation and analysis (https://research.spec.org/tools/submission.html). We thank the reviewer for this suggestion and included the artifact review approach in the introduction to the criteria section. The aspects of repeatability, reproducibility, and replicability are aimed more at the results of computational research rather than research software itself, we feel. Therefore, we didn’t include specific criteria in the list suggested to be used when evaluating research software for sustained funding. While the SPEC submission process is very clear, we could not find any concrete criteria applied during the review (except for requirements regarding the license). The authors write “We also argue that truly sustainable research software must ideally be published under a Free/Libre Open Source Software (FLOSS) license, and follow an open development model…” what I fully support (see for instance https://doi.org/10.1515/itit-2019-00402). However, later under the section heading “Legal aspects” this requirement is thwarted. I fully agree that legal aspects have to be considered, but the general bias of this section seems to be on commercial licensing of research software. For instance, the decision tree in Figure 1 starts with the question “Licensing planned?”. I assume that commercial licensing is meant, but this is not clear since the figures are not explained in the paper. Instead, the process should start with open sourcing the software. If licenses such as Apache or MIT are applied, the research institutions may later still commercialize the software if appropriate. Such open source licensing is also beneficial for start-ups that intend to provide professional services for the software. My experience with technology transfer units of German universities and research institutes is that they do not understand the ideas of open source business models (see for instance https://doi.org/10.1109/MC.2019.28981633). Their focus is on patents and commercializing licenses, sometimes also on start-ups. Conversely, in the software industry, one major motivation for open sourcing software is on improving the quality of software. I cite from https://doi.org/10.1109/ICSAW.2017.114 : “the open-source approach has some psychological effects: Developers show a tendency to apply higher quality standards if they know that the code will be publicly available.” For sustainability, quality is an important property of software. As further detailed below, we have moved the decision trees out of this manuscript as we see the problems and agree to your arguments. We like your suggested aspect of commercialization of FLOSS licensed software and included this aspect in the manuscript. The Figures 1-4 do more harm than good. They are daunting to researchers who intend to publish their code open source. These figures should be removed from the paper, they are useless without proper explanation. I suggest that the authors focus in the present paper on their main message (request for funding a German Software Sustainability Institute, which I fully support). Figures 1-4 could be moved to a separate paper, enriched with proper explanation. We thank the reviewer for this suggestion. Our initial thought was to place these decision trees in the supplemental material but did not realize that this is not the policy of f1000. The editorial team moved them into the main article. This is the reason why the decision trees appeared without additional information in the manuscript. We have now decided to take the Figures out and have published them together with documentation templates in a separate report via Zenodo unde a Creative Commons license, see https://zenodo.org/record/4327148#.X9n6ui337OQ. This report is now cited in the F1000 manuscript."
}
]
},
{
"id": "62872",
"date": "01 Jun 2020",
"name": "Radovan Bast",
"expertise": [
"Reviewer Expertise computational chemistry",
"research software engineering"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn \"An environment for sustainable research software in Germany and beyond: current state, open challenges, and call for action\" the authors identify challenges for research software sustainability in Germany and beyond.\nThey examine the current state of research software sustainability and challenges in motivating sustainable research software development, selection criteria for funding, personnel, funding, infrastructure, and legal aspects, and offer recommendations for addressing these challenges. These sections are accompanied and with a story using a fictional character Kim which helps to relate these aspects to typical career stages of a research software engineer.\nThe article is thoroughly researched, well-written, and offers an excellent overview of the challenges when building an environment for sustainable research software. Most of the discussed challenges and recommendations carry beyond Germany and are relevant and transferable to other countries.\nBelow I give few (minor) suggestions for consideration when improving the manuscript.\nRegarding the list of challenges under \"Why sustainable research software in the first place?\" (pages 5 and 6):\nInfrastructure issues: One design choice that often limits the use or usability of local infrastructure resources is that they are often bound to institutional user accounts and thus limit collaboration possibilities with collaborators in other institutions and countries. On the other hand, pooling of infrastructure resources which could enable collaboration across organizations can be limited by lack of authentication and authorization infrastructure (AAI) or legal constraints. Later in the paper the authors indeed mention AAI (page 12) but this could already be pointed out and connected earlier.\n\nLegal issues: Not only licensing is an issue but legal constraints or uncertainty about legal boundaries and identity federation can also limit the deployment of infrastructure services. Often the deployment and operation of infrastructure services is given to technical teams who may lack the legal support or expertise to clarify legal and privacy terms for the storage of data and processing of data.\n\nFunding issues: The challenge is not only that funding is scarce but also that it does not align well with pricing models of cloud infrastructure providers. It can be easier for research groups to spend a larger chunk of the budget towards the end of a year for hardware compared to pay possibly relatively modest monthly fees for a cloud service, which however may not fit into the budget forms. These budget constraints may also limit the possibility of pooling resources and sharing them with other research groups. Software cloud infrastructure is often not considered at all in the proposal. There is also a resistance among some of my research colleagues to pay 20-50 USD/ month for an infrastructure service which is sometimes solved by reinventing the service locally \"for free\".\n\nAnother mismatch between traditional funding models and support of software which \"must continue to exist\" to be sustainable (page 5), is the experience that it can take months or years until the software is picked up by other groups and contributions and questions start to roll in. But by that time the funding of the project stopped, the developer (team) may have already moved on to other positions and projects, and may not have the time to react and help, even though they still may have interest and the knowledge. Our traditional funding models consider the software to be \"done\" by the end of the project.\nSelection criteria for \"How to decide which software to sustain?\" (page 9):\nThe authors mention \"usage and impact\", \"software quality\", as well as \"maturity\". But I would like to see also \"openness and transparency\" among these. The reason is that we can expect the research community to adapt to these or any metrics and we will over time observe what we measure. Any set of metrics could be criticized as to some extent being arbitrary but the advantage of including \"openness and transparency\" is that the community as whole would benefit from such a metric [Enrico Glerean, \"Responsible conduct of research and questionable research practices\", presentation, slide 471].\nRegarding \"Who sustains research software?\":\nThe authors discuss the lack of recognition and awareness, as well as lack of career opportunities. It is also about respect and I was happy to see the sentence: \"Not all researchers that think of themselves as RSEs pursue a faculty position as their main career goal.\" I have experienced that RSEs are sometimes regarded as those who somehow \"failed\" to obtain a faculty position whereas many RSEs have chosen this position over a faculty position because it was a better fit for their career goals. This misunderstanding can lead to a lack of respect towards this position and this career choice and can lead to excellent personnel leaving the academic environment towards commercial employment, possibly not primarily for financial reasons but sometimes to be more respected and recognized.\nArchiving and software preservation (page 13:\nThe authors mention Docker but also Singularity should be mentioned as a tool since it is getting traction in particular on many-user systems such as higher performance computing clusters.\nLegal aspects (page 14):\nRe-licensing is mentioned in the story box and the text starts by pointing out that licensing is often not considered early enough in the project. Indeed re-licensing later in the project can be not only legally, but also organizationally very tricky, in particular for projects which developed over many years and involved many contributors in different organizations. This could be pointed out in the text as additional motivation to consider these very early in the project.\n\nI very much like the recommendation of providing a legal help desk for research groups to avoid the problem that out of uncertainty and fear of making a legal mistake some research groups end up not choosing any license at all which may limit further reuse of the software.\n\nThe manuscript presents a decision tree for contributors (Figure 1) and also discusses contributor license agreements. It could be useful to point out that without clear policies or legal help desks, individuals or organizations may be hesitant to contribute to a project because they may not feel confident having enough knowledge or authority to sign such agreements and too many legal steps and question can also raise the barrier to contribute, in particular for smaller projects. Also here clear guidelines and a support desk can help removing these barriers.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": [
{
"c_id": "6210",
"date": "26 Jan 2021",
"name": "Axel Loewe",
"role": "Author Response",
"response": "We thank you for the thorough review and constructive feedback regarding the manuscript. Below, we address the issues raised by you point-by-point. Our responses are set in italics. Regarding the list of challenges under \"Why sustainable research software in the first place?\" (pages 5 and 6): Infrastructure issues: One design choice that often limits the use or usability of local infrastructure resources is that they are often bound to institutional user accounts and thus limit collaboration possibilities with collaborators in other institutions and countries. On the other hand, pooling of infrastructure resources which could enable collaboration across organizations can be limited by lack of authentication and authorization infrastructure (AAI) or legal constraints. Later in the paper the authors indeed mention AAI (page 12) but this could already be pointed out and connected earlier. We thank the reviewer for their suggestion, and have included the mentioned issues in the respective list in the section “Why sustainable research software in the first place?”. Legal issues: Not only licensing is an issue but legal constraints or uncertainty about legal boundaries and identity federation can also limit the deployment of infrastructure services. Often the deployment and operation of infrastructure services is given to technical teams who may lack the legal support or expertise to clarify legal and privacy terms for the storage of data and processing of data. We thank the reviewer for their suggestion, and have included the mentioned issues in the respective list in the section “Why sustainable research software in the first place?” Funding issues: The challenge is not only that funding is scarce but also that it does not align well with pricing models of cloud infrastructure providers. It can be easier for research groups to spend a larger chunk of the budget towards the end of a year for hardware compared to pay possibly relatively modest monthly fees for a cloud service, which however may not fit into the budget forms. These budget constraints may also limit the possibility of pooling resources and sharing them with other research groups. Software cloud infrastructure is often not considered at all in the proposal. There is also a resistance among some of my research colleagues to pay 20-50 USD/ month for an infrastructure service which is sometimes solved by reinventing the service locally \"for free\". Another mismatch between traditional funding models and support of software which \"must continue to exist\" to be sustainable (page 5), is the experience that it can take months or years until the software is picked up by other groups and contributions and questions start to roll in. But by that time the funding of the project stopped, the developer (team) may have already moved on to other positions and projects, and may not have the time to react and help, even though they still may have interest and the knowledge. Our traditional funding models consider the software to be \"done\" by the end of the project. We thank the reviewer for these two comments. To address them, we have extended the “Funding issues” list item with a discussion of these issues. Selection criteria for \"How to decide which software to sustain?\" (page 9): The authors mention \"usage and impact\", \"software quality\", as well as \"maturity\". But I would like to see also \"openness and transparency\" among these. The reason is that we can expect the research community to adapt to these or any metrics and we will over time observe what we measure. Any set of metrics could be criticized as to some extent being arbitrary but the advantage of including \"openness and transparency\" is that the community as whole would benefit from such a metric [Enrico Glerean, \"Responsible conduct of research and questionable research practices\", presentation, slide 471]. We thank the reviewer for this suggestion and realized that indeed most aspects in this section (6-10, 12, 13) are actually related to openness and transparency. Therefore, we changed the title of this section to “Software transparency and quality”. Regarding \"Who sustains research software?\": The authors discuss the lack of recognition and awareness, as well as lack of career opportunities. It is also about respect and I was happy to see the sentence: \"Not all researchers that think of themselves as RSEs pursue a faculty position as their main career goal.\" I have experienced that RSEs are sometimes regarded as those who somehow \"failed\" to obtain a faculty position whereas many RSEs have chosen this position over a faculty position because it was a better fit for their career goals. This misunderstanding can lead to a lack of respect towards this position and this career choice and can lead to excellent personnel leaving the academic environment towards commercial employment, possibly not primarily for financial reasons but sometimes to be more respected and recognized. We fully agree with the reviewer and thank them for the renewed confirmation that this is seen as problematic not only by the authors. Archiving and software preservation (page 13): The authors mention Docker but also Singularity should be mentioned as a tool since it is getting traction in particular on many-user systems such as higher performance computing clusters. We thank the reviewer for this suggestion and now also mention Singularity and GUIX. However, we are not aiming for an exhaustive list, as options change dynamically and might even be specific to certain research communities. Legal aspects (page 14): Re-licensing is mentioned in the story box and the text starts by pointing out that licensing is often not considered early enough in the project. Indeed re-licensing later in the project can be not only legally, but also organizationally very tricky, in particular for projects which developed over many years and involved many contributors in different organizations. This could be pointed out in the text as additional motivation to consider these very early in the project. We thank the reviewer to point this out and it also nicely fits into the message of increasing the awareness of legal aspects early on in the project. We have added the suggested sentence in the manuscript. I very much like the recommendation of providing a legal help desk for research groups to avoid the problem that out of uncertainty and fear of making a legal mistake some research groups end up not choosing any license at all which may limit further reuse of the software. Thank you for supporting this idea. We have further included your idea of avoiding any license out of a fear to make legal mistakes. The manuscript presents a decision tree for contributors (Figure 1) and also discusses contributor license agreements. It could be useful to point out that without clear policies or legal help desks, individuals or organizations may be hesitant to contribute to a project because they may not feel confident having enough knowledge or authority to sign such agreements and too many legal steps and question can also raise the barrier to contribute, in particular for smaller projects. Also here clear guidelines and a support desk can help removing these barriers. We have decided to take the decision trees out of the manuscript to strengthen our point of publishing software under a FLOSS license. Instead, we published the decision trees together with documentation templates under a Creative Commons license via Zenodo: https://zenodo.org/record/4327148#.X9n6ui337OQ. In order to strengthen the point you addressed, we added some more details related to infrastructural investment."
}
]
}
] | 1
|
https://f1000research.com/articles/9-295
|
https://f1000research.com/articles/9-1231/v1
|
13 Oct 20
|
{
"type": "Review",
"title": "Dysplasia, malformation, or deformity? - explanation of the basis of hip development disorders and suggestions for future diagnostics and treatment",
"authors": [
"Jacek Dygut",
"Monika Piwowar",
"Jacek Dygut"
],
"abstract": "This publication focuses on processes which disrupt proper development of the hip. Four pathomechanisms underlying human developmental defects are described in literature, i.e. dysplasia, malformation, disruption, and deformity. In the case of hip development, arguably the greatest challenge involves confusion between dysplasia and deformity, which often leads to misdiagnosis, incorrect nomenclature, and incorrectly chosen treatment. The paper presents a description of hip joint development disorders in the context of their pathomechanisms. An attempt was made to answer the question whether these disorders are rooted in a primary disorder of tissue growth, resulting in its incorrect anatomy, or are the result of anatomical deformation with secondary modifications in tissue structures of a degenerative or adaptive nature, based on Deplesch-Heuter-Volkmann growth and remodeling laws. In addition, emphasis is placed on attention to the presence of the so-called clinically and diagnostically mute cases. The need to augment diagnostic procedures with genetic tests in order to increase the sensitivity of screening has also been suggested. Based on the arguments presented in the paper, a new division of developmental hip disorders has been proposed.",
"keywords": [
"dysplasia",
"malformation",
"deformity",
"hip development disorder"
],
"content": "Introduction\n\nMany studies of hip development disorders have been performed1, leading to a number of proposed diagnostic and treatment options2. Despite the impressive collection of studies documented in scientific publications, the etiology of developmental disorders of the hip joint still cannot be formulated with clarity1,3–5. Even so, the etiology of these diseases is known to be multifactorial, combining genetic factors (with varying intensity and expression periods) and environmental factors affecting the fetal and postnatal periods6,7. Normal hip growth and development depend on a genetically determined balance between the growth of the acetabular and triradiate cartilages and a properly located and centered femoral head6,8,9.\n\nExperimental studies have revealed that the development of the acetabulum depends on a coded geometric pattern10,11. The concave shape of the acetabulum results from the presence of a round femoral head inside. In addition, many other factors affect the acetabular depth, including growth within the acetabular cartilage, growth through apposition under the perichondrium layer, and growth of adjacent bones (iliac, sciatic and pubic)8,9.\n\nThe incidence of developmental hip growth disorders varies by population. It is close to 0% among newborns in China and Africa, but rises to 1% in Caucasian newborns, with the incidence of hip dislocations at approximately 0.1%12. Notably, these differences may result from environmental factors, such as the manner of childcare, rather than genetic issues. A positive family history of hip development disorders is found in 12–33% of patients and is more often observed in female children (80% of cases)12.\n\nAs described in literature, three independent but equivalent pathomechanisms underlying hip joint development disorders, i.e. malformation, dysplasia, and deformity, can be identified4. The fourth pathomechanism of developmental disorders – disruption – has been excluded from further considerations because its relevance to developmental hip disorders remains unproven. With the exception of deformity, which falls within the group the of so-called “packaging problems”, the three remaining pathomechanisms are collectively referred to as “production problems”12.\n\nMany recent scientific reports do not take into account the differences between the above-mentioned pathomechanisms. An equality sign is placed between them, treating them as synonyms usually grouped under one name, i.e. developmental dysplasia of the hip (DDH)13–15. Some publications use the following terms interchangeably: congenital hip dysplasia3, congenital hip dislocation16, developmental deformity of the hip17. Others contain statements such as “… malformation of anatomical structures occurs in dysplasia, which at the time of embryonic development were still normal ...” or “…developmental hip dysplasia is more deformity than malformation ...”12. According to the International Classification of Diseases and Health Problems (Q65 to Q79), the term “congenital hip dislocation” remains valid18.\n\nTo precisely explain the need to distinguish and organize these issues, we present and explain the following definitions, based on which a new division of hip developmental disorders is proposed.\n\nA congenital defect is a disorder present since birth. It is a general term, broadly describing the structural, behavioral, functional and metabolic damage that occurred in prenatal life, and which is diagnosed after birth or later in life19.\n\nMalformation (Latin: malformatio) is a term frequently used by English-speaking authors to refer to developmental disorders in general. More specifically, however, malformation represents just one of the four pathomechanisms of developmental disorders4. Malformation concerns developmental disorders, but only during the embryonic period. Referring to developmental changes which occur after this period as “malformation” is incorrect.\n\nMalformation of the mesenchymal primordium of the hip joint is a type of birth defect caused by a primary disorder of hip development during the embryonic period, during differentiation or organogenesis. The primary disorder affects cell proliferation, differentiation, migration, apoptosis or cell intercommunication processes. Primary impairment of cell function inhibits, delays or directs tissue development in the wrong direction, causing improper formation of anatomical structures of the hip4. Malformation underlies the development of congenital teratogenic hip dislocation.\n\nDysplasia of the hip is a type of disorder in which abnormally developing tissue (often excessively flaccid) results in faulty hip anatomy and evolves over time. Anatomical structures of the hip, normal during embryonic development, gradually become abnormal for various reasons10,20. Dysplasia may be environmentally or genetically conditioned. Dysplastic changes, along with malformation and disruption, may be collectively referred to as “production problems”. Hip dysplasia can occur both in the prenatal (early dysplasia) and postnatal (late dysplasia) period21. These disorders do not tend to self-heal12.\n\nDeformiy of the hip joint is an example of a developmental disorder in which properly developed structures are deformed during growth, as a result of mechanical factors. This can occur both in the pre- and postnatal periods. If the mechanical factor is active in the prenatal period, then we may refer to it as a “packaging problem”, associated with intrauterine fetal modeling. Disorders of this type are unlikely to cause growth disorders – rather, they tend to disappear with age and self-heal12.\n\nA developmental disorder is disorganization in the anatomical structure of the osteoarticular system that appears after some time, is absent or invisible at birth and has a tendency to either self-heal or worsen over time. In English literature, the term “structural defect” is often used in the context of developmental disorders to emphasize the anatomical nature of the defect.\n\nWhen referring to dysplasia and deformity, it is reasonable to introduce an additional term – “developmental disorder” – because we are talking about anomalies with a tendency to self-heal or become more severe over time4. Developmental dysplasia or developmental deformity of the hip, depending on the time of occurrence, can be early (primary – invisible and present at birth) or late (secondary – absent, and appearing after some time)12.\n\nAs disruption has not been described in the context of developmental hip joint disorders, and malformation is relatively easily diagnosed as a component of congenital anomalies, the main focus is on the distinction between dysplasia and deformity.\n\nThe aforementioned distinction is extremely important because it projects the course of treatment and prognosis of hip joint development disorders. It influences the choice of various conservative or surgical treatment strategies aimed at maintaining or restoring the normal growth potential of anatomical structures12. Misdiagnosis can lead to incorrect therapeutic management and, consequently, to deepening disability, thus significantly increasing the cost of treatment22,23.\n\nThe following is a discussion of the pathogenesis of developmental hip disorders with a focus on dysplasia and deformities.\n\n\nDiscussion\n\nThe pathomechanism of malformation cannot be the root cause of developmental hip disorder leading to dysplasia because it is only in the seventh week of life within the mesenchyme that the hip joint develops a fissure secreting the future femoral head and the acetabulum. Therefore, the first period when hip dislocation, and thus developmental hip dysplasia, may occur is the eleventh week of fetal life – the time when the hip joint is fully formed6. In the case of malformation, the most frequent causative factor is congenital anomalies syndrome, which generally does not pose major diagnostic difficulties. The effects of such congenital changes are present and visible immediately after childbirth24.\n\nIn the literature, dysplasia is considered in two aspects: dysplasia as a precancerous lesion (applies only to epithelial tissue, which is outside of the scope of this discussion), and dysplasia as a developmental disorder, involving incorrect organization or function of cells in a specific tissue (described as “production problems”), which is under consideration12. Dysplasia, which is a developmental disorder of the hip, can be grouped under the so-called osteoarticular dysplasia epiphyseal type20,25. It is characterized by abnormal growth potential of tissue structures underlying anatomical and functional changes in the growing hip8,9. In such cases, using the term dysplasia is fully justified.\n\nRisk factors for developmental dysplasia can include environmental or genetic conditions on both the mother and child side. Many scientific reports contain information on the impact of elevated level of biochemical factors on the occurrence of developmental dysplasia, e.g. female hormones (e.g. relaxin, estrogens) and biochemical markers of nutritional status (e.g. calcium, vitamins C and D)1. Regarding the relationship between the concentration of the hormone relaxin derived from the mother in the blood of the fetus and instability of hip joints, it was established that facts contradict the earlier assumption that hip instability coincides with increased relaxin concentrations in newborns. Instead, results indicate that hip instability frequently accompanies decreasing relaxin levels. The authors therefore assumed poorer mobilization of the pelvis and the birth canal during pregnancy as a result of the lower concentration of relaxin, which may result in greater pressure on the fetus in the perinatal phase26. Abnormalities caused by the disturbed balance of biochemical factors on the part of the mother can be expressed in the immaturity of the tissues of the child’s hip joint and the delay of their development in the prenatal period26–28. Such changes may be temporary and transient. With the right positioning of the hips, proper care of the newborn and then the baby, in most cases, the correct architecture of the anatomical elements of the hip can be restored7.\n\nOther studies report genetic disorders of the fetus underlying dysplastic changes in the hip joints. It has been confirmed that relaxation of ligaments and joint capsules, as well as irregularities in collagen metabolism, are associated with developmental dysplasia.\n\nIt has also been shown that some types of HLA A, B, and D, as well as mutations in specific genes or regulatory sequences, including genetic changes on chromosome 17 (17q21), predispose the child to developmental hip disorders of a dysplastic nature5,29,30. This group likely covers cases of developmental disorders characterized by prevalence of residual, recurrent and late forms that are resistant to treatment.\n\nIn situations (as assumed by the Delpesch law) where a correctly growing hip joint is affected by an external mechanical force, whether intracorporal (e.g. extra-articular contracture) or extracorporeal (e.g. incorrect position of the lower limb), leading to deformation of its anatomical structure, we are dealing with deformity12. Prolonged action of such factors, combined with ongoing growth, may result in ultimate subluxation or even full dislocation. In such cases, the term “deformity” is fully justified.\n\nIn deformity, change in the shape of the growing hip joint due to external extra-articular forces is not accompanied by disruption of the structure and tissue function in the initial phase, as is the case with dysplasia. Uneven distribution of forces acting on the roof of the growing acetabulum by the moving head leads to inhibition of growth of cartilage and bone tissue, their sclerosis and ultimately steep positioning of the acetabulum roof10. Atrophy of the acetabulum roof is accompanied by excessive bone growth within its fossa, i.e. in the unloaded zone. As a result of the loss of modeling and sliding out of the femoral head during acetabulum growth, the acetabulum bottom becomes bold, the acetabulum roof flattens and the acetabulum becomes shallow.\n\nThese processes of growth and remodeling of cartilage and bone tissue are of a secondary character and comply with Wolff-Delpesch laws, later developed by Hueter-Volkmann, Pauwels, and Arndt-Schmidt12.\n\nThe Hueter-Volkmann law also explains the presence of the most frequently observed pathological change in early postnatal hip dislocation, which is neolimbus (Ortolani positive symptom)6,12. The lack of physiological interaction (mutual pressure) between the head and the posterior edge of the acetabulum leads to excessive hypertrophy of the hyaline cartilage (neolimbus) in the upper, posterior and lower periphery of the acetabulum with the labrum curved out (pulled by a joint capsule in the dislocated hip)12.\n\nDeformity caused by mechanical factors, which is usually the result of intrauterine modeling, especially in the last trimester of pregnancy (hence the term “packaging problems”) usually does not cause disturbances in the growth potential of joint tissues, as observed in dysplasia or malformation. Instead, it exhibits a tendency to self-heal and rarely leads to relapse12.\n\nThis group includes cases of fetuses with abnormal breech position and ultra-position of limbs which self-heal or recover in the postnatal period, assisted by short-term conservative therapy7. Treatment involving restoration of the compact joint with concentric maintenance of the head in the acetabulum ensures optimal development conditions. If reposition is effectively maintained, the acetabulum, femoral head and femoral neck in anterversion undergo remodeling as a result of their normal growth potential. Restoration of the correct and stable joint connection between the femoral head and the acetabulum can lead to remodeling of the deformity and normalization of the morphology of the hip (due to developmental plasticity)12. The potential and growth time of the hip joint are closely related and depend on genetic and environmental factors12. These include genetic variations, e.g. of the SNP type, modulating the activity of proteins important from the point of view of tissue function, along with factors such as nutrition, general health, hormone concentration, mechanical forces and physiological age30. Therefore, during diagnostics and treatment, dysplastic and deformative cases should be considered together.\n\nTime can be an ally or an enemy, depending on whether growth potential remains normal. If it does, as in the case of deformity, then growth may promote development of correct anatomical structures (after correction and concentric arrangement of the elements of the hip joint). In the absence of normal growth potential, as with dysplasia, deformity of anatomical structures may deepen12 as growth progresses, even if a concentric position of the hip is achieved (recurrent, residual dysplasia resistant to treatment).\n\nIt is nearly impossible to distinguish between dysplasia and hip joint deformity on the basis of physical examination and imaging, e.g. ultrasound, X-ray and MRI. The procedures used in many countries, which call for imaging when Ortolani, Barlow and limited abduction tests are positive, may not be sufficient12,15. Because of the difficulty in correctly distinguishing between these two pathomechanisms, misdiagnosis often follows, and terminology is incorrectly applied. In the diagnosis and treatment of developmental hip joint disorders, there is a notable lack of uniform diagnostic and therapeutic standards in various countries around the world. This applies to the frequency of examinations and their complementarity15,31–34. There is also the danger of not detecting so-called clinically mute developmental disorders6,35. These are cases in which physical examination does not provide information about pathological changes at the joint level, which, however, become evident under imaging15. Literature describes cases of otherwise healthy children with normal physical examinations and radiographs of the hip in the first 3 months of life, who later developed hip dislocations21.\n\nThe opposite situation – involving diagnostically mute developmental disorders – may also occur. This condition occurs when physical examination clearly indicates a developmental disorder of the hip joint, and even its total displacement, while additional imaging tests do not confirm such abnormalities21. The joint can be anatomically normal but functionally abnormal, e.g. due to limitation of hip abduction. Passivity towards asymmetrical movement of the hip abduction can lead to a developmental hip disorder which often results in dislocation15. In this context, it should be noted that in cases of developmental hip disorder dysplasia and deformity, physical examination remains the priority, but it must be supplemented with imaging tests. Complementarity of physical examinations and imaging tests may reduce the number of undetected diagnostically and clinically silent cases.\n\nIn many cases of dysplasia in the first weeks of the child’s life, the doctor will not observe pathological anatomical changes of the hip and will not detect tissue changes despite the fact that they are present. These changes may lead to joint discongruity, deepening during hip development, as observed in late forms of dysplasia.\n\nIn the case of late dysplasia, diagnoses are most often made in a situation when there are degenerative changes in the hip joint presenting with pain in the 3rd or 4th decade of life6.\n\nTreatment implemented at this stage is symptomatic and clearly less effective than it might have been had the problem been noticed earlier. In order to avoid such situations, the possibility of implementing additional diagnostic tests should be considered. Given the development of high-throughput methods and the corresponding decrease in their cost, it is worth considering the implementation of a genetic testing procedure for detecting genetic variations responsible for pathological phenotypes. Knowledge about the genotype of patients with dysplastic changes would facilitate therapeutic planning in early cases of dysplasia, particularly with regard to abnormalities which are not phenotypically revealed until a later period in the patient’s life15. Such variations may explain the presence of short- and long-term tissue function disturbances, which lead to anatomical disorders of acetabulum development along with functional disorders of extra-articular tissues (contracture, excessive flaccidity)6,36,37.\n\nSeveral papers have been published on the topic of the genetic background of hip dysplasia; however, these studies always focus on changes occurring in specific genes. A study of the full range of genomic sequences with all potential variations (NGS sequencing studies) would allow us to describe the entire spectrum of variations comprising of the observed dysplastic changes29,30. Identification of e.g. single nucleotide changes (correlated with pathological phenotypes and affecting regulatory sequences which modulate protein functions) could significantly increase the level of diagnostic accuracy38,39.\n\nConsidering the above facts, it should be noted that nowadays there may still be clinically and diagnostically mute cases in neonatal hip tests, since genetic diagnostics are not yet routine. Integration of genetic testing with medical procedures would enable objective assessment of the situation at an early stage in both early and late dysplasia.\n\nDetection of differences at the level of genomic DNA, characteristic of the group of patients with dysplasia, would allow classification of this disorder with varying degrees of aggressiveness based on the obtained genetic profiles. In the future, this could serve as a tool for planning personalized therapy and become part of international standards15,31–34. Incorrect diagnosis resulting from incomplete patient data is the main cause of disability in childhood and adulthood, and treating such disability imposes a serious burden on state budgets40. With modern diagnostic tools the consequences of overlooking the defects could be largely avoided. In addition, this would sensitize the attending physicians to the possibility of late presentation of the abnormality, its resistance to treatment and eventual recurrence. Vigilance in the treatment process would facilitate thoughtful planning of the course of therapeutic management and thus improve the quality of life of patients in the future41.\n\nBased on the above considerations, a classification of developmental hip disorders was proposed depending on the etiology of the defect.\n\nI. Developmental disorders of the hip joint associated with “production problems”\n\nIn these abnormalities, the tissues forming the hip joint are primarily defective (dysplastic). Primary disturbed tissue development results in secondary disturbed hip joint anatomy.\n\n1) Teratogenic congenital hip dislocation\n\nThis is an example of a “tissue production” disorder which involves malformation at the embryonic stage, i.e. before the end of hip joint differentiation. It can be caused by genetic and/or biochemical and/or biophysical factors in the embryonic period.\n\n2) Developmental dysplasia of the hip (DDH)\n\nAnother “tissue production” disorder involving dysplasia, which may begin to manifest in the prenatal (fetal) period, i.e. from the time of hip joint formation (at 11–12 weeks of age) throughout the postnatal period. It may be caused by genetic and/or environmental factors.\n\nDepending on the time of onset, DDH can be divided into:\n\nA) Early (primary) developmental dysplasia of the hip\n\nThis disorder refers to the fetal phase of the prenatal period including the postnatal period up until the end of the 3rd month of life.\n\nB) Late (secondary) developmental hip dysplasia\n\nThis disorder refers to the postnatal period – after 3 months of age.\n\nII. Developmental disorders of the hip associated with so-called “packaging problems”\n\nIn these disorders, properly formed tissues of the anatomical structures of the hip joint are deformed due to prolonged presence of mechanical factors. An untreated deformity may, in the long term, cause secondary hip tissue changes in accordance with the remodeling and adaptation laws of Delpesch-Heuter-Volkmann – this mainly concerns bone tissue (atrophy and sclerosis in the overloaded zone, hypertrophy and low-density bone structure in the unloaded zone)42.\n\n1) Early (primary) developmental hip deformity\n\nThis disorder occurs in the prenatal (fetal) period and the postnatal period until the end of the 3rd month of life. Depending on the time of exposure to the deforming mechanical factor, we can distinguish a number of different risk factors. In the fetal period, these include: ultra position of the fetal lower limbs; breech position of the fetus; left hip joint – pressure on the sacrum prior to and during head delivery; oligohydramnios – intrauterine narrowness; primigravida.\n\nRisk factors present in the postnatal period (up until the end of the 3rd month of life) include incorrect diapering and extra-articular contractures.\n\n2) Late (secondary) developmental deformity of the hip\n\nThis disorder occurs in the postnatal period, after the 3rd month of life. Risk factors include improper care, neurogenic disorders (e.g. disorders of muscular balance in cerebral palsy, myelomeningocele, perinatal neuromuscular dystonia), inflammatory conditions (e.g. viral or bacterial hip inflammation), extra-articular contractures within e.g. adductor muscles of the hip caused by: idiopathic muscle fibrosis, ionizing radiation fibrosis, postinflammatory fibrosis (viral muscle damage, bacterial descent processes after abscesses) and fibrosing postpartum hematomas.\n\nIII. Developmental disorders of the hip associated with the so-called “packaging problems” and “production problems” (mixed)\n\nIn these disorders we deal with situations in which the deformity has a secondary effect on tissue quality, or deformity changes are superimposed upon dysplastic changes.\n\n1. Dysplastic disorder with secondary deformity changes\n\nAfter birth, the dysplastic hip joint with a disturbed congruence is affected by an additional iatrogenic mechanical external force, e.g. associated with incorrect diapers.\n\n2. Deformity or dysplastic disorder with secondary degenerative tissue changes\n\nProlonged maintenance of untreated deformity or dysplasia and the resulting lack of joint congruence may result in secondary degenerative changes. These changes are a direct consequence of improper nutrition of joint and extra-articular tissues caused by pathological intra- and extra-articular forces.\n\n\nConclusions\n\nDysplasia, malformation, and deformity are three of the four basic pathomechanisms leading to developmental hip disorders which are often not disambiguated at the clinical level and are therefore incorrectly named. Because malformation is relatively readable for the clinician, it is the least difficult to recognize. However, it is far more difficult to distinguish between dysplasia and deformity, because existing standards do not provide explicit methods along with a full range of diagnostic options. Nevertheless, this distinction is crucial and not merely a theoretical problem – it may influence medical practice, affecting patients’ quality of life and reducing treatment costs.\n\nTo achieve this goal, it is necessary to perform a full diagnostic process (physical examinations, imaging tests, genetic tests) in the immediate postnatal period to determine whether we’re dealing with:\n\n• an anatomically normal hip joint with a normal growth potential\n\n• teratogenic congenital dislocation of the hip with disturbed tissue growth potential in the malformation process (embryonic period)\n\n• a dysplastic hip joint with disturbed growth potential in the prenatal period (fetal period) (cases of early dysplasia)\n\n• an initially anatomically normal hip joint, but with a changed genotype that interferes with the potential for tissue growth in the postnatal period (cases of late dysplasia)\n\n• a deformed hip joint with normal growth potential occurring in the pre- and postnatal periods (cases of early and late developmental deformities)\n\n• a dysplastic hip joint with external deforming mechanical forces affecting it during the fetal or postnatal period\n\n• a deformed or dysplastic hip joint which is affected by internal or external mechanical forces (joint dyscongruence) resulting in secondary degenerative tissue changes.\n\nIn addition to routine procedures, specialists will likely have to rely on additional genetic testing to account for the possibility of recurrent residual dysplasia refractory to treatment despite a properly managed therapeutic process. In addition, such tests may help quickly explain the specific type of dysplasia and late deformities, preventing premature termination of diagnosis and treatment and thereby improving the quality of life of patients. The presented procedure would reduce the percentage of undetected diagnostically mute dysplasia and deformity.\n\n\nData availability\n\nNo data are associated with this article.",
"appendix": "References\n\nRhodes AML, Clarke NMP: A review of environmental factors implicated in human developmental dysplasia of the hip. J Child Orthop. 2014 [cited 2019 Dec 12]; 8(5): 375–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee CB, Mata-Fink A, Millis MB, et al.: Demographic Differences in Adolescent-diagnosed and Adult-diagnosed Acetabular Dysplasia Compared With Infantile Developmental Dysplasia of the Hip. J Pediatr Orthop. 2013 [cited 2019 Dec 12]; 33(2): 107–11. PubMed Abstract | Publisher Full Text\n\nKolb A, Windhager R, Chiari C: [Congenital hip dysplasia, screening and therapy]. Orthopade. 2015 [cited 2019 Sep 24]; 44(11): 917–26. PubMed Abstract | Publisher Full Text\n\nKorniszewski L: Dziecko z zespołem wad wrodzonych: diagnostyka dysmorfologiczna. Wydawn. Lekarskie PZWL; 2005 [cited 2019 Dec 12]; 259. Reference Source\n\nLoder RT, Skopelja EN: The Epidemiology and Demographics of Hip Dysplasia. ISRN Orthop. 2011 [cited 2019 Dec 12]; 2011: 238607. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWeinstein SL, Mubarak SJ, Wenger DR: Developmental hip dysplasia and dislocation: Part I. Instr Course Lect. 2004 [cited 2019 Dec 12]; 53: 523–30. PubMed Abstract\n\nWeinstein SL, Mubarak SJ, Wenger DR: Developmental hip dysplasia and dislocation: Part II. Instr Course Lect. 2004 [cited 2019 Dec 12]; 53: 531–42. PubMed Abstract | Publisher Full Text\n\nPonseti IV: Growth and development of the acetabulum in the normal child. Anatomical, histological, and roentgenographic studies. J Bone Joint Surg Am. 1978 [cited 2019 Dec 12]; 60(5): 575–85. PubMed Abstract | Publisher Full Text\n\nLindstrom JR, Ponseti IV, Wenger DR: Acetabular development after reduction in congenital dislocation of the hip. J Bone Joint Surg Am. 1979 [cited 2019 Dec 12]; 61(1): 112–8. PubMed Abstract\n\nHerring JA, Tachdjian MO, Texas Scottish Rite Hospital for Children: Tachdjian’s pediatric orthopaedics: from the Texas Scottish Rite Hospital for Children. 2014; 146.\n\nMorrissy RT, Weinstein SL, Kida B: Atlas of pediatric orthopaedic surgery. Lippincott Williams & Wilkins; 2001; 908. Reference Source\n\nDormans JP: Pediatric orthopaedics: core knowledge in orthopaedics. Mosby; 2005. Reference Source\n\nGuarniero R: DYSPLASIA OF HIP DEVELOPMENT: UPDATE. Rev Bras Ortop. 2015 [cited 2019 Sep 24]; 45(2): 116–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZamborsky R, Kokavec M, Harsanyi S, et al.: Developmental Dysplasia of Hip: Perspectives in Genetic Screening. Med Sci (Basel). 2019 [cited 2019 Jun 24]; 7(4): 59. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKotlarsky P, Haber R, Bialik V, et al.: Developmental dysplasia of the hip: What has changed in the last 20 years? World J Orthop. Baishideng Publishing Group Co; 2015; 6(11): 886–901. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFerreira N, Abreu M, Abreu A: Congenital Hip Dislocation: A Rare Case in Adulthood. Acta Med Port. 2018 [cited 2019 Sep 24]; 31(1): 68. PubMed Abstract | Publisher Full Text\n\nSiffert RS: Patterns of deformity of the developing hip. Clin Orthop Relat Res. 1981; 160: 14–29. PubMed Abstract\n\nWHO | International Classification of Diseases, 11th Revision (ICD-11). WHO. 2019. Reference Source\n\nCongenital anomalies. [cited 2019 Dec 16]. Reference Source\n\nLovell WW, Winter RB, Weinstein SL, et al.: Lovell and Winter’s pediatric orthopaedics. Wolters Kluwer Health/Lippincott Williams & Wilkins; 2014; 43. Reference Source\n\nRaimann A, Baar A, Raimann R, et al.: Late developmental dislocation of the hip after initial normal evaluation: a report of five cases. J Pediatr Orthop. 2007 [cited 2019 Sep 28]; 27(1): 32–6. PubMed Abstract | Publisher Full Text\n\nSewell MD, Eastwood DM: Screening and treatment in developmental dysplasia of the hip-where do we go from here? Int Orthop. 2011; 35(9): 1359–67. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAngliss R, Fujii G, Pickvance E, et al.: Surgical treatment of late developmental displacement of the hip. Results after 33 years. J Bone Joint Surg Br. 2005 [cited 2019 Dec 17]; 87(3): 384–94. PubMed Abstract | Publisher Full Text\n\nDietz FR, Morcuende JA: Embryology and development of musculoskeletal system. In: Lovell and Winter’s Pediartic Orthopaedics. 5th ed. Philadelphia: Lippincott Williams & Wilkins; 2001; 1–31.\n\nDysplazje kostno-stawowe | Reumatologia - Medycyna Praktyczna dla pacjentów. [cited 2019 Dec 16]. Reference Source\n\nKural B, Karapınar ED, Yılmazbaş P, et al.: Risk Factor Assessment and a Ten-Year Experience of DDH Screening in a Well-Child Population. Biomed Res Int. 2019 [cited 2019 Dec 17]; 2019: 7213681. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKasström H, Suzuki K, Olsson SE, et al.: Growth and remodeling of the hip joint and proximal femur in adolescent dogs. A scintimetric investigation with special reference to hip dysplasia and estradiol induced changes. Acta Radiol Suppl. 1975 [cited 2019 Dec 17]; 344: 75–80. PubMed Abstract | Publisher Full Text\n\nBeling CG, Gustafsson PO, Kasström H: Metabolism of estradiol in greyhounds and German shepherd dogs. An investigation with special reference to hip dysplasia. Acta Radiol Suppl. 1975 [cited 2019 Dec 17]; 344: 109–20. PubMed Abstract | Publisher Full Text\n\nUnger S: A genetic approach to the diagnosis of skeletal dysplasia. In: Clinical Orthopaedics and Related Research. Lippincott Williams and Wilkins; 2002; (401): 32–8. PubMed Abstract | Publisher Full Text\n\nKenanidis E, Gkekas NK, Karasmani A, et al.: Genetic Predisposition to Developmental Dysplasia of the Hip. J Arthroplasty. 2020; 35(1): 291–300.e1. Publisher Full Text\n\nSchaeffer EK, Ihdi Study Group, Mulpuri K: Developmental dysplasia of the hip: addressing evidence gaps with a multicentre prospective international study. Med J Aust. 2018 [cited 2019 Jun 24]; 208(8): 359–64. PubMed Abstract | Publisher Full Text\n\nPaton RW: Screening in Developmental Dysplasia of the Hip (DDH). Surgeon. 2017 [cited 2019 Jun 24]; 15(5): 290–6. PubMed Abstract | Publisher Full Text\n\nShaw BA, Segal LS, SECTION ON ORTHOPAEDICS: Evaluation and Referral for Developmental Dysplasia of the Hip in Infants. Pediatrics. 2016 [cited 2019 Jun 24]; 138(6): e20163107. PubMed Abstract | Publisher Full Text\n\nDormans JP: Pediatric orthopaedics: core knowledge in orthopaedics. Mosby; 2005; 515. Reference Source\n\nNamdari S, Pill SG, Mehta S: Orthopedic Secrets: Fourth Edition. Elsevier Inc.; 2014; 1–464. Reference Source\n\nRubini M, Cavallaro A, Calzolari E, et al.: Exclusion of COL2A1 and VDR as developmental dysplasia of the hip genes. Clin Orthop Relat Res. 2008 [cited 2019 Sep 18]; 466(4): 878–83. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang K, Shi D, Zhu P, et al.: Association of a single nucleotide polymorphism in Tbx4 with developmental dysplasia of the hip: A case-control study. Osteoarthritis Cartilage. 2010; 18(12): 1592–5. PubMed Abstract | Publisher Full Text\n\nBasit S, Hannan MA, Khoshhal KI: Developmental dysplasia of the hip: usefulness of next generation genomic tools for characterizing the underlying genes – a mini review. Clin Genet. Blackwell Publishing Ltd; 2016; 90(1): 16–20. PubMed Abstract | Publisher Full Text\n\nShi D, Dai J, Ikegawa S, et al.: Genetic study on developmental dysplasia of the hip. Eur J Clin Invest. 2012; 42(10): 1121–5. PubMed Abstract | Publisher Full Text\n\nSewell MD, Eastwood DM: Screening and treatment in developmental dysplasia of the hip-where do we go from here? Int Orthop. 2011; 35(9): 1359–67. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAngliss R, Fujii G, Pickvance E, et al.: Surgical treatment of late developmental displacement of the hip. Results after 33 years. J Bone Joint Surg Br. 2005 [cited 2019 Sep 29]; 87(3): 384–94. PubMed Abstract | Publisher Full Text\n\nBłoński M, Nowakowski A (ortopedia)., Mazurek TA, et al.: Ortopedia i traumatologia - podręcznik dla studentów Podręcznik rekomendowany przez Polskie Towarzystwo Ortopedyczne i Traumatologiczne. Wydaw. Naukowe Exemplum; 2017 [cited 2019 Dec 12]. Reference Source"
}
|
[
{
"id": "72952",
"date": "17 Nov 2020",
"name": "Grzegorz Szczęsny",
"expertise": [
"Reviewer Expertise Orthopaedic surgery",
"musculoskeletal trauma",
"bone physiology",
"fracture healing",
"malunion and non-union",
"skeletal infections",
"skeletal pathology and bone immunology"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn the paper, the Authors discuss a quite controversial idea that developmental dysplasia of the hip is not a pure dysplasia, but partly belongs to a group of malformations, partly to deformities and partly to developmental disorders. Thus, they proposed their own classification of the DDH basing their idea.\nUnfortunately, according to DDH, its classification of malformations, deformities or developmental disorders is very controversial, as DDH in fact does not fulfill the definitions of those conditions.\nAccording to definitions given by several widely accepted sources, including https://medical-dictionary.thefreedictionary.com/malformation, https://onlinelibrary.wiley.com/doi/pdf/10.1002/ajmg.a.36249 and Wikipedia, malformation is associated with a disorder of tissue development that often occurs in the first trimester. Deformation is equal to malformation (https://medical-dictionary.thefreedictionary.com/malformation).\nAuthors assumed that pathological changes that are observed in the hip joint affected by something that is nowadays called DDH could originate from malformation, deformity and developmental disorders. In the paragraph: “… Many recent scientific reports do not take into account the differences between the above-mentioned pathomechanism. An equality sign is placed between them, treating them as synonyms usually grouped under one name, i.e. developmental dysplasia of the hip (DDH)13–15. Some publications use the following terms interchangeably: congenital hip dysplasia3, congenital hip dislocation16, developmental deformity of the hip17. Others contain statements such as “… malformation of anatomical structures occurs in dysplasia, which at the time of embryonic development were still normal ...” or “…developmental hip dysplasia is more deformity than malformation ...”12. According to the International Classification of Diseases and Health Problems (Q65 to Q79), the term “congenital hip dislocation” remains valid18…” they use several footnotes that do not confirm their theses. Moreover, literature positions 13, 14 and 15 do not discuss malformation, deformation, deformity nor developmental disorders et al. In the position nr 14, the words deformity and developmental disorder were used one time only each and in a totally different meaning than that presented by Authors of the analyzed publication. The statement: “…Malformation of the mesenchymal primordium of the hip joint is a type of birth defect caused by a primary disorder of hip development during the embryonic period, during differentiation or organogenesis…” do not correspond to its definition, since malformation is associated with a disorder of tissue development. Disorders at the organ level are called dysplasia (https://en.wikipedia.org/wiki/Birth_defect).\nOther inaccuracies: “…The pathomechanism of malformation cannot be the root cause of developmental hip disorder leading to dysplasia because it is only in the seventh week of life within the mesenchyme that the hip joint develops a fissure secreting the future femoral head and the acetabulum. Therefore, the first period when hip dislocation, and thus developmental hip dysplasia, may occur is the eleventh week of fetal life – the time when the hip joint is fully formed6. In the case of malformation, the most frequent causative factor is congenital anomalies syndrome, which generally does not pose major diagnostic difficulties. The effects of such congenital changes are present and visible immediately after childbirth24…”. Methinks that it is not especially the fissure of the hip joint that secretes femoral head and acetabulum. Both structures are rather formed in consequence of endochondral ossification of cartilaginous structures of the three pelvic bones at the ypsilon cartilage and femoral head. “…dysplasia as a precancerous lesion…” could hardly be proven. Authors should focus on the subject of their paper - that is on hip dysplasia. The paragraph “…Early detection of cases of late dysplasia. In many cases of dysplasia in the first weeks of life, the doctor will not be able to observe clinically pathological signs and thus will not be able to “detect … dysplasia” despite the fact that they are present. These changes may lead to joint discongruity, deepening during hip development, as observed in late forms of dysplasia. In the case of late dysplasia, diagnoses are most often made in a situation when there are degenerative changes in the hip joint presenting with pain in the 3rd or 4th decade of life6…” seems to focus on “late” dysplasia. Nevertheless, that’s rather a degenerative joint disease that makes the problem, not hip dysplasia itself at the third and fourth decade of life. Authors should define the term “early detection … of late dysplasia”. Do they mean the diagnosis? If so - why an “early” diagnosis is important in such “antiquated” dysplasia? Proposed classification should be discussed in a group of investigators and practitioners, including geneticists, neonatologists, pediatricians, orthopedists, etc. Proposed treatment, albeit scanty, as well. In classification generally accepted definitions and terms should be abided by. Paper requires linguistic improvements and English language edit.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Partly",
"responses": [
{
"c_id": "6149",
"date": "04 Dec 2020",
"name": "Monika Piwowar",
"role": "Author Response",
"response": "We would like to thank the reviewer for his time and the opportunity to develop the topic and explain the issues that turned out to be ambiguous and unclear on the reception of our study. The posted comments made us aware which threads in our paper should be deeply clarified. We will publish an updated version of the manuscript soon. In our manuscript, we tried to organize many years of knowledge, documented in professional literature, which in some areas is misleading due to juggling by the terminology in some medical issues. For this reason, in the title of the manuscript, we have included the names of the main pathomechanisms of hip defects, i.e. dysplasia, deformity, malformation, which are often used as synonyms but are not. Based on the publication references, we have demonstrated the line of our reasoning and the proposed division of hip developmental defects derived from it. It is difficult for us to argue with the reviewer because the reviewer's comments are incorrect in the assumptions themselves, which we will try to justify below. Moreover, the conceptual set we use is not consistently used in two-sided argumentation. We understand, however, where it comes from, namely from the great mess in the nomenclature in which we also found ourselves at some point, and which prompted us to organize these issues. A discussion of DDH is only possible if the debaters are based on a consensus understanding of the same concepts. Therefore, in the manuscript, we have included a number of definitions (from professional literature) so that every attentive reader has no doubts about what we are writing about and what we are referring to, and in order to precisely and unambiguously describe the issue. We would like to emphasize the differences between dysplasia and deformation (Latin deformatio) in the medical sense, which perhaps in the original text of the manuscript is not emphasized enough to be properly perceived by the reviewer. In order to explain our view more clearly, we must return to the explanation of the terms we use. The interchangeable use of similar meaning words such as distortion and deformity, distortion and malformation, deformation (or deformity) and dysplasia is used in everyday language. On the level of medical considerations, however, it is not acceptable. And this is not just our opinion. Leading publication in the field of orthopedics J. P. Dormans \"Pediatric Orthopedics: Core Knowledge in Orthopedics\" in Chapter I by Steven L. Frick entitled \"Concepts of proper human growth and development in pediatric orthopedics\" explains the differences in the concepts of dysplasia, deformation and distortion, on which we have based our considerations. The term dysplasia refers to tissue whose structure is defective or poorly constructed; disturbed development causes disturbed anatomy. It is one of the four defect pathomechanisms. Following [Tachdjian's Pediatric Orthopedics [Chapter 01]], dysplasia is similarly defined, ie: „Dysplasias are structural defects caused be abnormal tissue differentiation as cell organize into tissues”. The term deformity (or deformation) refers to a properly developed anatomical structure that has been deformed by mechanical factors. This is another of the four defect pathomechanisms. In the publication [Tachdjian's Pediatric Orthopedics [Chapter 01]] one can find another confirmation of the above definition: „Deformations are defects in the form, shape, or site of body parts caused by mechanical stress. The mechanical stress, which may be intrinsic or extrinsic, alters or distorts tissue. Because the fetus grows considerably faster than the infant, fetus are more vulnerable to deformations”. Distortion is a pathological condition in which the shape of a part of the body has remodeled beyond the normal range. Distortion is therefore an overarching term that described both deformation and malformation, as well as dysplasia. In connection with the above, deformation (deformity) in medical terms can by no means be dysplasia, let alone malformation (Latin malformatio), but all of the above can lead to distortions of parts of the human body. The distinguishing of these terms is crucial for the correct diagnosis and subsequent therapeutic management, which is fraught with consequences. The effects of conservative or surgical treatment will be different in the case of Developmental Deformity of the Hip Joint and in the case of Developmental Dysplasia of the hip joint. The reviewer used the term deformity to refer to malformation as synonyms. In our study, we pay attention to the fact that such processes are commonly erroneous interchangeably used. Suffice it to mention that by the term malformation some scientists describe various issues, e.g. -> malformation as a pathomechanism of developmental defects, which is acting in the embryonic period -> malformation as a synonym of distortion and often for this reason incorrectly called deformity (deformation), which is another pathomechanism of developmental defects completely different from malformation -> malformation as a developmental defect causing confusion in the interpretation of medical scientific reports. In our publication, we understand malformation as one of the four pathomechanisms of defects alongside dysplasia, deformity and disruption as defined contained in [Tachdjian's Pediatric Orthopedics [Chapter 01]]: „Malformation are structural defects that result from interruption of normal organogenesis during the second month of gestation”. “ It is important to differentiate deformations form malformations. (…) Malformations cannot be corrected directly,, whereas deformations can often be revised relatively easily either by elimination the deforming force or by counteracting the force with stretching, casting, or bracing”. This particular pathomechanism only works in the embryonic period and underlies the teratogenic hip dislocation. The aim of the publication was to draw attention to the extent to which DDH is based on the pathomechanism of dysplasia, and to what extent on the pathomechanism of deformation, and to draw attention to the need to refine diagnostic schemes to distinguish these disease entities, which is currently not fully implemented despite the technical possibilities. Currently, as we claim, not every officially diagnosed developmental dysplasia is in fact a disease whose pathomechanism is pure dysplasia. Some of them are most likely a developmental deformity. The evidence showing the presence of diagnostically mute and clinically mute cases supports the above thesis. For this reason, we have proposed an orderly division of DDH along with the nomenclature. We are not claiming that the division presented in our manuscript is perfect. It certainly needs discussion and perhaps modification, but we hope it will contribute to serious consideration on this topic as it seems necessary. We hope that the description of the topics raised by the reviewer will help readers understand the essence of the matter and the need to organize the issue of DDH, as well as start discussing the extension of diagnostic schemes."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1231
|
https://f1000research.com/articles/9-683/v1
|
07 Jul 20
|
{
"type": "Method Article",
"title": "Maternal and perinatal Health Research Collaboration, India (MaatHRI): methodology for establishing a hospital-based research platform in a low and middle income country setting",
"authors": [
"Manisha Nair",
"Babul Bezbaruah",
"Amrit Krishna Bora",
"Krishnaram Bora",
"Shakuntala Chhabra",
"Saswati S. Choudhury",
"Arup Choudhury",
"Dipika Deka",
"Gitanjali Deka",
"Vijay Anand Ismavel",
"Swapna D. Kakoty",
"Roshine M. Koshy",
"Pramod Kumar",
"Pranabika Mahanta",
"Robin Medhi",
"Pranoy Nath",
"Anjali Rani",
"Indrani Roy",
"Usha Sarma",
"Carolin Solomi V",
"Ratna Kanta Talukdar",
"Farzana Zahir",
"Michael Hill",
"Nimmi Kansal",
"Reena Nakra",
"Colin Baigent",
"Marian Knight",
"Jenny J. Kurinczuk",
"Babul Bezbaruah",
"Amrit Krishna Bora",
"Krishnaram Bora",
"Shakuntala Chhabra",
"Saswati S. Choudhury",
"Arup Choudhury",
"Dipika Deka",
"Gitanjali Deka",
"Vijay Anand Ismavel",
"Swapna D. Kakoty",
"Roshine M. Koshy",
"Pramod Kumar",
"Pranabika Mahanta",
"Robin Medhi",
"Pranoy Nath",
"Anjali Rani",
"Indrani Roy",
"Usha Sarma",
"Carolin Solomi V",
"Ratna Kanta Talukdar",
"Farzana Zahir",
"Michael Hill",
"Nimmi Kansal",
"Reena Nakra",
"Colin Baigent",
"Marian Knight",
"Jenny J. Kurinczuk"
],
"abstract": "Background: Maternal and perinatal Health Research collaboration, India (MaatHRI) is a research platform that aims to improve evidence-based pregnancy care and outcomes for mothers and babies in India, a country with the second highest burden of maternal and perinatal deaths. The objective of this paper is to describe the methods used to establish and standardise the platform and the results of the process. Methods: MaatHRI is a hospital-based collaborative research platform. It is adapted from the UK Obstetric Surveillance System (UKOSS) and built on a pilot model (IndOSS-Assam), which has been extensively standardised using the following methods: (i) establishing a network of hospitals; (ii) setting up a secure system for data collection, storage and transfer; (iii) developing a standardised laboratory infrastructure; and (iv) developing and implementing regulatory systems. Results: MaatHRI was established in September 2018. Fourteen hospitals participate across four states in India – Assam, Meghalaya, Uttar Pradesh and Maharashtra. The research team includes 20 nurses, a project manager, 16 obstetricians, two pathologists, a public health specialist, a general physician and a paediatrician. MaatHRI has advanced standardisation of data and laboratory parameters, real-time monitoring of data and participant safety, and secure transfer of data. Four observational epidemiological studies are presently being undertaken through the platform. MaatHRI has enabled bi-directional capacity building. It is overseen by a steering committee and a data safety and monitoring board, a process that is not normally used, but was found to be highly effective in ensuring data safety and equitable partnerships in the context of low and middle income countries (LMICs). Conclusion: MaatHRI is the first prototype of UKOSS and other similar platforms in a LMIC setting. The model is built on existing methods but applies new standardisation processes to develop a collaborative research platform that can be replicated in other LMICs.",
"keywords": [
"Research platform",
"research model",
"epidemiology",
"low-and-middle income country",
"India",
"maternal health",
"perinatal health"
],
"content": "Introduction\n\nMaternal health is a global priority due to the large number of women becoming pregnant every year, an estimated 211 million1, and because of the growing disparity in maternal deaths across countries2,3. India has the second highest number of maternal deaths with ~45,000 deaths yearly2. The rate is much higher for some states, such as Assam in the Northeast of India. Assam has nearly half the population of the UK, and 6 women die every day as a result of pregnancy and childbirth complications4 compared with around one per week in the UK5. In addition, each year an estimated 5 million pregnant women in India experience a life-threatening complication. To improve care and outcomes, India needs large and robust studies to investigate the risk factors, management and outcomes of pregnancy complications and to find out why disease severity varies from state to state.\n\nIn a pilot project (called IndOSS-Assam) we demonstrated the feasibility of setting up a collaborative platform for maternal and perinatal health research jointly undertaken by Indian clinical collaborators and researchers at the University of Oxford6,7. This was modelled on the UK Obstetric Surveillance System (UKOSS)8 and showed that a hospital-based platform can be used to conduct large epidemiological studies and routine surveys to investigate pregnancy complications and management, and establish incidence and outcomes. UKOSS through its work over the past decade has contributed significantly to improving the safety and quality of care for pregnant women8. It has inspired several high-income countries to establish obstetric surveillance and research systems, which are being used to conduct national and multi-national studies to generate evidence to improve pregnancy care. However, there is no such system in low-and-middle income countries (LMICs) where more than 94% of all maternal deaths occur.\n\nOur pilot work in India not only justified the importance and urgency, but also demonstrated the need to further adapt and improve the pilot model to create a standardised collaborative platform for both research and research capacity building. This led to the establishment of the Maternal and perinatal Health Research collaboration, India (MaatHRI), a larger standardised collaborative research platform of 14 public and private hospitals across four states in India. The objective of this paper is to describe the methods used to establish and standardise the platform and the results of the process. MaatHRI means mother in Sanskrit.\n\n\nMethods\n\nMaatHRI is a hospital-based collaborative research platform established to: (i) regularly collect data on the prevalence of known and emerging life-threatening pregnancy complications; (ii) conduct large epidemiological studies to generate evidence to improve maternal and perinatal health in India; and (iii) develop research capacity and skills in the collaborating hospitals. It was built on the pilot system, but extensively expanded and standardised over a period of 18 months from May 2017 to September 2018. The following methods were used to establish the collaborative platform:\n\n1. Establishing a network of hospitals and clinical collaborators\n\n2. Setting up a high-quality secure system for data collection, storage, and transfer\n\n3. Developing a standardised laboratory infrastructure\n\n4. Implementing regulatory systems\n\nSuccessful completion of the pilot work in two hospitals in Assam allowed us to expand the network from two to nine government hospitals within Assam: six teaching hospitals and four district hospitals. In each hospital, we identified a lead collaborator who were obstetricians. Through their professional networks, we were able to reach out to other hospitals. A hospital was included in the network based on two criteria: (i) willingness of the hospital to participate in a large research collaboration and (ii) a high burden of maternal and perinatal deaths in the population covered by the hospital. Similar to the process used in the pilot work6, we mapped the hospitals to assess the spread and coverage of the population in each state.\n\nOne of the major reasons for success of the pilot work was having dedicated research staff for data collection and data entry. A pragmatic approach was adopted to develop a high quality secure electronic system to overcome the challenges of human resource constraints, lack of dedicated secure computer servers for data storage in the hospitals, and securely sharing data. The following methods were employed:\n\ni. Research nurses were appointed in each hospital and trained\n\nii. Electronic online data collection forms were developed for entering data\n\niii. Data are collated automatically in a cloud-based server located in India\n\niv. Quality assurance and data security procedures were established and implemented\n\nA laboratory infrastructure was created through a partnership with a private laboratory in India, Dr Lal PathLabs (LPL). LPL has a pan-India presence with a network of sample collection centres, regional laboratories and a national reference laboratory in New Delhi, India. Their existing service delivery model was adapted to the requirements of the MaatHRI platform through extensive consultations between the Indian clinical collaborators, and experts at the University of Oxford and LPL. The following services were agreed and are being currently used to standardise the laboratory infrastructure:\n\n- Service 1: Provide blood collection kits with instructions to all study hospitals\n\n- Service 2: Train MaatHRI research nurses to collect and prepare blood samples\n\n- Service 3: Transport samples at ambient conditions from the hospitals to the laboratory\n\n- Service 4: Standardise blood assays\n\n- Service 5: Produce standardised test reports\n\nWe tested the model in a run-in phase before full implementation.\n\nThe steering committee constituted for the pilot work (IndOSS-Assam) was expanded to form the MaatHRI steering committee. The committee includes representatives from all the collaborating hospitals, the University of Oxford, Indian policy advocates and experts in statistics and ethics. As MaatHRI is a research platform set up to conduct studies on a long-term basis, an independent ‘Data safety and monitoring board’ (DSMB) was set up, including members from India and the UK who are not associated with the MaatHRI platform. A DSMB charter was drafted outlining the roles and responsibilities of the members and how the board will function to provide independent safety review of participants and data, and guidance for observational studies during the course of the ongoing projects. Since the studies currently undertaken through the platform are observational studies, review of adverse event data and reports of serious adverse events (SAEs) are not currently applicable to MaatHRI DSMB. However, should randomised controlled trials be conducted through MaatHRI in the future we would expect the DSMB to be involved in reviewing this type of information.\n\nThe MaatHRI platform and the ongoing studies have been approved by the institutional review boards (IRB) of each coordinating Indian institution, namely: Srimanta Sankaradeva University of Health Sciences, Guwahati, Assam (No.MC/190/2007/Pt-1/126); Nazareth hospital, Shillong, Meghalaya (Ref No. NH/CMO/IEC/COMMUNICATIONS/18-01); Emmanuel Hospital Association, New Delhi (Ref. Protocol No.167); Mahatma Gandhi Institute of Medical Sciences, Sevagram, Maharashtra (Ref No. MGIMS/IEC/OBGY/118/2017); and the Institute of Medical Sciences, Banaras Hindu University, Varanasi, Uttar Pradesh (No.Dean/2018/EC/290). The project has also been approved by the Government of India’s Health Ministry’s Screening Committee, the Indian Council of Medical Research, New Delhi (ID number 2018-0152) and by the Oxford Tropical Research Ethics Committee (OxTREC), University of Oxford, UK (OxTREC Ref: 7-18).\n\n\nResults\n\nWe were able to establish a network of 14 hospitals by September 2018 across four states in India – Assam, Meghalaya, Uttar Pradesh, and Maharashtra. After establishing the network, two more hospitals joined MaatHRI, but two government district hospitals left the collaboration. A lack of interest in research and high patient load were the main reasons given by the lead collaborators of the departing hospitals. The MaatHRI platform currently includes a network of 14 hospitals (11 Government and 3 private). Figure 1 shows the distribution of the network across India and within the state of Assam. The 14 hospitals together conduct about 100,000 deliveries per year. The network includes an Indian research team of 20 nurses, a project manager, 16 obstetricians, two pathologists, a public health specialist, a general physician and a paediatrician.\n\nData and biological sample collection: Depending on the patient load and related participant recruitment rates, one or two research nurses have been appointed in each collaborating hospital for the MaatHRI work. The nurses are responsible for recruiting study participants, providing participant information and obtaining informed consent, collecting data and blood samples, and following up participants. The research nurses were specifically trained to undertake these activities. In addition, a project manager has been appointed to manage the research nurses and provide supportive supervision.\n\nData entry and storage: Our original plan was for research nurses to collect data in online electronic forms using tablet PCs enabling automatic collation in the Microsoft Azure cloud computing platform (Microsoft Corporation) with servers located in India; there is no provision for storing data on the tablets. However, after an initial trial we found that direct data entry in an online form was not possible due to problems with internet connections in several hospitals and the sensitivity associated with a nurse standing with a tablet PC next to a very sick woman. It was therefore decided that paper forms would be used to collect data in these hospitals and the nurse would enter the data immediately afterwards into the online data portal and then destroy the paper forms. Each hospital has a unique Login ID and password to access the data collection forms and their collated data on the online portal.\n\nQuality assurance and data security: The electronic data collection forms have checks and validations to flag logical errors. The project manager is responsible for monitoring data entry on a day-to-day basis. Red flags are raised for errors and incomplete forms immediately so that the research nurse can rectify the errors before the participant is discharged from the hospital. Data stored in the cloud server are encrypted and password protected. Each collaborating hospital can only view and download its own data. Identifiable information are collected for follow-up of participants, but these can only be viewed by the authorised hospital staff and cannot be downloaded by anyone. Once the data collection is complete, in preparation for analysis, all identifiable information is completely delinked from the clinical data to generate pseudonymised analysis files. We have developed secure mechanisms for transferring data within India and between India and the UK with recommended level of end-end-encryption.\n\nDr Lal Pathlabs (LPL) provides the laboratory infrastructure for MaatHRI. The following services were tested in a trial run before being fully incorporated into the platform.\n\nService 1: Blood collection kits with instructions to all study hospitals. LPL provides the required blood collection kits with specific written guidance to all study hospitals for collecting, processing and packing the blood samples.\n\nService 2: Train MaatHRI research nurses to collect and prepare blood samples. Technical experts from the laboratory trained the MaatHRI staff (project manager and research nurses) to collect, centrifuge and pack samples before the start of studies. When required, a phlebotomist from their collection centre provided supportive supervision to the research nurses during the initial few weeks to correct or prevent any errors.\n\nThe MaatHRI research nurses collect, centrifuge and pack blood samples as per instructions in transportation boxes ready for collection by LPL. A standard test requisition form for each participant is filled in by the obstetrician caring for the participant. This form only includes the participant ID, age and a barcode to ensure participant confidentiality and blinding to minimise reporting bias. The test results are only used for research purposes and not for the provision of clinical care.\n\nService 3: Transport samples at ambient conditions from the hospitals to the laboratory. A designated person from the LPL collection centre collects the boxes from the hospital. These are transported via road to the nearest regional laboratory where they are checked and then shipped via air to the national laboratory in New Delhi. A flow-chart describing the transportation process from the hospitals to the LPL National Reference Laboratory is shown in Figure 2 and the network is presented in a map in Figure 3. Time in transit is regularly monitored by LPL and reported for each participant along with their test results.\n\nService 4: Standardising blood assays. All samples are processed and analysed in the LPL National Reference Laboratory based at New Delhi. The assay methods, traceability and performance characteristics are discussed by experts from the University of Oxford’s Wolfson laboratory and LPL before including a test in the study. Table 1 shows the traceability and Table 2 shows the performance characteristics for assays that are commonly used for the epidemiological studies undertaken using the MaatHRI platform. The details of specific tests will be presented in subsequent publications. Traceability and assay performance monitoring are important for standardisation of laboratory procedures and quality control. If the quality of a blood sample is compromised in transit, it is not processed, and the site-collaborator and research nurse are advised to collect a fresh sample. The laboratory runs quality control checks daily for each assay (twice a day for some) and monitors their mean coefficient of variation and standard deviation. The results are shared as part of a performance monitoring plan during monitoring and feedback meetings. In addition, LPL also runs a quarterly Quality Improvement Programme.\n\nNGSP - National Glycohemoglobin Standardization Program; CLSI – Clinical and Laboratory Standards Institute; HbF – Fetal haemoglobin; HbA2 - Haemoglobin Subunit Alpha 2; NA - Not applicable\n\nCAP - College of American Pathologists; CAP PT - College of American Pathologists Proficiency Testing programme; HbF – Fetal haemoglobin; HbA2 - Haemoglobin Subunit Alpha\n\nThe LPL National Reference Laboratory is accredited by the following bodies – College of American Pathologists (CAP); National Accreditation Board for Testing and Calibration (NABL); British Standards Institution (Quality Management System ISO 9001: 2015, FS 60411).\n\nService – 5: Test reports. Test reports are securely made available to the site-collaborator in each hospital through their usual communication channel. Data from the reports are entered in the electronic forms by the research nurse.\n\nMaatHRI steering committee has met biannually since the platform was established in September 2018. The role of the steering committee is to guide the platform in terms of vision, scope, equitable partnership, and research and training priorities. It is also responsible for communicating the results of the studies undertaken through MaatHRI to the Ministry of Health and Family Welfare (MoHFW), Government of India.\n\nThe DSMB periodically reviews participant recruitment, data safety and confidentiality, ethical issues and data quality, and examines whether the overall safety and feasibility of the MaatHRI project is acceptable. Although conventionally DSMB is set up for individual studies, we found that setting up a DSMB for the research platform that has oversight of all studies undertaken through the platform could be an effective way to ensure data safety. The DSMB has met twice since MaatHRI was established in September 2018 and membership includes two obstetricians (one from the UK and one from India), one paediatrician (from India), one biostatistician (from the UK), and one expert in bioethics (from India), all with prior experience and expertise in observational epidemiological studies. They were nominated by the study investigators.\n\nOne survey and three observation studies are currently being undertaken through the platform. A monthly survey of nine life-threatening complications of pregnancy has been in progress since July 2018. The complications are defined using standard definitions and include eclampsia, pre-eclampsia, postpartum haemorrhage, maternal peripartum infection, septic abortion, uterine rupture, heart failure during pregnancy and postpartum, transient peripheral neuropathy, and Japanese encephalitis complications.\n\nThe epidemiological studies undertaken are informed by the knowledge and hypothesis generated during the pilot work for IndOSS-Assam. They include: (i) an unmatched case-control study examining the risk factors, clinical characteristics, and outcomes of heart failure in pregnant and postpartum women; (ii) a prospective cohort study investigating the safety of induction and augmentation of labour in pregnant women with anaemia; and (iii) a nested study within the prospective study comparing the coagulation parameters in pregnant women with and without anaemia. Of these, the nested coagulation study is complete, and the other two studies will be completed by June 2022. The monthly survey will continue as long as the collaborative platform exists.\n\n\nDiscussion\n\nMaatHRI, a collaborative research platform, modelled on UKOSS, was successfully established to conduct hospital-based research to improve care and outcomes for mothers and babies in India. It includes 14 public and private hospitals across four states in India, which together conduct about 100,000 deliveries per year. The platform is standardised in terms of data collection, equipment, and laboratory methodology, and employs strict measures for participant confidentiality and data security. It is monitored by two regulatory bodies: a steering committee and an independent DSMB. One survey and three epidemiological studies are being undertaken through the platform.\n\nMaatHRI is the first prototype of UKOSS and other similar platforms9 in a low and middle income country (LMIC). Within this setting, it covers the most deprived and vulnerable population groups. The MaatHRI platform, although built on models of existing surveillance and research platforms in high income countries, is more advanced in terms of using current best practices for standardisation of data and laboratory parameters, monitoring data and participant safety, and secure transfer of data within and between countries. All biological samples are analysed at the LPL National Reference Laboratory. The precision, performance and quality of each laboratory parameter are documented and maintained to a high level. The laboratory partnership also benefits from subsidised costs from LPL for each test, at a rate that is 40% less than their commercial price, with no additional costs for transportation and project management. The laboratory has also started tests for the MaatHRI project, which they did not offer previously. This involved completing extensive validation processes. In addition to high quality and standardisation of the laboratory procedures, the pseudonymised laboratory model ensures confidentiality of participants and minimises reporting bias.\n\nAnother advantage of the MaatHRI platform is the ability to undertake long term follow-up studies of participants. Identifiable information collected locally from participants helps to locate each participant by hospital staff for follow-up. All studies currently undertaken through the platform have a follow-up component with the potential to generate participant cohorts, based on informed consent, for long term follow-up of the effects of pregnancy complications. Adequate measures have been put in place for securely storing the identifiable information and destroying it after the cohorts for long term follow-up have been established. An independent MaatHRI DSMB monitors data safety and participant confidentiality on an ongoing basis, thereby ensuring confidence and trust on the research platform.\n\nWhile the platform is established and is currently running three epidemiological studies, the process to develop capacity for research and further improving pregnancy care will continue and is an integral part of the MaatHRI collaboration. The focus is on bi-directional skills development and capacity building through mutual learning between the collaborators in India and the UK. The platform is also being used to develop the research capacity of early career researchers (MSc and PhD students and post-doctoral researchers) interested in working in maternal and perinatal health in an LMIC setting.\n\nMaatHRI is a collaboration of hospitals that covers deprived populations, some of which are located in remote rural areas of India. While this provides the opportunity to conduct research to improve the health of mothers and babies in areas of the country that have the highest burden of maternal and perinatal deaths, it also poses challenges related to resources and capacity. Appointing new research nurses to collect data and blood samples ensured that the MaatHRI platform was not depriving the hospitals of their scarce human resource. This has created an employment opportunity for nurses in the field of research, which is not a usual job for trained nurses in India. However, the challenge associated with this was the need for extensive training and constant supervision of the nurses. Furthermore, most of the collaborating hospitals had not been involved in a project of this scale and intensity encompassing not just implementation, but designing, standardising and developing the project as equal partners. Therefore, it took more than 20 months of continuous engagement with staff and collaborators to achieve the desired level of quality and standardisation for the MaatHRI platform.\n\nWithin the resource constraints, a further challenge is achieving a balance between an ideal collaborative research platform and a pragmatic solution. For example, the ideal platform would have collected data electronically on tablets using online forms, but this was not feasible due to a lack of good internet connectivity in the remote hospitals and cultural sensitivities. Therefore, paper forms are used in some hospitals. However, to mitigate risks and as advised by the DSMB, we have developed a documented process of securely storing and destroying the paper forms within an agreed timeline for each hospital.\n\nCosts related to research staff, standardised laboratory parameters, programming data collection forms, and storing data on Microsoft Azure make studies undertaken through the MaatHRI platform more expensive compared with existing similar systems in the UK8, Europe and Australia9. It is our belief, however, that the benefits of generating high quality scientific evidence to answer important and urgent clinical research questions that will save the lives of thousands of future mothers and babies, outweigh these additional costs.\n\n\nConclusion\n\nIn summary, the methods that we have used to develop the MaatHRI platform make it a unique and high-quality research resource using a model that can be replicated in other LMICs. Since being established in September 2018, MaatHRI has already secured further funding, including industry funding. One epidemiological study is complete and two others are in various stages of participant recruitment and data collection. We intend to make the data generated through the MaatHRI platform available to researchers for secondary analysis. In addition to research impact, our approach to building the platform on the premise of equitable partnership between all collaborators and developing research capacity in the collaborating institutions will further contribute to the sustainability of MaatHRI.\n\n\nData availability\n\nNo data is associated with this article.",
"appendix": "Acknowledgments\n\nWe thank Prof. U C Sarma, retired Vice Chancellor of Srimanta Sankaradeva University of Health Sciences, Guwahati, Assam for his valuable contribution in establishing the MaatHRI platform. We also thank Prof. Hem Kanta Sarma, Professor and Head of the Department of Obstetrics and Gynaecology, Jorhat Medical College and Hospital, Assam for his contribution during the initial phase of setting up MaatHRI.\n\nA previous version of this study is available as a preprint on Authorea, https://doi.org/10.22541/au.158931017.74200443\n\n\nReferences\n\nWorld Health Organization: The World health report : 2005 : make every mother and child count. Geneva: World Health Organization, 2005. Reference Source\n\nGraham W, Woodd S, Byass P, et al.: Diversity and divergence: the dynamic burden of poor maternal health. Lancet. 2016; 388(10056): 2164–2175. PubMed Abstract | Publisher Full Text\n\nWHO, UNICEF, UNFPA, et al.: Maternal mortality: Levels and trends 2000 to 2017. Geneva: World Health Organisation, 2019. Reference Source\n\nOffice of Registrar General and Census Commissioner India: Annual Health Survey 2012-13: Fact sheet. New Delhi: Ministry of Home Affairs, Government of India, 2014.\n\nKnight M, Bunch K, Tuffnell D, et al.: Saving Lives, Improving Mothers’ Care - Lessons learned to inform maternity care from the UK and Ireland Confidential Enquiries into Maternal Deaths and Morbidity 2015-17. Oxford: National Perinatal Epidemiology Unit, University of Oxford, 2019. Reference Source\n\nNair M, Choudhury MK, Choudhury SS, et al.: IndOSS-Assam: Investigating the feasibility of introducing a simple maternal morbidity surveillance and research system in Assam, India. BMJ Glob Health. 2016; 1(1): e000024. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNair M, Choudhury MK, Choudhury SS, et al.: The association between maternal anaemia and pregnancy outcomes: a cohort study in Assam, India. BMJ Glob Health. 2016; 1: e000026. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKnight M, Lindquist A: The UK Obstetric Surveillance System: Impact on patient safety. Best Pract Res Clin Obstet Gynaecol. 2013; 27(4): 621–30. PubMed Abstract | Publisher Full Text\n\nKnight M, INOSS: The International Network of Obstetric Survey Systems (INOSS): benefits of multi-country studies of severe and uncommon maternal morbidities. Acta Obstet Gynecol Scand. 2014; 93(2): 127–31. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "66616",
"date": "27 Jul 2020",
"name": "Rajmohan Panda",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors describe the setting up of a consortium and network for doing observational studies and RCTs in the future, as such the study does not describe any specific method. It relies more on the aims and objectives of the network and the process followed. The opportunities and challenges are also listed and potential solution in the real world outlined.\n\nThe network is essential for amplifying the work in maternal and new-born health and can serve as a valuable resource for providing evidence and conducting epidemiological studies in these backwards states in India.\n\nThe description could have touched on some of the areas to make it more robust.\n\nThe process of engagement with partners, the MOUs signed and the IPR shared, the shared value systems, the larger universe of health systems and how they plan for risk communication to audiences as well as decision makers including the obstetricians in the network. The incentives and the academic partnerships could have been better described. Often in the north south partnerships, the hospitals and intervention centres are left our to be nothing more than data collectors. Critical questions like how they will benefit, how the population will benefit, do they envisage community panels patient panels and how do they plan to engage them as part of the network. What academic institutes are involved in the global south, what resources do they bring? How does the platform plan for long term sustainability and scalability?\n\nThe ethics part is well described but does not give clarity of was the ethics for setting up the platform or for carrying out the observational studies.\n\nIs the rationale for developing the new method (or application) clearly explained? Partly\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "5803",
"date": "17 Aug 2020",
"name": "Manisha Nair",
"role": "Author Response",
"response": "We thank the reviewer for the valuable comments. We have updated the manuscript as advised. Please find below our responses indicating how each comment was addressed. We believe that the revised manuscript has benefited from the comments and suggestions.Yours sincerely,Associate Professor Manisha Nair (corresponding author)Nuffield Department of Population Health, University of OxfordResponse to the reviewer’s comments Comment: The authors describe the setting up of a consortium and network for doing observational studies and RCTs in the future, as such the study does not describe any specific method. It relies more on the aims and objectives of the network and the process followed. The opportunities and challenges are also listed and potential solution in the real world outlined. Comment: The network is essential for amplifying the work in maternal and new-born health and can serve as a valuable resource for providing evidence and conducting epidemiological studies in these backwards states in India.Reply: We thank the reviewer for the comment and for endorsing the value of the MaatHRI platform.Comment: The description could have touched on some of the areas to make it more robust. The process of engagement with partners, the MOUs signed and the IPR shared, the shared value systems, the larger universe of health systems and how they plan for risk communication to audiences as well as decision makers including the obstetricians in the network.Reply: We agree with the reviewer that these are important points and have therefore added a the following sections to the ‘methods’ and ‘results’ sectionsIn methods – under ‘Regulatory systems’ we have added the following paragraph:“Academic collaboration agreements were signed between the University of Oxford and all collaborating institutions in India. These were discussed extensively with each institution’s legal team to agree on the terms and conditions including data security and sharing, confidentiality, and intellectual property (IP) rights. To ensure equitable partnership, IP is jointly owned by the collaborating institutions in proportion to the respective contribution of each institution.”We have updated the ‘Regulatory systems’ in the results section as follows:“If the DSMB estimates a potential risk to participant or data security, it will be communicated to the MaatHRI steering committee who has the responsibility to mitigate the problem as soon as possible. The risk and measures taken to mitigate it will also be communicated to all collaborating hospitals and obstetricians as lessons learnt. If the risk or compromise has the potential to cause harm to any participant, the information will be communicated with the participant (at risk) by the site investigator. The DSMB has met twice since MaatHRI was established in September 2018 and membership includes two obstetricians (one from the UK and one from India), one paediatrician (from India), one biostatistician (from the UK), and one expert in bioethics (from India), all with prior experience and expertise in observational epidemiological studies. They were nominated by the study investigators. So far, no risk or compromise to participant or data security has been identified.”The collaboration is based on the shared vision of generating scientific evidence to reduce maternal and perinatal mortality and morbidity in India and other LMICs. We have now stated this clearly in the ‘Results section’ under ‘MaatHRI network of hospitals’.Comment: The incentives and the academic partnerships could have been better described. Often in the north south partnerships, the hospitals and intervention centres are left out to be nothing more than data collectors. Critical questions like how they will benefit, how the population will benefit, do they envisage community panels patient panels and how do they plan to engage them as part of the network.Reply: These are crucial points and I thank the reviewer for raising them. We absolutely agree that there could be power imbalance between collaborating organisations. This is not only true for international collaborations between the North and South, but also between academic collaborations within a country. We have made especial efforts through the regulatory systems to make the partnerships equitable and as already mentioned in the discussion section, our focus is on bi-directional skills development and capacity building through mutual learning between the collaborators in India and the UK.Indian collaborating obstetricians are the ones who suggest topics for research based on the needs of the local population. For example, based on the observation of high number of cases of health failure during pregnancy, this was proposed as a research topic by one of the collaborating obstetricians. In addition, the conditions included in the monthly survey of pregnancy complications and death were decided by the obstetricians. Thus, all research projects undertaken through the platform are co-developed by the Indian and the UK collaborators. Since the research undertaken is only on conditions that are known to affect pregnant women in India, the results from the studies will benefit the future mothers in India as well as mothers in other LMICs. We have now clearly stated the above in the results sections of the revised article in the section ‘Studies currently being undertaken through the MaatHRI platform’.We are working to achieve a more active and extensive process to involve the public and patients (pregnant women, mothers and their families) and civil societies working to improve the health and wellbeing of mothers and babies in India. This has been added to the discussion section.Comment: What academic institutes are involved in the global south, what resources do they bring?Reply: As advised by the reviewer, we have added the following sentence to the results section: “A list of the Indian collaborating institutions are available on the MaatHRI website (https://www.npeu.ox.ac.uk/maathri).”In addition to co-developing the studies, another major resource that the collaborators contribute is their time. None of the site-collaborators or any clinician involved in the project charge their time contribution to the project. This is another testament of their passion towards the shared vision. This information is also added in the revised version.Comment: How does the platform plan for long term sustainability and scalability?Reply: This is an important point and some details were already included in the conclusion section. We have further clarified this by adding the following: “We welcome research organisations from India and the UK to use this standardised research platform to undertake studies that are in line with the vision of MaatHRI and will contribute towards reducing the high burden of maternal and perinatal deaths in India.”Comment: The ethics part is well described but does not give clarity of was the ethics for setting up the platform or for carrying out the observational studies.Reply: As advised by the reviewer, we have reworded sentences to make this more clear: “Ethics approval was obtained for setting up the platform and to undertake the observational studies from……”“The platform and the studies have also been approved by the Government of India’s…..”"
}
]
}
] | 1
|
https://f1000research.com/articles/9-683
|
https://f1000research.com/articles/10-46/v1
|
25 Jan 21
|
{
"type": "Case Report",
"title": "Case Report: unilateral condylar hyperplasia",
"authors": [
"Shishir Shetty",
"Shrihari Guddadararangiah",
"Shrihari Guddadararangiah"
],
"abstract": "Case: This report describes a clinical case of unilateral condylar hyperplasia (CH) with unique, atypical morphology. An important feature of this report is the documentation of a series of clinical photographs of the patient, showing a gradual increase in facial asymmetry associated with the CH. The main symptom reported in this case was facial asymmetry. The main intraoral clinical features observed in the patient were contralateral crossbite and ipsilateral open bite associated with CH. Surgical reshaping of the condyle was the treatment plan for this case. Conclusions: The main take away point from this case is the importance of obtaining previous photographs of the patient at different ages during case diagnosis, which helps the clinician to determine the approximate time of commencement of CH. This case also highlights the imaging features of rarely observed atypical shape of the hyperplastic condyle.",
"keywords": [
"Mandibular condyle",
"hyperplasia",
"panoramic radiography",
"computed tomography"
],
"content": "Introduction\n\nCondylar hyperplasia (CH) is a rare condition associated with excessive condylar bone growth1. Adams first described CH in the year 18362. CH often occurs unilaterally and manifests clinically as facial asymmetry3. Apart from facial asymmetry, occlusal discrepancies, chin deviation and temporomandibular joint discomfort are commonly associated with CH4. CH is known to be self-limiting in nature, normally commences during puberty, progresses gradually and sometimes may be only recognized at the age of 25–30 years5. Diagnosis of CH is usually made using a combination of clinical findings and imaging features6. The aim of this case report is to present clinical and imaging features of unilateral CH. One of the significant points of the present report is that the progression of CH-associated facial asymmetry has been described using a series of photographs of the patient. The unique finding is the atypical shape of the hyperplastic condyle observed.\n\n\nCase report\n\nA 31-year-old Asian male mechanic reported to the dental clinic in January 2013 with complaint of asymmetry of the face over the past five years. The patient felt that the asymmetry had increased during the first three years (of the five-year duration) but had remained constant over the next two years. There was no history of pain, discomfort or clicking sounds in the temporomandibular joints (TMJ). However, the patient had a history of lower jaw trauma during a sports event at the age of 15 years. No other family member had a similar condition.\n\nClinical examination of the patient revealed facial asymmetry due to the deviation of the chin to the left side of the face (Figure 1). Examination of the right TMJ revealed a bony swelling in the right preauricular area. Intraoral examination revealed posterior open bite on the right side (Figure 2) and posterior crossbite on the left side (Figure 3). A series of photographs of the patient at the age of 18 years (Figure 4a), 24 years (Figure 4b), and 27 years (Figure 4c) was evaluated. No evidence of facial asymmetry was noticed at 18 years. Mild features of asymmetry were noticed at 24 years and obvious features of asymmetry were noticed at 27 years. A panoramic radiograph revealed the presence of a beak shaped hyperplastic right condyle (Figure 5a). The posteroanterior skull view revealed increased length of the condylar neck on the right side (Figure 5b). A coronal computed tomography (CT) scan showed enlargement of the right condyle with beak like projection on the medial aspect (Figure 6a). An axial CT scan revealed the antero-medial projection of the beak like enlargement (Figure 6b).\n\nBased on the patient’s history, clinical features and imaging findings a diagnosis of hyperplasia of the right condyle was made. The patient was advised surgical treatment of the CH. Unfortunately, the patient was not willing to undergo surgical correction and long term follow up was not possible.\n\n\nDiscussion\n\nCH is characterized by unilateral or bilateral increase in the volume of the mandibular condyle, often leading to facial asymmetry, jaw deviation and malocclusion7. The exact etiological factor of CH is still unclear, although endocrine alterations, metabolic hyperactivity, trauma, and genetic factors8 have been implicated. In our patient there was history of trauma at the age of 15 years which could be the possible etiological factor. CH occurs predominantly in women9 with a recently published meta-analysis revealing that 64% of cases occurred in females10. In our report the patient was a 31-year-old male. The female predominance has been attributed to hormonal factors, particularly estrogen10,11. Estrogen regulates bone growth and is found in the articular cartilage and growth plates11,12.\n\nCH usually occurs between the ages of 10 and 30 years and most cases occur between adolescence until the end of pubertal growth9. However, some cases of CH also occur after puberty. In our patient the CH seems to have occurred after puberty, as evident in the photographs taken at 24 years. The use of serial photographs of patients with CH at different ages helps physicians to estimate the approximate time of occurrence of the condition. This method was used to estimate the time of occurrence of CH in our patient.\n\nAnother important finding that depends on the time of occurrence of CH is posterior open bite. It has been observed that if CH occurs during puberty the occlusal plane usually inclines as a result of dental compensation, but if CH occurs after puberty posterior open bite may be evident13,14. In our case posterior open bite was observed, suggesting that the CH must have occurred after the growth phase ended.\n\nThe main clinical feature of unilateral CH is enlargement of the same side of the face and flattened appearance of the contralateral side15. These clinical features were observed in our case.\n\nAlthough the combination of clinical findings and imaging features is required for the diagnosis of CH, a radiological examination showing elongation of the neck and head of the condyle is necessary for a definitive diagnosis16. Osteoma, osteochondroma and resorption of the contralateral condyle are the important differential diagnoses for unilateral CH17. Condylar osteomas are extremely rare in occurrence. Condylar osteomas can be differentiated from CH radiographically, since osteomas tend to exhibit a mixed radiolucent-radiopaque appearance, unlike CH which are radiopaque18.\n\nCondylar osteochondromas can be differentiated using CT imaging. In the case of condylar osteochondroma the coronal and sagittal CT sections tend to reveal a growth arising from the morphologically normal condyle in contrast to the uniform enlargement of condylar head which is characteristic of CH19. Panoramic radiographs are useful for comparing both the condyles in a single image although the view is two dimensional20. Panoramic radiographs are good for screening condyles but not considered suitable for the quantitative analysis of condyles and follow up of patients with unilateral CH21. We used a panoramic radiograph to screen our patient and evaluation of the panoramic view revealed hyperplasia of the right condyle. CT imaging aids in multiplanar imaging of the condyles22. A recently conducted retrospective CT based study revealed a significant increase in condylar length and other dimensions on the hyperplastic side when compared to the normal side23. CH characteristically appears as a uniform enlargement of the condylar head19,24. In our patient a beak like projection in the anteromedial direction was observed in the axial CT section which was atypical of CH.\n\nGrowth activity of the CH can be assessed using single-photon emission computed tomography (SPECT)25. In SPECT the unilateral hyperplastic condyle is quantitatively compared to the normal contralateral side26. A 0–5% difference in activity is usually observed between normal condyles. If the difference in activity is greater than 10% between two condyles, CH is suspected in the condyle with increased activity27,28. SPECT could not be performed in our patient because of financial constraints. Prior to initiating treatment for patients with CH several factors such as the level of facial asymmetry, psycho-social consequences of the facial change, functional changes and malocclusion have to be considered4. Treatment options for CH include high condylectomy with or without orthognathic surgery and orthodontic treatment29. Unfortunately, our patient was not willing to undergo surgical treatment of CH.\n\n\nConclusion\n\nPrompt diagnosis is very important for successful management of CH. Apart from a patient’s history and clinical findings, serial photographs of the patient from the past 10 to 15 years also provides vital information about the approximate time of occurrence and progression of the CH. Hence it is advisable to study serial photographs of patients with CH during the diagnostic stage.\n\n\nInformed consent\n\nWritten informed consent for publication of their clinical details and clinical images was obtained from the patient.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.",
"appendix": "References\n\nAlmeida LE, Zacharias J, Pierce S: Condylar hyperplasia: An updated review of the literature. Korean J Orthod. 2015; 45(6): 333–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIannetti G, Cascone P, Belli E, et al.: Condylar hyperplasia: Cephalometric study, treatment planning, and surgical correction (our experience). Oral Surg Oral Med Oral Pathol. 1989; 68(6): 673–81. PubMed Abstract | Publisher Full Text\n\nHovell JH: Condylar hyperplasia. Br J Oral Surg. 1963; 1: 105–111. PubMed Abstract | Publisher Full Text\n\nOlate S, Netto HD, Rodriguez-Chessa J, et al.: Mandible condylar hyperplasia: a review of diagnosis and treatment protocol. Int J Clin Exp Med. 2013; 6(9): 727–37. PubMed Abstract | Free Full Text\n\nBharathi SC, Senthilnathan S, Kumar LD, et al.: Unilateral condylar hyperplasia: A case report and review of literature. J Int Soc Prev Community Dent. 2014 ; 4(1): 67–70. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShankar U, Chandra S, Raju BH, et al.: Condylar hyperplasia. J Contemp Dent Pract. 2012; 13(6): 914–7.\n\nGelada K, Halli R, Hebbale M, et al.: Unilateral Condylar Hyperplasia: A Case Report. Ann Clin Case Rep. 2018; 3: 1524. Reference Source\n\nHansson T, Oberg T, Carlsson GE, et al.: Thickness of the soft tissue layers and the articular disk in the temporomandibular joint. Acta Odontol Scand. 1977; 35(2): 77–83. PubMed Abstract | Publisher Full Text\n\nNitzan DW, Katsnelson A, Bermanis I, et al.: The clinical characteristics of condylar hyperplasia: experience with 61 patients. J Oral Maxillofac Surg. 2008; 66(2): 312–318. PubMed Abstract | Publisher Full Text\n\nRaijmakers PG, Karssemakers LH, Tuinzing DB: Female predominance and effect of gender on unilateral condylar hyperplasia: a review and meta-analysis. J Oral Maxillofac Surg. 2012; 70(1): e72–e76. PubMed Abstract | Publisher Full Text\n\nTalwar RM, Wong BS, Svoboda K, et al.: Effects of estrogen on chondrocyte proliferation and collagen synthesis in skeletally mature articular cartilage. J Oral Maxillofac Surg. 2006; 64(4): 600–609. PubMed Abstract | Publisher Full Text\n\nYu S, Xing X, Liang S, et al.: Locally synthesized estrogen plays an important role in the development of TMD. Med Hypotheses. 2009; 72(6): 720–722. PubMed Abstract | Publisher Full Text\n\nMehrotra D, Dhasmana S, Kamboj M, et al.: Condylar hyperplasia and facial asymmetry: report of five cases. J Maxillofac Oral Surg. 2011; 10(1): 50–56. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNorman JE, Painter DM: Hyperplasia of the mandibular condyle. A historical review of important early cases with a presentation and analysis of twelve patients. J Maxillofac Surg. 1980; 8(3): 161–75. PubMed Abstract\n\nLópez DF, Corral CM: Condylar hyperplasia: characteristics, manifestations, diagnosis and treatment. A topic review. Rev Fac Odontol Univ Antioq. 2015; 26(21): 425–446. Reference Source\n\nWen B, Shen Y, Wang CY: Clinical Value of 99Tcm-MDP SPECT Bone Scintigraphy in the Diagnosis of Unilateral Condylar Hyperplasia. Sci World J. 2014; 2014: 256256. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTripathi T, Srivastava D, Neha, et al.: Differential diagnosis and treatment of condylar hyperplasia. J Clin Orthod. 2019; 53(1): 29–38. PubMed Abstract\n\nValente L, Tieghi R, Mandrioli S, et al.: Mandibular Condyle Osteoma. Ann Maxillofac Surg. 2019; 9(2): 434–438. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAndrade NN, Gandhewar TM, Kapoor P, et al.: Osteochondroma of the mandibular condyle - Report of an atypical case and the importance of computed tomography. J Oral Biol Craniofac Res. 2014; 4(3): 208–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKubota Y, Takenoshita Y, Takamori K, et al.: Levandoski Panographic analysis in the diagnosis of hyperplasia of the coronoid process. Br J Oral Maxillofac Surg. 1999; 37(5): 409–411. PubMed Abstract | Publisher Full Text\n\nNolte JW, Karssemakers LH, Grootendorst DC, et al.: Panoramic imaging is not suitable for quantitative evaluation, classification, and follow up in unilateral condylar hyperplasia. Br J Oral Maxillofac Surg. 2015; 53(5): 446–50. PubMed Abstract | Publisher Full Text\n\nMutoh Y, Ohashi Y, Uchiyama N, et al.: Three-dimensional analysis of condylar hyperplasia with computed tomography. J Craniomaxillofac Surg. 1991; 19(2): 49–55. PubMed Abstract | Publisher Full Text\n\nLópez DF, Botero JR, Muñoz JM, et al.: Mandibular and temporomandibular morphologic characteristics of patients with suspected unilateral condylar hyperplasia: a CT study. Dental Press J Orthod. 2020; 25(2): 61–68. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAvinash KR, Rajagopal KV, Ramakrishnaiah RH, et al.: Computed tomographic features of mandibular osteochondroma. case report. Dentomaxillofac Radiol. 2007; 36(7): 434–436. PubMed Abstract | Publisher Full Text\n\nAlyamani A, Abuzinada S: Management of patients with condylar hyperplasia: A diverse experience with 18 patients. Ann Maxillofac Surg. 2012; 2(1): 17–23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPripatnanont P, Vittayakittipong P, Markmanee U, et al.: The use of SPECT to evaluate growth cessation of the mandible in unilateral condylar hyperplasia. Int J Oral Maxillofac Surg. 2005; 34(4): 364–368. PubMed Abstract | Publisher Full Text\n\nSaridin CP, Gilijamse M, Kuik DJ, et al.: Evaluation of temporomandibular function after high partial condylectomy because of unilateral condylar hyperactivity. J Oral Maxillofac Surg. 2010; 68(5): 1094–1099. PubMed Abstract | Publisher Full Text\n\nYang Z, Reed T, Longino BH: Bone Scintigraphy SPECT/CT Evaluation of Mandibular Condylar Hyperplasia. J Nucl Med Technol. 2016; 44(1): 49–51. PubMed Abstract | Publisher Full Text\n\nSembronio S, Tel A, Costa F, et al.: An Updated Protocol for the Treatment of Condylar Hyperplasia: Computer-Guided Proportional Condylectomy. J Oral Maxillofac Surg. 2019; 77(7): 1457–1465. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "78064",
"date": "29 Jan 2021",
"name": "Prashant M. Battepati",
"expertise": [
"Reviewer Expertise Pediatric dentistry",
"Preventive Dentistry"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript provides sufficient details regarding the specified case. Though the conclusion or the findings of the case report does not add to the available information, the case report as a whole can prove as a source of information for the practitioners looking for information related to such specific cases.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes",
"responses": []
},
{
"id": "78061",
"date": "02 Feb 2021",
"name": "Renita Lorina Castelino",
"expertise": [
"Reviewer Expertise Oral Medicine",
"Oral Radiology",
"Forensic odontology",
"Innovations in dentistry",
"3D imaging and printing",
"Lasers."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIt is a well written manuscript covering all the required aspects for diagnosing a case of condylar hyperplasia. The authors have provided a well documented case substantiated with appropriate images and radiographs. This will help students and practitioners to comprehend the clinical and radiographic features of condylar hyperplasia and its progression.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-46
|
https://f1000research.com/articles/10-44/v1
|
23 Jan 21
|
{
"type": "Research Article",
"title": "Reverse vaccinology approach towards the in-silico multiepitope vaccine development against SARS-CoV-2",
"authors": [
"Vipul Kumar",
"Sudhakar Kancharla",
"Prachetha Kolli",
"Manoj Jena",
"Vipul Kumar",
"Sudhakar Kancharla",
"Prachetha Kolli"
],
"abstract": "Background: The novel severe acute respiratory syndrome related corona virus-2 (SARS-CoV-2) belongs to the “Coronaviridae” family and order “Nidovirales”, which has caused the pandemic coronavirus disease 2019 (COVID-19). SARS-CoV-2 has been spread in more than a 100 countries, and more than a million have lost their lives. Vaccination and immunization could be an effective strategy to combat fatal COVID-19. Methods: For identification of effective vaccine candidate against COVID-19, various immunoinformatics online tools and softwares were used to predict epitopes. Cytotoxic T cell epitopes, helper T cell epitopes, and B cell epitopes from three structural polyproteins (Spike, Membrane, and Nucleocapsid (SMN) based on the binding affinity towards MHC, antigenicity, non-allergenicity, and non-toxicity) were identified for vaccine development. The multiepitope based vaccine was constructed linking two additional adjuvants human beta-defensin-3 and human beta-defensin-2 at N and C terminal, respectively. Results: The constructed vaccine sequence was found to be a good antigen and non-allergen for the human body. The constructed vaccine was docked with the TLR-3 receptor. The docked complex was further taken for molecular dynamics simulations and RMSD was calculated, which showed stable binding of the complex. The codon adaptation index (CAI) of 0.92 and GC content of 55.5% for E. coli (K12 strain) suggested efficient expression of the predicted vaccine. Conclusion: The current study can be helpful in the reduction of time and cost for further experimental validations and could give a valuable contribution against this pandemic.",
"keywords": [
"SARS-CoV-2",
"COVID-19",
"Immunoinformatics",
"multiepitope",
"docking",
"simulations."
],
"content": "Introduction\n\nCoronaviruses (CoVs) belong to the family of coronaviridae in the order Nidovirales, and have single-strand positive-sense RNA1. The size of the RNA of coronavirus is the largest among the viruses (~30 kb)2. They have glycoprotein projections on the envelope, which gives the corona appearance. CoVs are pathogens mainly involved in respiratory and gastrointestinal diseases in a wide range of animals and humans1,2. CoVs are divided into four sub-categories, namely alpha, beta, gamma, and delta, out of which alpha and beta coronavirus are known to infect humans1,3. From alpha and beta, four strains are responsible for the common cold, and two strains were found to be responsible for severe acute respiratory syndrome (SARS-CoV) and Middle East respiratory syndrome CoV (MERS-CoV)4,5. Recently in December 2019, a novel coronavirus SARS-CoV-2 was detected from patients of novel coronavirus disease 2019 (COVID-19) in the Wuhan, China6-8. The symptoms of COVID-19 infection include headache, fever, pneumonia, and asthenia9,10. A severe and fatal outbreak of this virus has taken many lives and has created enormous economic loss worldwide. The treatment and prevention from this infection is the need of the hour.\n\nCOVID-19 has given a serious and tough challenge to biomedical research scientists and researchers all around the world. Much research is looking at repurposing antiviral drugs, and developing new drugs and vaccines against the SARS-CoV-211-15. Here, an attempt has been given to construct an in-silico vaccine, which can be further validated through experimental assays and could play a major role in the management of this pandemic. In this study, three structural proteins form SARS-CoV-2, based on the antigenicity has been selected for the construction of the vaccine. The first structural protein is Spike (S) glycoprotein, which has been reported to be a crucial surface protein of SARS-CoV-2, which facilitates the entry of the virus inside the host cell. It has been reported that for the entry of the SARS-CoV-2, S protein first binds with Angiotensin-Converting Enzyme-2 (ACE-2) receptor, and then is primed by the host serine protease (TMPRSS2)16,17. This priming of the S protein lets it to fuse into the host cell membrane and entry inside the cell. The second crucial structural protein, which induces a strong immune response, is Membrane (M) glycoprotein. It plays a crucial role in virus morphogenesis and assembly by interacting with several other viral proteins18. The third structural protein chosen for vaccine construct is Nucleocapsid (N) phosphoprotein; it links the viral genome to the envelope. It consists of two domains, N terminal and C terminal, and both can bind to RNA. It has been reported that the C terminal domain facilitates the physical interaction of the RNA genome and envelope19,20. All these three structural proteins are predicted to be good antigens and could induce the immune response.\n\nIn this pandemic situation, an immunoinformatics approach could be a fast, scientifically sound, and reliable option for quicker vaccine development. These three proteins chosen for the present study were predicted to be good antigens, which gives the opportunity to predict B and T cell epitopes. When naive B cells interact with the antigenic B cell epitopes via its transmembrane bound antibody, they differentiate into two types of cells plasma and memory cells21,22. Plasma cells lack receptors, but they produce a large number of antibodies against the antigen. Memory cells express membrane-bound antibody molecules, but they are functionally inactive unless they encounter the same antigen again23-25. Furthermore, T cell epitopes are recognized by Major Histocompatibility Class (MHC), a glycoprotein present on the variety of the cells, which display the antigen to T cells26,27. Antigen-presenting MHCs are divided into major two classes, MHC class-I are expressed on nucleated cells while MHC class-II are only expressed by antigen-presenting cells. Class II MHC interacts with T helper cells and activates B cells via cytokines, while Class-I MHC interacts with cytotoxic T cells, which kills virus-infected host cells28,29. Further to know the ability of a constructed vaccine for inducing innate as well as antigen-specific acquired immunity, the constructed vaccine must dock with Toll-Like Receptor-3 (TLR-3). TLRs are mainly expressed on various leukocytes such as dendritic cells, natural killer cells, and cells of adaptive immunity such as T cells and B cells30,31. Hence, in this study, an attempt has been made to construct the multiepitope vaccine consists of Helper T cells (HTLs), Cytotoxic T Cells (HTLs), and B cell epitopes, which could interact with TLR-3 and generate the immune response. This constructed multiepitope vaccine may induce both humoral as well as cell-mediated immune responses.\n\n\nMethods\n\nThe complete sequence of all three structural polyproteins from SARS-CoV-2 reference sequence (NC045512.2) were retrieved from NCBI on the basis of their antigenicity. The spike (S) glycoprotein (YP_009724390.1), Membrane (M) glycoprotein (YP_009724393.1) and Nucleocapsid (N) phosphoprotein (YP_009724397.2) were retrieved in FASTA format. These three proteins together are referred to as SMN (Spike, Membrane, and Nucleocapsid) polyprotein in this study.\n\nFirst, the CTL epitopes for SMN polyproteins were predicted using Netctl 1.2 server32. Prediction of the epitopes depends on three major attributes: (1) binding affinity of MHC-1 class; (2) ability of the proteasome cleavage; and (3) TAP transport efficiency. The first two are predicted with the artificial neural network algorithm while third one using weight matrix. For the prediction of the epitopes threshold for epitopes, identification was chosen to be 0.75, weight on C terminal cleavage was set on 0.15, while the weight on TAP transport efficiency was set on 0.05. The predicted epitopes were ranked according to the combined score.\n\nFor the prediction of HTL epitopes, the IEDB MHC II server was used33. The species/locus was selected as Human/HLA-DR, and a 7-allele HLA reference set was selected for the prediction. Further, 15 mer length of the epitopes were retrieved and ranked according to the percentile. The percentile rank is given after comparing the peptides score with five million 15 mers from the SWISSPROT database. The higher percentile value means a lower binding affinity of MHC-II. For further refinements of the HTL epitopes, these selected HTL epitopes were subjected to investigate whether they can induce IFN gamma immune response using the IFN epitope server34. For IFN gamma inducing epitopes selection, the Motif/SVM hybrid approach was chosen, and the model was set to be IFN gamma versus non-IFN gamma. Finally, the epitopes whose results were positive for the IFN gamma response were chosen for the in-silico vaccine development.\n\nB cell epitopes were predicted using the ABCpred server35. This server predicts B cell epitopes using recurrent neural network algorithm. For the identification of the epitopes, the threshold was set on 0.51, while the window length for the prediction was chosen to be 16, keeping overlapping filter on. Top predicted epitopes having scored more than 0.9 was only chosen for the development of the candidate vaccine. Further, after the construction of the vaccine, linear as well as discontinuous conformational B cell epitopes were identified in the vaccine construct using ElliPro, an online server36. Elipro predicts the antibody epitopes taking protein antigen tertiary structure as input.\n\nThe important attributes such as the antigenicity, allergenicity and toxicity were predicted for all the predicted epitopes individually as well as after construction of the vaccine. First of all, the antigenicity was investigated using the VaxiJen 2.0 server37, and only probable antigen epitopes were chosen for the construction of the vaccine. Further, the allergenicity was predicted using the AlgPred server38 and only non-allergenic epitopes were selected. Finally, all the epitopes were investigated for toxicity using the ToxinPred server39 and non-toxic epitopes were selected. All the predicted epitopes had to cross all these barriers. The overall construct of the vaccine was also tested for these attributes.\n\nThe vaccine sequence was constructed using the best identified CTL, HTL, and B cell epitopes. For the construction of the sequence at the N terminal and C terminal, an adjuvant was added using EAAAK linkers. While HTL epitopes were linked using GPGPG, linkers and CTL epitopes were linked using AAY linkers. In the C terminal, HHHHHH was added for the easy purification of the vaccine.\n\nThe physicochemical properties such as molecular weight, PI, half-life, aliphatic index, and hydropathicity were predicted using online tool ProtParam40.\n\nProtein secondary structure prediction gives further opportunity to predict the tertiary structure as well as gives information about the activity and function of the protein. The secondary structure of the final multiepitope vaccine sequence was predicted by the free online web tool CFSSP41.\n\nThe tertiary structure of the constructed vaccine was predicted using the Rosetta web tool42. Rosetta tool applies a deep neural network algorithm to predict the inter-residue distances as well as orientations. Then these orientations are converted to smooth inter-residue constraints followed by gradient descent energy minimization. Further, the coarse-grained models are generated, and full atom refinement is done. The validation of the model has been done through Ramachandran plot analysis using VADAR web tool43. Further, the modelled structure was validated through the ProSA web tool44, which gives the quality Z score of the modelled protein based on the already known similar size of the proteins crystal structures.\n\nTo achieve a more stable structure, the predicted structure was further taken for molecular dynamics (MD) simulations using Gromacs software45. The structure was minimized using the steepest descent algorithm with 50000 steps, followed by NVT and NPT equilibration for 100 picoseconds, followed by MD simulations of 500 picoseconds. The last frame of the MD trajectory was taken for further analysis.\n\nFor molecular docking, the last frame from the MD simulations of the constructed vaccine was taken, and the TLR-3 structure was retrieved from Protein Data Bank (PDB; ID 1ZIW). The downloaded structure was prepared and processed for docking using dock prep tool UCSF Chimera software. For the docking, the vaccine construct and TLR-3 was uploaded to patchDock server46. Further, for refinement of the rigid body molecular docking solutions, FireDock server was used47. It gave the best 10 docked confirmation based on global energy and Van der Waal’s interactions.\n\nFinally, for expressing the constructed multiepitope, the vaccine needs to be expressed in the suitable vector inside the prokaryotic system. Hence, reverse translation and codon optimization were analysed using the Java codon adaptation web tool (Jcat)48. The codon optimization was performed for E. coli strain K12 as a host. Jcat gives the codon adaption index (CAI) and percentage GC content as output. The CAI gives the information of codon usage, generally score between 1 and 0.8, while GC contents should be between 40 % to 70%, values lying outside the given margin is suggested to be inefficient49.\n\n\nResults\n\nThe amino acid sequence of all three (SMN) structural proteins were retrieved from the NCBI database in fasta format. The proteins were investigated for antigenicity by Vaxijen web tool, and it was found that all the three chosen proteins could be good antigens. The default threshold of 0.4 was chosen as the criteria for the antigenicity in the Vaxijen tool. The spike protein showed a score of 0.46; membrane glycoprotein showed a score of 0.51; while nucleocapsid protein showed a score of 0.50. Hence, all three proteins were chosen for further predictions of B cell and T cell epitopes and the construction of the vaccine.\n\nCTL epitopes were predicted using Netctl 1.2 server for all the three selected proteins. A total of 38 epitopes was predicted from spike glycoprotein; 10 epitopes were predicted from membrane glycoprotein; while 9 were predicted from nucleocapsid protein. Out of all these predicted CTL epitopes, only 8 were selected for the construction of the vaccine, based on a high binding affinity towards MHC-I, antigenicity, non-allergenicity, and non-toxicity predictions, as shown in Table 1.\n\nHTL epitopes were predicted using the IEDB MHC II server for all the three SMN structural proteins. Finally, 4 HTL epitopes were selected on the basis of binding affinity, antigenicity, non-allergenicity, and non-toxicity, as shown in Table 2. Four human alleles and position of predicted epitopes are HLA-DRB1*07:01 (166-180), HLA-DRB4*01:01 (298-312), HLA-DRB5*01:01 (232-246), HLA-DRB5*01:01 (345-359).\n\nFor the prediction of B cell epitopes, ABCpred server was used. Based on the binding score (>0.9), non-allergenicity and non-toxicity, a total of four B cell epitopes were finally selected, as shown in Table 3.\n\nThe four B cell epitopes, four HTL epitopes and 8 CTL epitopes were selected for vaccine construction, which fulfilled all the criteria of binding affinity, antigenicity, non-toxicity and non-allergenicity. Besides these epitopes, two adjuvants were also added at the N terminal (human beta defensin-3) and at C terminal (human beta defensin-2) of the vaccine to increase the antigenicity. Adjuvant were linked via EAAAK linkers to the epitopes, HTL epitopes were linked via GPGPG linkers, while CTL epitopes were linked with AAY linkers, as shown in Figure 1. The constructed vaccine sequence was again checked for antigenicity, non-allergenicity, non-toxicity and it fulfilled all the criteria.\n\nAt the C terminal, an adjuvant human B defesin3 has been added, and then it is linked with B cell epitopes using EAAAK linkers. B cell epitopes are linked with HTL with GPGPG linkers, and HTL are linked with CTL with AYY linkers. At N terminal, another adjuvant human B defensin-2 has been linked with six histidine sequences.\n\nThe physiochemical parameters of the vaccine sequence were predicted by the ProtParam server. The molecular weight of the construct was predicted to be 38.8 KDa, and the theoretical PI value was 9.92. The predicted half-life in E. coli was more than 10 hours, and the instability index in the test tube was found to be stable. The aliphatic value of the vaccine sequence was 58.7 and the grand average of hydropathicity (GRAVY) was -0.348.\n\nSecondary structure prediction was made using the CFSSP web tool. The result showed the presence of helix at 44.5%, sheet at 35.6%, and turns at 14 %.\n\nThe 3D structure of the multiepitope predicted vaccine was predicted using the Rosetta web tool. It uses de-novo structure prediction using deep neural network algorithm to predict the inter-residue distances as well as orientations. Then these orientations are converted to smooth inter-residue constraints followed by gradient descent energy minimization. Further, coarse-grained models are generated, and full atom refinement is done. It gave five best-predicted models, and based on the TM score, one model was selected for further investigation, as shown in Figure 2A. Further to validate the predicted model, Ramachandran plot analysis was done, and results showed that 96.3% residues were in the favourable region, 2.5% were in the allowed region while ~ 1% were in the outlier region (Figure 3). Additionally, the PROSA web tool was used to predict the quality of the modelled vaccine, which predicted a Z score of -6.34. Ramachandran plot and Z score have suggested that the predicted model of protein was valid and could be taken for further analysis.\n\n(A) The crude 3D modelled structure of the vaccine (grey) has been superimposed with the simulated model (yellow). (B) The top 3 conformational B cell epitopes predicted in the vaccine has been shown with yellow spheres.\n\nElipro predicts the antibody epitopes taking protein 3D structure as input. Linear B, as well as discontinuous conformational epitopes, were identified in the vaccine construct using ElliPro, an online server. A total of 8 linear epitopes were predicted, and the sequence of the top 3 epitopes have been reported in Table 4 and has been shown structurally in Figure 2B. Various discontinuous epitope residues were predicted from vaccine sequence length 232-253 (21 epitope residues), between 299-357 (55 epitope residues), between 1-54 (52 epitope residues), between 69-128 (33 epitope residues) and between 168-176 (9 epitope residues) were predicted. The individual score of each of the discontinuous epitopes has been shown in Figure 4.\n\nThe modelled structure of the vaccine was taken through energy minimization, equilibration, and MD simulations before docking. The last frame from the simulated trajectory was taken further for docking. The simulated structure has been compared with the crude modelled structure, as shown in Figure 2A. The TLR-3 structure was retrieved from PDB (ID 1ZIW). The downloaded structure was prepared and processed for docking using the dock prep tool UCSF Chimera software50. The simulation was done using the PatchDock server and further refined using FireDock. The best-docked complex had global energy of -14.91 Kcal/mol, and attractive Van der Waal’s energy was -18.1 Kcal/mol, which shows a decent binding affinity of the vaccine towards TLR-3. Further, the best binding pose was investigated for polar interactions using discovery studio visualizer51 between TLR-3 and vaccine, and it was found that GLN352, SER428, ILE370 of TLR-3 was making the hydrogen bond with TYR260, ARG321, and LYS166 of vaccine respectively (Figure 5).\n\nJcat was used for the optimization of the codon for the proper expression of the protein. E. coli strain K12 was chosen as a host, with additional options such as avoid rho-independent transcription terminators, avoid prokaryotic ribosome binding sites, and avoid Cleavage Sites of Restriction Enzymes. The CAI gives the information of codon usage, generally score between 1 and 0.8, while GC contents should be between 40 % to 70%, values lie outside the given margin is suggested to be inefficient. CAI of the optimized nucleotide sequence of the vaccine was found to be 0.92, with a GC content of 55.6%, which indicates the effective expression of the protein in the E. coli.\n\n\nDiscussion\n\nThe structural polyproteins from SARS-CoV-2 were selected for developing a multiepitope based vaccine. Initially, four structural proteins were chosen, namely spike, membrane, envelope, and nucleocapsid protein, based on their antigenicity prediction, role in facilitating the entry of the virus, and packaging inside the host cells. However, when CTL, HTL, and B cell epitopes were predicted, it was found that epitopes from envelope protein were not able to satisfy the criteria of non-allergenicity and antigenicity simultaneously, and hence envelope protein was not considered further in the study. While epitopes selected from SMN structural polyprotein were satisfying, all the criteria such as antigenicity, non-allergenicity, non-toxicity, and high binding affinity towards MHC and also HTL epitopes were able to induce IFN gamma immune response. The constructed multiepitope vaccine from the selected epitopes from all the three SMN polyproteins was again investigated for antigenicity, and it was found that vaccine construct predicted to be a potent antigen with score 0.60 (predicted by Vaxijen). Further, the vaccine construct was looked for its allergenicity, and it was found that the vaccine was a non-allergen with a score of -0.59 (threshold was set on -0.4, predicted by AlgPred tool). Further physicochemical parameters were analysed for vaccine sequence, and it was predicted to have a molecular weight of 38.8 KDa, PI of 9.92, and half-life inside the E. coli >10 hours, which shows that protein can easily express and isolated. The 3D model of the constructed vaccine sequence was predicted from the Rosetta web server. Eight conformational linear B cell epitopes were found in the modelled structure of the vaccine, as predicted by the Elipro web tool. It was evident from the prediction that the constructed vaccine model could easily produce adaptive immune response specific to the SARS-CoV-2 antigens. Further, to investigate the ability of the modelled vaccine to interact with TLR receptors on immune cells, the TLR-3 receptor was docked with the modelled vaccine. The results showed that the modelled vaccine had a good binding affinity towards TLR-3, and it was found that GLN352, SER428, ILE370 of TLR-3 was making the hydrogen bond with TYR260, ARG321, and LYS166 of vaccine, respectively. This interaction of vaccine with TLR-3 was predicting that vaccine have the potential to generate both innate as well as adaptive humoral and cell-mediated immune responses. For efficient expression of the protein inside the E. coli, codon optimization was done to improve the translation and transcription efficiency. The constructed vaccine sequence was reverse translated, and CAI and GC content were assessed, taking E. coli (K12) as a host organism. The CAI index of 0.92 and GC content of 55.6 % and half-life was already predicted to be more than 10 hours, suggests the efficient expression of recombinant protein inside the E. coli. This immunoinformatics study suggests that the predicted vaccine could generate specific adaptive immunity against SARS-CoV-2 and could provide a valuable contribution to the management of the COVID-19. This predicted vaccine candidate strongly warrant in-vitro and in-vivo study for the practical implications.\n\n\nConclusions\n\nIn this study, a multiepitope (CTL, HTL, and B cell) vaccine construct has been predicted and modelled through immunoinformatics techniques. The predictions suggest that the constructed vaccine could generate both humoral and cell-based adaptive immunity towards SARS-CoV-2. Further, it was also predicted that it may easily be expressed inside the E. coli strain (K12). This immunoinformatics study may reduce the expenditure and time for vaccine research and may give a significant value in the management of COVID-19. This in silico prediction warrants the in-vitro and in-vivo study to test the practical implications of the predicted vaccine.\n\n\nData availability\n\nNCBI: Severe acute respiratory syndrome coronavirus 2 isolate Wuhan-Hu-1, complete genome, Accession number NC045512.2: https://www.ncbi.nlm.nih.gov/nuccore/NC_045512.2/\n\nNCIB: surface glycoprotein [Severe acute respiratory syndrome coronavirus 2], Accession number YP_009724390.1: https://www.ncbi.nlm.nih.gov/protein/YP_009724390.1/\n\nNCBI: membrane glycoprotein [Severe acute respiratory syndrome coronavirus 2], Accession number YP_009724393.1: https://www.ncbi.nlm.nih.gov/protein/YP_009724393.1\n\nNCBI: nucleocapsid phosphoprotein [Severe acute respiratory syndrome coronavirus 2], Accession number YP_009724397.2: https://www.ncbi.nlm.nih.gov/protein/YP_009724397.2\n\nProtein Data Bank: Human Toll-like Receptor 3 extracellular domain structure, Accession number 1ZIW: http://doi.org/10.2210/pdb1ZIW/pdb",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nA previous version of this article is available on Research Square: https://doi.org/10.21203/rs.3.rs-31779/v1\n\n\nReferences\n\nDenison MR, et al.: Coronaviruses: an RNA proofreading machine regulates replication fidelity and diversity. RNA Biol 2011; 8(2): 270–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFehr AR, Perlman S: Coronaviruses: an overview of their replication and pathogenesis. Methods Mol Biol 2015; 1282: 1–23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYe ZW, et al.: Zoonotic origins of human coronaviruses. Int J Biol Sci 2020; 16(10): 1686–97. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBhattacharya S, et al.: Recent apprise on coronavirus and its terrible insinuations. Virusdisease 2020; 1–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVabret A, et al.: Human coronavirus NL63, France. Emerg Infect Dis 2005; 11(8): 1225–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi Q, et al.: Early Transmission Dynamics in Wuhan, China, of Novel Coronavirus-Infected Pneumonia. N Engl J Med 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHui DS, et al.: The continuing 2019-nCoV epidemic threat of novel coronaviruses to global health—The latest 2019 novel coronavirus outbreak in Wuhan, China. Int J Infect Dis 2020; 91: 264–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuang C, et al.: Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. Lancet 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLechien JR, et al.: Olfactory and gustatory dysfunctions as a clinical presentation of mild-to-moderate forms of the coronavirus disease (COVID-19): a multicenter European study. Eur Arch Otorhinolaryngol 2020; 1–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNicastri E, et al.: National institute for the infectious diseases \"L. Spallanzani\", IRCCS. Recommendations for COVID-19 clinical management. Infect Dis Rep 2020; 12(1). PubMed Abstract | Publisher Full Text | Free Full Text\n\nSindhu T, et al.: Antiviral screening of Clerodol derivatives as COV 2 main protease inhibitor in Novel Corona Virus Disease: In silico approaches. Asian J Pharm Clin Res 2020; 10(2): 60–4. Publisher Full Text\n\nYadav AR, Mohite SK: A Novel Approach for Treatment of COVID-19 with Convalescent Plasma. RJPDFT 2020; 12(3): 227–30. Publisher Full Text\n\nShankhdhar PK, et al.: Turmeric: Plant Immunobooster against COVID-19. Res Rev J Pharmacogn Phytochem 2020; 12(3): 174–7. Publisher Full Text\n\nGoswami S, et al.: A Meticulous Interpretation on a Sanguinary Disease COVID-19. Res Rev J Pharmacogn Phytochem 2020; 12(3): 231–3. Publisher Full Text\n\nNaresh B: A Review of the 2019 Novel Coronavirus (COVID-19) Pandemic. Asian J Pharm Sci 2020; 10(3): 233–8. Publisher Full Text\n\nHoffmann M, et al.: SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor. Cell 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIwata-Yoshikawa N, et al.: TMPRSS2 Contributes to Virus Spread and Immunopathology in the Airways of Murine Models after Coronavirus Infection. J Virol 2019; 93(6). PubMed Abstract | Publisher Full Text | Free Full Text\n\nPrajapat M, et al.: Drug targets for corona virus: A systematic review. Indian J Pharmacol 2020; 52(1): 56–65. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrunel J, et al.: Sequence of events in measles virus replication: role of phosphoprotein-nucleocapsid interactions. J Virol 2014; 88(18): 10851–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSurjit M, Lal SK: The SARS-CoV nucleocapsid protein: a protein with multifarious activities. Infect Genet Evol 2008; 8(4): 397–405. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNemazee D: Mechanisms of central tolerance for B cells. Nat Rev Immunol 2017; 17(5): 281. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOleinika K, et al.: B Cell Activation and B Cell Tolerance. The Autoimmune Diseases: Elsevier 2020; 171–87. Publisher Full Text\n\nSchmitt ME, et al.: The B cell antigen receptor of IgE-switched plasma cells regulates memory IgE responses. J Allergy Clin Immunol 2020. PubMed Abstract | Publisher Full Text\n\nGold MR, Reth MG: Antigen receptor function in the context of the nanoscale organization of the B cell membrane. Annu Rev Immunol 2019; 37: 97–123. PubMed Abstract | Publisher Full Text\n\nBrink R, Phan TG: Self-reactive B cells in the germinal center reaction. Annu Rev Immunol 2018; 36: 339–57. PubMed Abstract | Publisher Full Text\n\nKarunakaran KP, et al.: Discordance in the Epithelial Cell-Dendritic Cell Major Histocompatibility Complex Class II Immunoproteome: Implications for Chlamydia Vaccine Development. J Infect Dis 2020; 221(5): 841–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPeters B, et al.: T Cell Epitope Predictions. Annu Rev Immunol 2020; 38. Publisher Full Text\n\nGaudino SJ, Kumar P: Cross-talk between antigen presenting cells and T cells impacts intestinal homeostasis, bacterial infections, and tumorigenesis. Front Immunol 2019; 10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMiller CT: The Expression Of Connexin-43 By CD11c+ Dendritic Cells Is Required to Maintain CD4+ Foxp3+ Regulatory T Cell Population in Peripheral Lymphoid Organs.2019. Publisher Full Text\n\nRich RR, Chaplin DD: The human immune response. Clinical Immunology: Elsevier 2019; e1: 3–17. Publisher Full Text\n\nToubi E, Vadasz Z: Innate immune-responses and their role in driving autoimmunity. Autoimmun Rev 2019; 18(3): 306–11. PubMed Abstract | Publisher Full Text\n\nLarsen MV, et al.: Large-scale validation of methods for cytotoxic T-lymphocyte epitope prediction. BMC Bioinformatics 2007; 8: 424. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang Q, et al.: Immune epitope database analysis resource (IEDB-AR). Nucleic Acids Res 2008; 36(Web Server issue): W513–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDhanda SK, et al.: Designing of interferon-gamma inducing MHC class-II binders. Biol Direct 2013; 8: 30. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSaha S, Raghava GP: Prediction of continuous B-cell epitopes in an antigen using recurrent neural network. Proteins 2006; 65(1): 40–8. PubMed Abstract | Publisher Full Text\n\nPonomarenko J, et al.: ElliPro: a new structure-based tool for the prediction of antibody epitopes. BMC Bioinformatics 2008; 9: 514. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDoytchinova IA, Flower DR: VaxiJen: a server for prediction of protective antigens, tumour antigens and subunit vaccines. BMC Bioinformatics. 2007; 8: 4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSaha S, Raghava GP: AlgPred: prediction of allergenic proteins and mapping of IgE epitopes. Nucleic Acids Res 2006; 34(Web Server): W202–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGupta S, et al.: In silico approach for predicting toxicity of peptides and proteins. PLoS One 2013; 8(9): e73957. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGarg VK, et al.: MFPPI - Multi FASTA ProtParam Interface. Bioinformation 2016; 12(2): 74–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKumar TA: CFSSP: Chou and Fasman secondary structure prediction server. Wide Spectrum 2013; 1(9): 15–9.\n\nYang J, et al.: Improved protein structure prediction using predicted interresidue orientations. Proc Natl Acad Sci U S A 2020; 117(3): 1496–503. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWillard L, et al.: VADAR: a web server for quantitative evaluation of protein structure quality. Nucleic Acids Res 2003; 31(13): 3316–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWiederstein M, Sippl MJ: ProSA-web: interactive web service for the recognition of errors in three-dimensional structures of proteins. Nucleic Acids Res 2007; 35(Web Server): W407–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHess B, et al.: GROMACS 4: Algorithms for Highly Efficient, Load-Balanced, and Scalable Molecular Simulation. J Chem Theory Comput 2008; 4(3): 435–47. PubMed Abstract | Publisher Full Text\n\nSchneidman-Duhovny D, et al.: PatchDock and SymmDock: servers for rigid and symmetric docking. Nucleic Acids Res 2005; 33(Web Server): W363–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMashiach E, et al.: FireDock: a web server for fast interaction refinement in molecular docking. Nucleic Acids Res 2008; 36(Web Server): W229–32. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrote A, et al.: JCat: a novel tool to adapt codon usage of a target gene to its potential expression host. Nucleic Acids Res 2005; 33(Web Server): W526–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDas S, et al.: Comparative Analysis of Predicted Gene Expression among Crenarchaeal Genomes. Genomics Inform 2017; 15(1): 38–47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoddard TD, et al.: Software extensions to UCSF chimera for interactive visualization of large molecular assemblies. Structure 2005; 13(3): 473–82. PubMed Abstract | Publisher Full Text\n\nBIOVIA DS: Discovery Studio Modeling Environment.2020."
}
|
[
{
"id": "77965",
"date": "09 Feb 2021",
"name": "Sivaraman Natarajan",
"expertise": [
"Reviewer Expertise Genomics",
"Single-cell Technology",
"and Cancer Biology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nKumar et al. in the manuscript attempted in-silico vaccine development against SARS-CoV-2 using publicly available SARS-CoV2-Sequence. They used three different crucial structural proteins (glycoproteins) from Helper T cell, Cytotoxic T cell, and B cell epitope. The epitope was predicted using publicly available tools. After predicting the epitope, they attempted an in-silico vaccine construction by applying various conditions including adding linked adjuvants to predict the efficiency of the vaccine. This immunoinformatics approach is very interesting given the SARS-CoV-2 evolution of new mutant strains, help to predict effective vaccine in-silico.\nAs the author mentioned in the manuscript it lacks any invitro experiment to confirm any of the predictions as proposed in the manuscript. However, the in-silico finding from the study will be immensely beneficial to the researcher who studies the SARS-Co-V2.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "77969",
"date": "17 Feb 2021",
"name": "Tirumala Bharani K. Settypalli",
"expertise": [
"Reviewer Expertise Molecular genetics",
"Immunology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe current research article by Kumar et al., utilized in silico Reverse vaccinology approach to develop multiepitope vaccine against SARS-CoV-2. Overall, the manuscript was well written and is a valuable contribution at this time of the pandemic.\nGiven the pandemic situation, the in silico approach is a reliable and fast option for quicker vaccine development. The authors have utilized various immunoinformatic online tools to predict T-cell and B-cell epitopes from three structural polyproteins and constructed vaccine sequence.\n\nAlthough the current study lacks the in vitro / in vivo studies, the immunoinformatic approach is helpful to reduce the cost and time for further experimental validation and can be a valuable contribution during the pandemic.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-44
|
https://f1000research.com/articles/10-43/v1
|
23 Jan 21
|
{
"type": "Research Article",
"title": "Does endoscopic thoracic sympathectomy improve the quality of life of patients with primary hyperhidrosis? A single center retrospective review",
"authors": [
"Ahmad Farouk Musa",
"Khit Kar Hui",
"Jeswant Dillon",
"Rusli Bin Nordin",
"Khit Kar Hui",
"Jeswant Dillon",
"Rusli Bin Nordin"
],
"abstract": "Background: Endoscopic thoracic sympathectomy (ETS) is renowned as an effective surgical treatment for primary hyperhidrosis (PHH) and believed to improve patients’ quality of life (QOL). This study aimed to evaluate the quality of life (QOL) of patients with PHH after ETS compared to before ETS, and to determine whether compensatory sweating (CS) affects QOL of patients. Methods: This is a single-centre retrospective review of patients who had undergone ETS at the National Heart Center [Institut Jantung Negara (IJN)], Malaysia. In total, 62 patients from January 2014 to December 2018 were recruited. Medical records were first reviewed for all relevant data, prior to making telephone interview to administer the questionnaire. A modified questionnaire with validated components was used to assess the patients’ QOL. Patient satisfaction, symptom resolution, recurrence and occurrence of CS were also asked during the interview. Data were analysed using IBM SPSS Statistics 25.0. Results: A total of 46 patients (response rate: 74.2%) completed the questionnaire, with 95.7% reporting improvement in the total QOL score (Mean difference = 113.54, SD=70.79, 95% CI = 95.52 – 134.57, p<0.001). There was remarkable symptom resolution for palmar HH as 97.8% reported dry hands, whereas majority of patients with palmar-plantar HH reported persistent sweating from feet HH. CS rate was 89.1%. In terms of severity of CS, 6 (14.6%) reported mild, 17 (41.5%) moderate, and 18 (43.9%) had severe CS. The severity of CS as well as the number of locations have a significant effect on the QOL reported (p=0.022 and p=0.008, respectively). Conclusion: ETS is an effective treatment for PHH in improving the QOL of patients, even long term. The occurrence of CS did not affect the QOL, but severity of CS and number of locations involved in CS affect the QOL of patients.",
"keywords": [
"primary hyperhidrosis (PHH)",
"endoscopic thoracic sympathectomy (ETS)",
"quality of life (QOL)",
"compensatory sweating (CS)"
],
"content": "Introduction\n\nPrimary hyperhidrosis (PHH) is a chronic and distressing condition not caused by a medical condition but thought to be caused by over-stimulation of cholinergic receptors on eccrine glands. An excessive sweating beyond what the body is used for homeostatic temperature regulation remains the main pathologic feature.1 Since eccrine glands are found commonly at the palms, axilla, face, and soles, these are the areas mainly affected by PHH.2 Physiologically, it is thought that the negative feedback loop is impaired in such patients, which also helps to explain how a physiologic response becomes pathologic.3\n\nDifferent prevalence rates for PHH have been cited around the globe ranging from 0.6% to 16.3%,4 and despite the myriad of treatment options, endoscopic thoracic sympathectomy (ETS) is still the most effective with a high rate of patient satisfaction despite the occurrence of compensatory sweating (CS) elsewhere in the body.5,6,7,8 Although the pathogenesis of CS remains unknown, it is believed that CS occurs due to an abnormal thermoregulatory response after sympathectomy due to altered feedback mechanism in the hypothalamus, whereby residual sweat glands increase their activity in an attempt to compensate for the loss of neural regulated sweat glands.9\n\nEver since the advent of ETS, CS has become the most feared complication because if severe, will significantly affect the quality of life (QOL) of patients.10 It is deemed the worst side effect of ETS, which impacts on daily activities.11,12,13 However, only a few studies have assessed CS and if its intensity affects the QOL of patients, but the results were inconsistent. Dias et al 14 proposed that CS potentially worsens postoperative QOL. Chang and colleagues’15 retrospective review reported that overall satisfaction of patients was inversely correlated with the severity of CS. On the other hand, one study concluded that postoperative QOL did not depend solely on the severity of CS, but also on the ability of patients to tolerate the situation.16\n\nIt is also worth noting that besides traditional outcome data, complication rate and morbidity, as we have presented in our previous research17, and QOL analyses were also used as benchmarks of the success of such surgical procedure. But it is also noted that some QOL questionnaires are lacking in quality, and in deriving substantive conclusions.18 Some researchers19,20 used generic QOL instruments such as the Short Form 36 (SF36) that would look at all aspects of the patients’ health status, but it lacks specificity. These questionnaires might not be able to detect disease-specific changes in the QOL of patients post-surgery21. Whereas disease-specific questionnaires might be able to measure the change in the QOL of patients suffering from a specific disease.22,23\n\nWe noted that, for assessing the QOL of patients after ETS, two of the most commonly used disease-specific questionnaires are by Keller et al24 that assesses the hyperhidrosis scale using 15 questions concerning daily life; and Milanez de Campos et al25 that has four different domains with 20 questions that cover sweating symptoms, intimacy, emotional response, and special circumstances.\n\nThe aims of this study are to evaluate the QOL of patients with PHH who had undergone ETS compared to their QOL pre-operatively, and to determine whether CS would affect the QOL of patients post-operatively.\n\n\nMethods\n\nThis study was conducted from January to July 2019 at the National Heart Institute (IJN) in Kuala Lumpur which was a five-year retrospective review of the QOL of patients who underwent bilateral endoscopic thoracic sympathectomy (BETS) for the treatment of PHH at the National Heart Institute of Malaysia [Institut Jantung Negara (IJN)] from January 2014 to December 2018. With the approval of the IJN Research Ethics Committee (IJNREC), all patients’ medical records were screened to ensure the completion of data. A complete medical record would include the proper and adequate documentation of patient’s demographics, preoperative consultation record, medical and surgical record, laboratory results and postoperative follow-up notes. All eligible patients were contacted through phone calls for the administration of the questionnaire, which comprises of preoperative and postoperative scoring.\n\nInclusion criteria: bilateral endoscopic thoracic sympathectomy cases; completed medical records; consented for an interview.\n\nExclusion criteria: missing data from the medical record; medical condition that resulted in secondary hyperhidrosis; redo endoscopic thoracic sympathectomy; failure to give consent for an interview\n\nThis constituted a total of 62 patients who all had moderate to severe PHH prior to surgery, involving more than one location.\n\nEthical approval was obtained from both the IJN Research Ethics Committee (IJNREC/207/2017) and Monash University Human Research Ethics Committee (MUHREC/9214). The study was also registered with the National Medical Research Register (NMRR-17-3133-39469). Verbal consent to participate in the study was obtained over the telephone. This method of consent was approved by the Research Ethics Committees, as set out in the study protocol submitted to them.\n\nAll research procedures were done in accordance with the ethical regulations set by the IJN Ethics Committee (IJNEC), Monash University Human Research Ethics Committee (MUHREC) and it abides with the Helsinki Declaration revised in 2013.\n\nAll patients were assessed for the severity of hyperhidrosis before surgery. Once the anaesthetic assessment is completed, elective surgery of BETS was scheduled. All BETS were performed by two different surgeons with the same level of experience and using similar techniques. All surgeries were performed in a standardized manner using the same equipment under general anaesthesia. None of the surgeries was converted into open surgery. In our institute, BETS was conducted in such a way that two incisions were made for the access of ports and thoracoscope. The first incision was made around the mid-axillary line over the fifth intercostal space. After deflating the lung of the operative side, the pleural space was entered by blunt dissection. Then, a second incision was made around the submammary fold under endoscopic guidance. For each side of the surgery, two ports of either 5mm or 10mm were used with a 30 degree thoracoscope. Once the ganglion was identified, the sympathetic chain would be interrupted using excision method. The same procedure was then repeated on the contralateral side at the same level of interruption in all cases. All specimens excised were sent for histopathology examination for the confirmation of ganglion excision. Patients were hospitalised for a total duration of three to four days to monitor for postoperative complications such as pneumothorax and wound management. All patients were subjected to chest X-rays which were reviewed by surgeons before discharge.\n\nFor this study, all surgical and inpatient notes were analysed and recorded including the total operative time, days of hospital stay, pain score, postoperative complications, and medications used.\n\nAll except one patient had at least one follow-up after surgery. About 63% of them attended follow-up twice and one of them had it thrice. During follow up, patients were assessed for symptom resolution, occurrence and severity of CS, pain, wound healing as well as other problems resulting from surgery.\n\nAll questionnaires were administered during the period of January until July 2019. The questionnaire used for this study was the validated Health Related Quality of Life (HRQOL) Questionnaire in PHH which is a modified questionnaire based on studies by Milanez de Campos et al25 and Keller and colleagues24, which was designed to specifically evaluate preoperative and postoperative improvement of QOL in patients with hyperhidrosis. After obtaining the demographic data from the case records, the patients were interviewed via telephone on the severity of their PHH and also any possible recurrence, and on CS (Extended data: Appendix 146). This is followed by the second set of questionnaires that assessed the impact of QOL in four major domains, namely the functional, social, personal and psychological domains (Extedned data: Appendix 246).\n\nThe questionnaire starts with a single question of ‘In general, how would you rate your quality of life before and after treatment?”, in which patients will be asked to rate from 0 to 10, with 0 being the worst and 10 being excellent. This was followed by a total of 29 questions with 17 questions on functional domain, 3 questions on social domain, 2 for personal domain and 7 questions on psychological domain. Within the functional domain, the questions were further divided into 9 questions that focus on palmar sweating, 5 questions on feet sweating, and 3 questions on axillary sweating. The final total score of QOL can therefore possibly range from 0 to 290. In this study, QOL was assessed in three major forms that are: a) general QOL, b) QOL score as per domain and c) total QOL score of all domains, as illustrated in Figure 1.\n\nFurtermore, patients were asked to rate their overall satisfaction level from 0 being the worst, 5 being neutral, to 10 being excellent, and the score was further categorised into very unsatisfied (0-1), unsatisfied (2-4), neutral (5), satisfied (6-8) and completely satisfied (9-10). For the purpose of this study, there were additional open-ended questions after each section of the domain to address all potential issues that might be missed from the questionnaire. A good mix between closed-ended and limited number of open ended-questions allows for a smooth flow of interview while adequately addressing patients’ concern. All questions were asked in a simple and clear manner without medical jargons to avoid ambiguity or misunderstanding. At the end of the interview, we also asked if patients regretted the decision to undergo surgery.\n\nAll complete patients’ medical records were reviewed by a single investigator to record relevant data. Data collection was performed by a single investigator throughout the study in order to reduce potential bias secondary to inconsistent administration of questionnnaire and data handling. Prior to interview, a separate reference list with only patients’ name and contact details was generated. This ensured the blinding of investigator from the level of interruption that patients had underwent, therefore reducing investigator bias. Before the interview, a template of conversation was created for a good conversation flow. This template included a standard introduction to the identity of investigator, purpose of the project, assurrance of patients’ confidentiality, and ended by obtaining patients’ informed consent. In addition, pilot tests were run with ten volunteers to gain feedback to improve the process of interview. These were patients from our previous cohort of study on ETS17 and no changes were made to the questionnaire after the pilot study.\n\nPrior to the interview, patients were asked to choose their language of preference, between Malay or English for better communication. This was done on a background of an investigator who is fluent in both languages. Good rapport was built by asking daily questions to make patients feel more at ease in expressing themselves. In order to reduce performance bias, all patients were reassured that there will be no coercion on their response and honest feedback is greatly appreciated for the benefit of scientific research. To reduce recall bias, all patients were informed at the beginning of the phone call that the interview would take an average of 30 minutes; hence, appointment will be made at another time if patients were occupied at that particular moment. Besides, for each of the questions asked, patients were given adequate time to recall without any prompts. This was strictly adhered to across all interviews to minimise recall bias. When asked to rate the QOL from a scale of 0 to 10 with regards to hyperhidrosis, a standard explanation was given to all patients that was: 0 represents the worst QOL you could imagine, while 10 represents excellent QOL, and 5 being a neutral response.\n\nData entry and analysis were performed using IBM® SPSS Statistics version 25.0. Descriptive statistics were reported as percentages for discrete variables, continuous data were reported as means with standard deviations (SDs) for parametric data and medians with interquartile ranges (IQRs) for non-parametric data. Internal consistency of questions was tested separately for each domains using Cronbach’s Alpha reliability test. Questions from each subset of functional domain and psychological domain demonstrate excellent internal consistency. Personal domain demonstrates poor consistency and hence may impose a limitation to our analysis (Table 1). Tested variables were first explored for its normality, outliers and linear relationship and those which met the assumptions will be analysed using parametric analysis whereas those that did not will be analysed using non-parametric analysis. Correlations between factors affecting the occurrence of CS were assessed using Spearman’s rho or Pearson’s correlation. Coefficient of more than 0.7 will be considered significant. Differences between categorical variables were analysed using Fisher’s exact test or chi-squared test, depending on the assumptions. Differences for independent continuous variables were analysed using independent t-test or Mann-Whitney test for parametric and non-parametric data, respectively. The differences in QOL score rated by patients before and after operation were compared using the paired t-tests for normally distributed data, or Wilcoxon signed-rank test for skewed data. As the improvement of QOL in certain domains was not normally distributed, all comparisons of QOL improvement in each domains were done using the Wilcoxon test. The value of significance was taken at p<0.05 for all the analysis.\n\n\nResults\n\nA total of 62 patients were included based on the criteria of complete medical records and having BETS done between January 2014 and December 2018. One patient was excluded for having undergone the surgery twice. All relevant medical records were screened thoroughly to ensure that there was no underlying medical condition that may be a confounder to secondary hyperhidrosis such as anxiety disorder, diabetes, or hyperthyroidism. The final diagnosis of PHH in each patient was confirmed through consultation record after the exclusion of possible secondary causes of hyperhidrosis. Thirteen patients were lost to contact, and two patients did not consent to the phone interview for personal reasons. Finally, 46 patients (response rate: 74.2%) consented to the telephone interview, and all of them completed the HRQOL questionnaire (Figure 2).\n\nThe studied population comprised of 30 (65.2%) and 16 female (34.8%) patients. In total, 36 of patients were ethnic Malays (84.8%), 6 were Chinese (13%) and 1 was Indian (2.2%), which appropriately represented the racial distribution in Malaysia. None of the patients had any medical conditions that might predispose them to secondary hyperhidrosis. Blood tests were done on all patients to ensure there were no cases of thyroid dysfunction, or diabetes. Upon first consultation, no patient reported any anxiety disorder which might worsen the severity of sweating.\n\nThe age of the study population ranged from 13 to 55 years old, with 84.8% of the study population being younger than 30 years old. The distribution was positively skewed with a median of 20.5 years old. The BMI of the study population is normally distributed with a mean of 23.02 kg/m2 (SD=4.04). When classified into categories, 1 (2.2%) patient was underweight, 25 (54.3%) were normal, 16 (34.8%) were overweight and 4 (8.7%) were obese (Table 2).\n\nLocation of hyperhidrosis\n\nOnly two patients had sweating from a single location (4.3%), as most of them had hyperhidrosis in two locations (73.9%), with the remaining reported PHH involving more than two locations (21.7%). Combined palmar-plantar hyperhidrosis was present in 73.9% of the study population, with the rest distributed in all different possible combinations, as summarised in Table 3. Broken down to each location and its individual reported frequencies, the palm was the most commonly reported location by all patients (100%), followed by the soles of the feet [43 (93.5%) patients].\n\nSeverity\n\nAll patients experienced moderate to severe sweating from the palms and soles, which prompted them to seek surgical treatment. Reportedly, 45.7% had moderate palmar hyperhidrosis while 54.3% had severe condition. None had mild sweating prior to surgery. Some described dribbling of sweat from hands and feet, which embarrassed and impaired their daily activities. The majority of patients described worsening of sweating with hot weather and relieved by staying in cool condition. A small proportion of patients described worsening of sweating by stress. In some cases, sweating was persistent even under cool conditions, leading to a very poor QOL before operation.\n\nLevel of sympathectomy\n\nIn our centre, majority of BETS were performed at a level of T2-T3 (78.3%), the frequency reported is summarised in Table 3. A T2 level has been well recognised to increase the risk of CS and other complications if interrupted26,27. Hence, we further categorise patients into T2-involved and T2-spared group to study the difference between both groups.\n\nOccurrence of compensatory sweating\n\nAlthough the definition of CS remains vague, it is essentially a condition of increased sweating at a previously normal location, to an extent of being noticed by patients. In our study, CS was reported in up to 41 (89.1%) of patients. We further assessed these patients for the location(s) involved and its severity. Data on the onset of CS first becoming noticeable was also recorded. When assessing the locations involved in CS, every patient was asked regarding the presence of CS at locations such as the face, axilla, trunk, abdomen or groin, and lower limbs. They were then grouped into patients who had CS from one location, two locations and more than two locations. In total, 38 (92.7%) patients reported having CS on the body, followed by lower limbs (48.8%). CS at the abdominal and groin region was reported by 10 (24.4%) patients, and axillary region by 4 (9.8%) patients. We noticed that 64.3% of patients who experienced CS first noticed the sweating within the first month of the surgery. Among these, 31.7% noticed it as early as during the first week of surgery. Less than a quarter (13.0%) noticed CS more than six-months after the surgery. Patients with CS were also asked to rate the severity of sweating as mild, moderate or severe according to the pre-set standard definition (Table 4). The majority of them had moderate CS (41.5%) or severe CS (43.9%). Only 14.6% of them described mild sweating and this information is summarised in Table 4.\n\nEvolution of compensatory sweating\n\nThe majority of patients reported no improvement in the intensity of CS. Only four (9.8%) patients reported complete resolution of CS over time, and three (7.3%) reported gradual improvement with time. In those who claimed that CS had resolved, one described having CS on their trunk for around seven-months to a year, before it progressively resolved. The other two patients descibed having CS for more than a year which slowly improved and then resolved. There was a patient who described having CS at multiple locations including the trunk, abdomen and groin, and thigh but CS from the thigh had resolved after it occurred for a week, even though CS from other locations remained.\n\nPatient satisfaction\n\nIn general, there was great satisfaction reported by patients with a median score of 7 (IQR=4). There was an ascending trend of frequency from the worst to excellent as shown below (Table 5). In this study, we took the value of 5 as neutral as explained to all patients before the administration of questionnaire.\n\nCompensatiry sweating and patient satisfaction\n\nWe did not find any significant difference in satisfaction between the group of patients with CS and without CS (p=0.219) (Table 6).\n\nData is not normally distributed.\n\na Mann-Whitney test\n\nWe noticed that comparing patients with mild to moderate (non-severe) CS with those who had severe CS, the satisfaction reported was highly and significantly lower in the latter group (p=0.001) (Table 7).\n\nData is not normally distributed.\n\na Mann-Whitney test\n\n* significant difference (p=0.001)\n\nGeneral QOL\n\nIn our study, 37 (80.4%) patients reported improvement in the general QOL after surgery, whereas 5 (10.9%) reported no changes with general QOL and 4 (8.7%) reported worse QOL after surgery. All four patients who reported worse general QOL after surgery had severe CS involving more than one location. It was demonstrated in this study that the improvement in the general QOL of patients after surgery was highly statistically significant (p<0.001) (Table 8).\n\nData is not normally distributed.\n\na Wilcoxon test. IQR: Inter Quartile Range;\n\n* significant difference (p < 0.001)\n\nQOL improvement by domain\n\nFor comparison of QOL improvement in each of the four domains, the mean changes in score were adjusted to give an average improvement of score per question. It can be observed in Figure 3 below that all domains have demonstrated an increment in the median score for postoperative QOL, although it was not prominent in the axilla (p=0.087). Besides, the functional domain of hands had shown the greatest improvement as compared to other domains with average increment of 5.86 per question (p<0.001). This was followed by the social domain, which also showed substantial improvement with an average increment of 5.12 per question (p<0.001).\n\nSummary of improvement in QOL\n\nTable 9 summarizes the preoperative and postoperative median score for each question with its respective p value. It was noted that all domains had significant improvement in QOL score except for the axillary domain.\n\na Questionnaire in Extended data: Appendix 2, adapted from Milanez de Campos et al 25 and Keller et al24;\n\nb Wilcoxon-test. Data is not normally distributed;\n\n* significant difference (postoperative-preoperative) (p <0.001).\n\nTable 10 summarizes the median score for each domain and its respective average increment in score per question and p values. This is followed by Table 11 that describes the change in QOL in each domain.\n\na Wilcoxon-signed rank test. Data is not normally distributed;\n\n** significant difference (p<0.001)\n\nTotal sum of QOL\n\nWe also summed up the total score from all domains to compare the changes before and after surgery. Of all patients, only two (4.3%) had a reduction in the total QOL score, whereby the rest of them (95.7%) reported improvement in total QOL. By comparing the differences between the preoperative and postoperative total QOL score using paired t-test, there was a statistically significant difference (p<0.001), with a mean difference of 113.54 (SD=70.79, 95% CI=95.52 - 134.57) (Table 12).\n\na Paired t-test. Data is normally distributed. Assumptions were fulfilled;\n\n** significant at p < 0.0001\n\nFactors affecting QOL\n\nTo study the factors associated with QOL after surgery, selected variables were initially tested for their correlation with general and total QOL score using Spearman’s and Pearson’s correlation. It was noted that the severity of CS and number of location(s) involved in CS were significantly associated with both general and total QOL (Table 13).\n\na Spearman’s rho correlation.\n\nb Pearson’s correlation. Assumptions were fulfilled. r = correlation coefficient;\n\n* significant (p<0.05);\n\n** significant (p=0.002)\n\n*** significant (p<0.001); CS, compenstory sweating.\n\nIt is interesting to note that when comparing the postoperative general QOL and total QOL between patients with CS and without CS, no significant difference was noted (p=0.303 and p=0.167 respectively) (Table 14). However, when we compared the general QOL reported by patients with non-severe CS and those with severe CS, there was a significantly lower general QOL reported by the latter (p<0.001) (Table 15). We also performed an analysis between the severity of CS with the total QOL score. By running an independent t-test, the results were consistent with the above analysis in which the mean total QOL score was noted to be significantly lower in the group with severe CS (Mean=217.17 vs 184.22, p=0.008) (Table 15).\n\na Mann-Whitney test. Data is not normally distributed.;\n\nb Independent t-test.\n\nData is normally distributed. All assumptions were fulfilled.\n\na Mann-Whitney test. Data is not normally distributed;\n\nb Independent T-test. Data is normally distributed. All assumptions were fulfilled;\n\n* significant (p=0.008);\n\n** significant (p=0.001)\n\nPost-hoc Bonferroni was performed to improve the accuracy of such comparison. It was found that there was still a significant association between two factors (p=0.022) with a significant difference in mean total QOL reported between patients with moderate and severe CS (p=0.019) (Table 16).\n\na One-way ANOVA test was applied. Assumptions were fulfilled;\n\n* significant at p<0.05;\n\nb Post-hoc analysis: Bonferroni test was applied. Significant difference was found between moderate and severe compensatory sweating (p=0.019).\n\nAnother significant finding is with regard to the number of locations of CS. By using Kruskal-Wallis test, we noticed a significantly lower general QOL reported in patients with CS involving more than two locations, as compared with those involving one location or two locations (p=0.008). Post-hoc Dunn test had demonstrated, in particular, a significant difference in median score between patients with one location and more than two locations (p=0.024) (Table 17). As illustrated in Figure 4, we can see an obvious downward trend of the median general QOL score from ‘one location’ box to ‘two locations’ and to ‘more than two locations’.\n\na Kruskal-Wallis test was applied. Assumptions were fulfilled;\n\n* significant at p < 0.05;\n\nb Post-hoc analysis: Dunn test was applied. Significant difference was found between patients with only 1 location and more than 2 locations (p=0.024).\n\nSimilarly, we also ran a test to study the association between number of CS location(s) and total QOL. By using individual t-test, patients with more than two locations of CS reported significantly lower total QOL score as compared to those with one location of CS, with a mean difference of 39.48 (95% CI=4.44 - 74.52, p=0.029) (Table 18).\n\na Independent t-test. Data is normally distributed. All assumptions were fulfilled;\n\n* significant at p <0.05\n\n\nDiscussion\n\nOur study has demonstrated a significant improvement in the general QOL of patients postoperatively, which is in keeping with most of the literature that supports the effectiveness of ETS in improving the general QOL of patients with PHH.12,13,14,22,28,29,30,31,32,33 By summing up the total QOL score from all domains, we noticed 44 (95.7%) patients had an improvement in the total QOL score with a mean improvement of score of 113.54 (SD=70.79, 95% CI=95.52 - 134.57), supporting the overall effectiveness of ETS in improving QOL (p<0.001) (Table 12).\n\nAmong all domains that were asked in the questionnaire, functional domain of hands demonstrated the most profound improvement in QOL after the operation, as all patients (100%) reported an improvement in the QOL score after surgery, with a highest average improvement per question and mean score improvement of 53.48 (95% CI=46.91 - 60.04, p<0.001) (Tables 10 and 11). This further supports the effectiveness of ETS in resolving sweating from palms. Apart from the functional domain, all other domains had also shown significant improvement in the QOL score when compared to the preoperative state. However, for the functional domain of axilla, the improvement was not statistically significant (p=0.087) (Table 10). This can be explained by the majority occurrence of CS on the trunk which also had extended effect to the axillary region, leading to less QOL score given in this domain.\n\nOverall satisfaction score was reported to be a median of 7 (IQR=4), with the most common cause of dissatisfaction being the occurrence of CS. As we compared satisfaction between non-severe CS and severe CS, there was significantly lower overall satisfaction in patients who had severe CS (Median=9.00 vs 5.00, p=0.001) (Table 13). However, the occurrence of CS alone was not found to have a profound impact on overall satisfaction (Table 14). This is because satisfaction also comprises other elements other than the outcome of surgery such as consultation, hospitalisation services and other services provided.\n\nWhen we examined the relationship between CS and QOL, we noticed that the severity of CS had a significant impact on the general and total QOL score. Comparing patients with non-severe CS, patients with severe CS reported much lower general and total QOL score (p=0.001 and p=0.008 respectively) (Table 15). For greater accuracy, further analysis was run using one-way ANOVA. It was noted that there was still a significant association between the severity of CS and total QOL score after the operation with p=0.022. Patients who reported moderate CS had higher total QOL score than those with severe CS. (Mean difference=36.54, 95% CI=4.83 - 68.26, p=0.022) (Table 16).\n\nFurthermore, patients with a lower number of location(s) involved in CS reported better general and total QOL score. By using the t-test, patients who had CS in a single location reported a significantly higher total QOL as compared to patients with more than two locations of CS, with a mean difference of 39.48 (95% CI=4.44 - 74.52, p=0.029) (Table 18). In short, even though we could not comment on the association between the occurrence of CS and QOL, we found significant association between both severity of CS and number of location(s) of CS with QOL score.\n\nWe also believe that there is a common assumption that CS will disappear or improve with time. However, Herbst et al34 did not agree with this assertion as from their study for over fourteen years, the authors concluded that up to 67% of patients still reported permanent CS without improvement. In our study, the majority of patients (82.9%) with CS had persistent CS without improvement. Only 9.8% of patients reported resolved CS over time, with gradual improvement of CS reported in 7.3% of patients. This is of course a higher rate of CS persistence if compared to the study by Herbst et al34 but of a shorter duration. The inconsistency of our data with the current literatures available with the lack of supporting literatures necessitates more studies to validate such assumptions.\n\nOne issue that must be discussed here is the relationship between CS and the extent or level of sympathectomy, which has been the subject of intense debate and controversies among surgeons. Although the optimal level of interruption and method of interruption remains debatable, it was suggested by The Society of Thoracic Surgeon expert consensus35 for the surgical treatment of hyperhidrosis that the optimal operation for palmar hyperhidrosis is a T3 interruption with cauterizing or clipping as it has been shown to result in close to 100% symptom resolution. However, it is also reasonable to approach for T4 interruption. The expert consensus35 also mentioned that interruption at T3 as compared with T4, poses a slightly higher risk of CS but provided better symptom resolution. Even though it is impossible to predict the risk of CS, avoiding interruption at T2 and limiting the extent of interruption are currently the best method to reduce the risk of CS.\n\nWe would summarize that studies with high reliability and strength have shown that higher interruption resulted in greater risk of complications36. Our centre, which carried out most surgeries involving T2 level, did not report any case of Horner’s syndrome and few cases of gustatory sweating in five (10.9%) patients. Amongst the 8.7% of patients who had T2 spared, all of them experienced CS as well. However, this interpretation is reserved with a marked limitation noted in our study as our centre performed surgical interruption at the level of T2-T3 in the majority of cases (78.3%) with a great portion of patients falling into the T2-involved group (91.3%). Besides, we could not yield a similar finding to those found in our previous study17 as our sample size is relatively small compared to previous studies. Furthermore, as this study is a continuation of the previous study17, different study populations used in the different studies may explain variations in the results. However, we believe that more surgeries sparing T2 should be carried out in order for further analysis on such association to be done.\n\nAfter an extensive literature search, we found that most studies have investigated CS and QOL as two separate entities. Among the few literatures that associated CS with QOL, the findings were inconsistent with one another. And in consonant to what was claimed by Hajjar et al28 that there was no relationship between CS and postoperative QOL, our study did not find any significant relationship between the occurrence of CS itself and postoperative QOL as well (Table 14). This finding contradicted Dias et al14 who proposed that CS potentially worsens postoperative QOL. As we explored the effect of CS on QOL on other aspects, we found a significant association between QOL and severity of CS where poorer QOL was reported in patients who experienced severe CS (Tables 15 and 16). It was observed in certain cases that some patients had severe CS but reported a higher QOL than patients who had milder form of CS. This was because patients’ ability to tolerate CS is subjective and will affect the QOL that they perceived. This finding was entirely consistent with findings reportedly done by Leiderman et al12 and Wolosker et al16.\n\nTo date, we did not find any literature that has clearly demonstrated a relationship between the number of CS location(s) with QOL, not to mention that this factor might not be known to researchers. However, as described previously, we found a significantly lower postoperative QOL reported by patients in our study who had more than two locations involved in CS as compared to those with one location (Tables 17 and 18, and Figure 4). Therefore, we suggest the number of location(s) involved in CS to be considered as a confounder to postoperative QOL for future research. The clinical implication of this study is that with the high incidence of CS reported in this study and our previous study,17 which can be associated with the involvement of T2 resection, it may be proposed to our institute regarding reducing ETS surgery that involves T2 resection. This suggestion is backed by the literature, which claims that symptom resolution was not significantly different between T2-involved and T2-spared group, but incidence of CS was significantly higher in the T2-involved group26,36. In addition, the expert consensus35 of The Society of Thoracic Surgeons also supported the idea that ETS should avoid involving the T2 level as it may result in more complications and significantly higher risk of CS.\n\nGiven the relative rarity of this surgery with less than 15 cases performed per year at our institute, we believe a longer duration of such retrospective study is more effective in capturing a higher number of patients. However, there were a few limitations in this study. First of all, the study design requires inclusion of telephonic follow-up. As with any research that involves telephone follow-up, reaching all participants was a difficult and tedious process because of changed or disconnected numbers and hence a few patients were lost. In addition, since the data was pre-recorded, there may be inaccuracy and inadequacy with regards to the history and examination notes. Furthermore, as data was collected from a single centre, this may potentially introduce selection bias. In addition, as the QOL questionnaire was applied postoperatively, there was also a risk of recall bias. We have taken a few steps, as mentioned in Methods, that were carried out and strictly adhered in order to reduce recall bias. Additionally, Cronbach alpha reliability test was run and showed poor internal consistency for the questions in the personal domain, which might affect the analysis that involve total QOL score. Finally, as this study also looked into the occurrence of CS and factors affecting it, a small sample size in the group of patients without CS has limited our analysis related to the occurrence of CS.\n\n\nConclusion\n\nThis study has managed to answer the pre-set research questions, on the QOL of patients after ETS compared with that before, as well as the secondary question of whether or not CS affects QOL of patients. Our study has proven the effectiveness of ETS in providing sustained improvement of QOL after the surgery, with remarkable resolution of sweating from the palms. Furthermore, overall satisfaction was reportedly high and not affected by the presence of CS or other long-term complications. These findings on the efficacy of ETS is largely in coherence with many other studies37,38,39,40,41,42,43,44,45 that recognise ETS as an effective treatment for PHH with a good safety profile. With regards to CS, our institute reported a relatively high prevalence (89.1%) of the occurrence of CS, which can be attributed to two reasons. Firstly, the majority of our surgeries were performed involving the level of T2, which has been highly regarded by many researchers to increase the risk of CS and other complications. Secondly, as this rate is comparable to a study28 done in a desert climate country of Saudi Arabia which reported prevalence of 92%, hence leading us to think that a hot climate, like in our own country, could be a confounder to CS.\n\nTo answer the second research question, we found no significant association between the occurrence of CS and QOL of patients. However, we noticed a significant relationship between severity of CS and QOL. In addition, we discovered a potential relationship between the number of location(s) involved in CS and QOL. It is worth noting that such association has never been clearly described or studied in other research before, to the best of our knowledge. We hope that this study will provide a research background for future researchers to confirm the number of CS locations as a confounder to postoperative QOL. Knowing the inadequacy of our study due to the small sample size, we believe an extension of current study with inclusion of more patients with spared T2 will be able to provide more convincing answers to the many questions that were limited due to its sample size.\n\n\nData availability\n\nHarvard Dataverse. Replication Data for: Does endoscopic thoracic sympathectomy improve the quality of life of patients with primary hyperhidrosis? A single center retrospective review, https://doi.org/10.7910/DVN/REQ7Y746\n\nThis project contains the following underlying data:\n\n• Set 1: Raw Data (.tab)\n\n• Set 2: Output Data (PDF)\n\nHarvard Dataverse. Replication Data for: Does endoscopic thoracic sympathectomy improve the quality of life of patients with primary hyperhidrosis? A single center retrospective review, https://doi.org/10.7910/DVN/REQ7Y746\n\nThis project contains the following extended data:\n\n• Set 3: Appendix 1: Questionnaire - Set 1 (PDF)\n\n• Set 4: Appendix 2: Questionnaire - Set 2 (PDF)\n\nData are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).",
"appendix": "Acknowledgements\n\nThe authors would like to extend our gratitude to the Senior Manager of the Clinical Research Department, National Heart Institute, Mr Mohd Faizal Ramli, and all the Clinical research staff, for their assistance in making this study possible.\n\n\nReferences\n\nSammons JE, Khachemoune A: Axillary hyperhidrosis: a focused review. J Dermatolog Treat. 2017; 28(7): 582–590. PubMed Abstract | Publisher Full Text\n\nMenzinger S, Quenan S: Evaluation and management of hyperhidrosis. Rev Med Suisse. 2017; 13(556): 710–714. PubMed Abstract\n\nRomero FR, Haddad GR, Miot HA, et al.: Palmar hyperhidrosis: clinical, pathophysiological, diagnostic and therapeutic aspects. An Bras Dermatol. 2016; 91(6): 716–725. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEstevan F, Wolosker MB, Wolosker N, et al.: Epidemiologic analysis of prevalence of the hyperhidrosis. An Bras Dermatol. 2017; 92(5): 630–634. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDumont P, Denoyer A, Robin P: Long-term results of thoracoscopic sympathectomy for hyperhidrosis. Ann Thorac Surg 2004; 78: 1801–1807. PubMed Abstract | Publisher Full Text\n\nGossot D, Galetta D, Pascal A, et al.: Long-term results of endoscopic thoracic sympathectomy for upper limb hyperhidrosis. Ann Thorac Surg 2003; 75: 1075–1079. PubMed Abstract | Publisher Full Text\n\nChang YT, Li HP, Lee JY, et al.: Treatment of palmar hyperhidrosis: T4 Level compared with T3 and T2. Ann Surg 2007; 246: 330–336. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJeganathan R, Jordan S, Jones M, et al.: Bilateral thoracoscopic sympathectomy: results and long-term follow-up. Interact Cardiovasc Thorac Surg 2008; 7: 67–70. PubMed Abstract | Publisher Full Text\n\nChou SH, Kao EL, Lin CC, et al.: The importance of classification in sympathetic surgery and a proposed mechanism for compensatory hyperhidrosis: experience with 464 cases. Surg Endos. 2006; 20(11): 1749–1753. PubMed Abstract | Publisher Full Text\n\nCai S, Huang S, An J, et al.: Effect of lowering or restricting sympathectomy levels on compensatory sweating. Clin Auton Res. 2014; 24(3): 143–9. PubMed Abstract | Publisher Full Text\n\nMilanez de Campos JR, Kauffman P, Faustino CB, et al.: Upper extremity sympathectomy. J Vis Surg. 2018; 4: 180. Publisher Full Text\n\nLeiderman DBD, Milanez de Campos JR, Kauffman P, et al.: The relation between age and outcomes of thoracic sympathectomy for hyperhidrosis: The older the better. J Thorac Cardiovasc Surg. 2018; 156(4): 1748–56. PubMed Abstract | Publisher Full Text\n\nFukuda JM, Varella AYM, Teivelis MP, et al.: Video-assisted thoracoscopic sympathectomy for facial hyperhidrosis: The influence of the main site of complaint. Ann Vasc Surg. 2018; 46: 337–44. PubMed Abstract | Publisher Full Text\n\nDias LI, Miranda EC, Toro IF, et al.: Relationship between anxiety, depression and quality of life with the intensity of reflex sweating after thoracoscopic sympathectomy for treatment of primary hyperhidrosis. Rev Col Bras Cir. 2016; 43(5): 354–9. PubMed Abstract | Publisher Full Text\n\nChang YT, Li HP, Lee JY, et al.: Treatment of palmar hyperhidrosis: T(4) level compared with T(3) and T(2). Ann Surg. 2007; 246(2): 330–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWolosker N, Yazbek G, de Campos JR, et al.: Quality of life before surgery is a predictive factor for satisfaction among patients undergoing sympathectomy to treat hyperhidrosis. J Vasc Surg. 2010; 51(5): 1190–1194. PubMed Abstract | Publisher Full Text\n\nMusa A, Gandhi V, Dillon J, et al.: A retrospective review on minimally invasive technique via endoscopic thoracic sympathectomy (ETS) in the treatment of severe primary hyperhidrosis: Experiences from the National Heart Institute, Malaysia. F1000Res. 2018; 7(670). PubMed Abstract | Publisher Full Text | Free Full Text\n\nVelanoch V: The quality of quality of life studies in general surgical journals. J Am Coll Surg 2001; 193: 288–296. PubMed Abstract | Publisher Full Text\n\nLau WT, Lee JD, Dang CR, et al.: Improvement in quality of life after bilateral transthoracic endoscopic sympathectomy for palmar hyperhydrosis. Hawaii Med J 2001; 60: 126–137. PubMed Abstract\n\nYoung O, Neary P, Keaveny TV, et al.: Evaluation of the impact of transthoracic endoscopic sympathectomy on patients with palmar hyperhidrosis. Eur J Vasc Endovasc Surg 2003; 26: 673–676. PubMed Abstract | Publisher Full Text\n\nPanhofer P, Zacheri J, Jakesz G, et al.: Improved quality of life after sympathetic block of upper limb hyperhidrosis. Br J Surg 2006; 93: 582–586. PubMed Abstract | Publisher Full Text\n\nMilanez de Campos JR, Kauffman P, de Campos Werebe E, et al.: Quality of life, before and after thoracic sympathectomy: report on 378 operated patients. Ann Thorac Surg 2003; 76: 886–891. PubMed Abstract | Publisher Full Text\n\nNeumayer C, Zacherl J, Holak G, et al.: Limited endoscopic thoracic sympathetic block for hyperhidrosis of the upper limb: reduction of compensatory sweating by clipping T4. Surg Endosc 2004; 18: 152–156. PubMed Abstract | Publisher Full Text\n\nKeller SM, Bello R, Vibert V, et al.: Diagnosis of palmar hyperhidrosis via questionnaire without physical examination. Clin Auton Res 2009; 19: 175–181. PubMed Abstract | Publisher Full Text\n\nMilanez de Campos JR, Kauffman P, de Campos Werebe E, et al.: Questionnaire of quality of life in patients with primary hyperhidrosis. J Pneumologia 2003; 29(4): 178–191. Publisher Full Text\n\nSang HW, Li GL, Xiong P, et al.: Optimal targeting of sympathetic chain levels for treatment of palmar hyperhidrosis: an updated systematic review. Surg Endosc. 2017; 31(11): 4357–69. PubMed Abstract | Publisher Full Text\n\nCheng A, Johnsen H, Chang MY: Patient Satisfaction after Thoracoscopic Sympathectomy for Palmar Hyperhidrosis: Do Method and Level Matter?. Perm J. 2015; 19(4): 29–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHajjar WM, Al-Nassar SA, Al-Sharif HM, et al.: The quality of life and satisfaction rate of patients with upper limb hyperhydrosis before and after bilateral thoracic sympathectomy. Saudi J Anaesth 2019; 13: 16–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVannucci F, Araujo JA: Thoracic sympathectomy for hyperhidrosis: from surgical indications to clinical results. J Thorac Dis. 2017; 9(Suppl 3): S178–S92. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLembranca L, Wolosker N, Milanez de Campos JR, et al.: Videothoracoscopic Sympathectomy Results after Oxybutynin Chloride Treatment Failure. Ann Vasc Surg. 2017; 43: 283–7. PubMed Abstract | Publisher Full Text\n\nGuo JG, Fei Y, Huang B, et al.: CT-guided thoracic sympathetic blockade for palmar hyperhidrosis: Immediate results and postoperative quality of life. J Clin Neurosci. 2016; 34: 89–93. PubMed Abstract | Publisher Full Text\n\nBaroncello JB, Baroncello LR.Z, Schneider EG.F, et al.: Evaluation of quality of life before and after videothoracoscopic simpathectomy for primary hyperhidrosis. Rev Col Bras Cir. 2014; 41(5): 325–30. PubMed Abstract | Publisher Full Text\n\nWolosker N, Milanez de Campos JR, Kauffman P, et al.: Evaluation of quality of life over time among 453 patients with hyperhidrosis submitted to endoscopic thoracic sympathectomy. J Vasc Surg. 2012; 55(1): 154–6. PubMed Abstract | Publisher Full Text\n\nHerbst F, Plas EG, Függer R, et al.: Endoscopic thoracic sympathectomy for primary hyperhidrosis of the upper limbs. A critical analysis and long-term results of 480 operations. Ann Surg 1994; 220(1): 86–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCerfolio RJ, Milanez de Campos JR, Bryant AS, et al.: The Society of Thoracic Surgeons expert consensus for the surgical treatment of hyperhidrosis. Ann Thorac Surg. 2011; 91(5): 1642–8. PubMed Abstract | Publisher Full Text\n\nZhang W, Wei Y, Jiang H, et al.: T3 versus T4 thoracoscopic sympathectomy for palmar hyperhidrosis: a meta-analysis and systematic review. J Surg Res. 2017; 218: 124–31. PubMed Abstract | Publisher Full Text\n\nKumagai K, Kawase H, Kawanishi M: Health-related quality of life after thoracoscopic sympathectomy for palmar hyperhidrosis. Ann Thorac Surg. 2005; 80: 461–6. PubMed Abstract | Publisher Full Text\n\nSugimura H, Spratt EH, Compeau CG, et al.: Thoracoscopic sympathetic clipping for hyperhidrosis: Long-term results and reversibility. J Thorac Cardiovasc Surg. 2009; 137: 1370–6. PubMed Abstract | Publisher Full Text\n\nPrasad A, Ali M, Kaul S: Endoscopic thoracic sympathectomy for primary palmar hyperidrosis. Surg Endosc. 2010; 24: 1952–7. PubMed Abstract | Publisher Full Text\n\nOncel M, Sadi Sunam G, Erdem E, et al.: Bilateral thoracoscopic sympathectomy for primary hyperhydrosis: A review of 335 cases. Cardiovasc J Afr. 2013; 24: 137–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVorkamp T, Foo FJ, Khan S, et al.: Hyperhidrosis: Evolving concepts and a comprehensive review. Surgeon. 2010; 8: 287–92. PubMed Abstract | Publisher Full Text\n\nDewey TM, Herbert MA, Hill SL, et al.: One-year follow-up after thoracoscopic sympathectomy for hyperhidrosis: Outcomes and consequences. Ann Thorac Surg. 2006; 81: 1227–32. PubMed Abstract | Publisher Full Text\n\nVanderhelst E, De Keukeleire T, Verbanck S, et al.: Quality of life and patient satisfaction after video-assisted thoracic sympathicolysis for essential hyperhidrosis: A follow-up of 138 patients. J Laparoendosc Adv Surg Tech A. 2011; 21: 905–9. PubMed Abstract | Publisher Full Text\n\nYanagihara TK, Ibrahimiye A, Harris C, et al.: Analysis of clamping versus cutting of T3 sympathetic nerve for severe palmar hyperhidrosis. J Thorac Cardiovasc Surg. 2010; 140: 984–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHamm H: Impact of hyperhidrosis on quality of life and its assessment. Dermatol Clin. 2014; 32: 467–76. PubMed Abstract | Publisher Full Text\n\nMusa AF: Replication Data for: Does endoscopic thoracic sympathectomy improve the quality of life of patients with primary hyperhidrosis? A single center retrospective review. In: Harvard Dataverse, V2, UNF:6:DCEVdfDz4RdD04Rttb5T1Q== [fileUNF] 2020. Publisher Full Text"
}
|
[
{
"id": "85018",
"date": "10 Jun 2021",
"name": "Nelson Wolosker",
"expertise": [
"Reviewer Expertise Hyperhidrosis"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIntroduction\nMusa A et al. present their experience with patients operated for hyperhidrosis due to bilateral sympathectomy. Initially, they offer a current review of the characteristics of the disease, its epidemiology, therapeutic alternatives, and forms of evaluation of surgical treatment are presented. The forms of evaluation of the results are exposed, presenting the analysis of quality of life as one of the main methods. The authors analyze the existing questionnaires. This study is based on an analysis of quality of life created by a group of researchers that contemplates the positive characteristics of several previous questionnaires. A retrospective study of 46 patients (initially 62) conducted in Malaysia comprising five years of bilateral sympathectomy between 2014 and 2018.\nThe failure in the introduction is the lack of a paragraph dedicated to the presentation of the questionnaire created for this study.\nMethods Clear. A retrospective study of 62 patients conducted in Malaysia comprising 5 years of BETS, between 2014 and 2018. The surgical techniques, methodology, and statistical analysis are straightforward.\nSurgical techniques Well described.\nQuiz Well presented.\nStatistical analysis Proper.\nResults Number of cases included. The authors explained in detail why the final number of patients studied has decreased to 46. The demographics of the patients were presented in detail, similar to the literature. They were patients with severe hyperhidrosis who presented the resolution for palmar HH in 97.8% of the patients, and the majority of patients with palmar-plantar HH reported persistent sweating from feet HH. Results consistent with the literature. Compensatory Sudoresis occurred in 89.1% of patients to some degree. Despite this, 95.7% reported improvement in the total QOL.\nDemographics Well presented.\nLocation of hyperhidrosis Presented in great detail.\nSeverity Clear.\nLevel of Sympathectomy Clearly presented.\nCompensatory Hyperidrosis Well presented.\nQOL Well presented and interesting.\nDiscussion Proper. The discussion was very appropriate, and once again, it is observed that ETS is a suitable method for the treatment of palmar hyperhidrosis, despite compensatory hyperhidrosis.\nConclusions Adequate.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "96393",
"date": "11 Oct 2021",
"name": "Stamatis Gregoriou",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors report improvement of QOL in patients with multiple site focal hyperhidrosis undergoing sympathectomy and presenting with compensatory sweating. Data on patients with multiple site focal hyperhidrosis are scarce in the literature so the data presented are of merit. The paper could improve with the following information added:\nAccording to all guidelines surgical procedures are 3rd line treatment in severe focal hyperhidrosis when other therapies have failed or are contraindicated. In the paragraph describing severity, data of severity in a validated scale such as HDSS would be welcome. Furthermore, data on previous therapies used by the patients should be presented as patients having tried most of the 1st or 2nd line treatments unsuccessfully would be more prone to have no compromise in their QOL even with CS.\n\nAre there any data on sites of hyperhidrosis not treated with ETS (such as the craniofacial region) developing exacerbation of hyperhidrosis after treatment? This is not included in the strict CS definition but has been mentioned in literature after other therapies (e.g. in my article: Gregoriou S. et al. Effects of botulinum toxin-a therapy for palmar hyperhidrosis in plantar sweat production. Dermatol Surg. 2010 Apr;36(4):496-8.1). The article I mention evaluated worsening of hyperhidrosis in sites other than the ones treated. Since compensatory hyperhidrosis is defined as novel appearance of hyperhidrosis in sites that were previously sweating normally, it would be useful to know (if such information is available) if in the patients reporting facial hyperhidrosis (that could not improve after ETS) hyperhidrosis exacerbated after ETS.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-43
|
https://f1000research.com/articles/10-42/v1
|
22 Jan 21
|
{
"type": "Study Protocol",
"title": "IMPORTANCE trial: a provisional study-design of a single-center, phase II, double-blinded, placebo-controlled, randomized, 4-week study to compare the efficacy and safety of intranasal esketamine in chronic opioid refractory pain",
"authors": [
"Mauricio Fernandes",
"Magdalena Schelotto",
"Philipp Maximilian Doldi",
"Giovanna Milani",
"Abul Andrés Ariza Manzano",
"Doriam Perera Valdivia",
"Alexandra Marie Winter Matos",
"Yasmin Hamdy Abdelrahim",
"Shaza Ahmed Hamad Bek",
"Benito K. Benitez",
"Vanessa Luiza Romanelli Tavares",
"Abdulrahim M. Basendwah",
"Luis Henrique Albuquerque Sousa",
"Naiara Faria Xavier",
"Tania Zertuche Maldonado",
"Sarah Toyomi de Oliveira",
"Melisa Chaker",
"Michelle Menon Miyake",
"Elif Uygur Kucukseymen",
"Kinza Waqar",
"Ola M.J. Alkhozondar",
"Ricardo Bernardo da Silva",
"Guilhermo Droppelmann",
"Antonio Vaz de Macedo",
"Rui Nakamura",
"Felipe Fregni",
"Mauricio Fernandes",
"Magdalena Schelotto",
"Giovanna Milani",
"Abul Andrés Ariza Manzano",
"Doriam Perera Valdivia",
"Alexandra Marie Winter Matos",
"Yasmin Hamdy Abdelrahim",
"Shaza Ahmed Hamad Bek",
"Benito K. Benitez",
"Vanessa Luiza Romanelli Tavares",
"Abdulrahim M. Basendwah",
"Luis Henrique Albuquerque Sousa",
"Naiara Faria Xavier",
"Tania Zertuche Maldonado",
"Sarah Toyomi de Oliveira",
"Melisa Chaker",
"Michelle Menon Miyake",
"Elif Uygur Kucukseymen",
"Kinza Waqar",
"Ola M.J. Alkhozondar",
"Ricardo Bernardo da Silva",
"Guilhermo Droppelmann",
"Antonio Vaz de Macedo",
"Rui Nakamura",
"Felipe Fregni"
],
"abstract": "Background: Cancer is the second leading cause of death globally. Up to 86% of advanced cancer patients experience significant pain, while 10-20% live in chronic pain. Besides, increasing prescription of opioids resulted in 33,000 deaths in the US in 2015. Both reduce patients’ functional status and quality of life. While cancer survival rates are increasing, therapeutic options for chronic opioid refractory pain are still limited. Esketamine is the s-enantiomer of ketamine, with superior analgesic effect and less psychotomimetic side effects. Intranasal esketamine was approved by the FDA for treatment-resistant depression. However, its use in chronic cancer pain has never been tested. Therefore, we propose a phase II, randomized, placebo-controlled trial to evaluate the efficacy and safety of intranasal esketamine in chronic opioid refractory cancer pain. Methods and analysis: We will recruit 120 subjects with chronic opioid refractory pain, defined as pain lasting more than 3 months despite optimal therapy with high dose opioids (>60 mg morphine equivalent dose/day) and optimal adjuvant therapy. Subjects will be randomized into two groups: intranasal esketamine (56mg) and placebo. Treatment will be administered twice a week for four consecutive weeks. The primary outcome is defined as reduction in the Numeric Pain Rating Scale (NPRS) after first application. Secondary outcomes include NPRS reduction after four weeks, the number of daily morphine rescue doses, functional status and satisfaction, and depression. Conclusion: This study may extend therapeutic options in patients with chronic pain, thus improving their quality of life and reducing opioid use. Trial registration: Clinical Trials.gov, NCT04666623. Registered on 14 December 2020",
"keywords": [
"Cancer pain",
"Esketamine",
"Pain measurement",
"Opioid refractory"
],
"content": "Introduction\n\nRefractory pain is a challenging condition in Oncology. More than half of the patients with advanced stage cancer experienced pain1–3 and about 10 to 20% of the pain in cancer is refractory4. The etiology is complex, with a mix of nociceptive, neuropathic and inflammatory mechanisms5, resulting in a situation that can be difficult to control. Tolle et al., described in 2000 that around 46% of family members of terminal patients referred that there is lack of adequate pain treatment when death is imminent6. When opioids are not enough to relieve pain, it is necessary to associate an adjuvant, such as ketamine, as recommended by current National Comprehensive Cancer Network guidelines in pain not responding to other analgesics5,7,8.\n\nKetamine was discovered in 1980, described as a non-competitive antagonist of N-methyl-D-aspartate (NMDA) receptor causing the inhibition of excitatory glutamate receptors in the central nervous system5. It is commonly used as a general anesthetic agent, administered by intravenous or intramuscular route, but other usage and routes for this drug are being studied. Ketamine can cause not only dissociative anesthesia but also an analgesic and antidepressant effect5. Low doses of ketamine, have minimal effect on cardiovascular and respiratory function, produce analgesia, modulate central sensitization, hyperalgesia and opioid tolerance9.\n\nOne of the ketamine enantiomers largely studied nowadays for other potential use is esketamine, the most effective optical isomer of ketamine3. The advantage of this enantiomer is that it induces less dissociative effects and more powerful analgesia and anesthesia10. As an antidepressant, esketamine was approved by the FDA10, and as an analgesic it is not labeled yet due to the lack of studies assessing its efficacy and safety in relieving pain10,11.\n\nIntranasal esketamine use is approved by the FDA for refractory major depression12. There are a couple of recently completed or ongoing trials testing intranasal esketamine for pain13,14. A pilot study recently published found similar results in analgesia between esketamine plus midazolam via intranasal, and intravenous morphine by patient controlled analgesia in spinal surgery patients13.\n\nEsketamine has shown promising results superior to those observed for ketamine due to its more potent analgesic effect, associated with less psychomimetic side effects referred above15,16. However, a large randomized controlled trial is needed to prove that this drug is effective to relieve pain, is safe, and causes a reduction in opioid consumption.\n\nFinding a better approach to manage severe and refractory cancer pain through rapid and efficient analgesia would redound to the benefits of the health care of these patients. Having an alternative to escalating doses of opioid would allow a reduction in opioid consumption in this population. This would decrease the risk of overdose and the frequent and unwanted side effects associated with high doses of these drugs. Opioid consumption is an epidemic health care concern in the United States and many other countries and efforts are strongly directed to combat it17.\n\nThis research aims to assess the efficacy and safety of a four-week treatment with intranasal esketamine (56 mg) twice a week, combined with opioid analgesic and adjuvant standard therapy in the management of adult patients with severe and opioid refractory chronic cancer pain.\n\nThe primary aim is to investigate the analgesic efficacy of intranasal esketamine as an add-on therapy to potentially improve opioid resistant pain in cancer patients. We will evaluate whether intranasal esketamine alleviates chronic pain in this population by assessing the reduction with the Numeric Pain Rating Scale (NPRS) after the first dose of esketamine compared to baseline18.\n\nOur secondary aims are:\n\n1. Assessment of safety19 of this novel treatment. This will be measured by a self-reported and questionnaire-guided adverse events report.\n\n2. Evaluation if esketamine treatment reduces the number of morphine rescues required daily by the patient.\n\n3. To compare the patients’ functional status between both treatment groups.\n\n4. To evaluate the response of depression score to esketamine therapy\n\n5. To analyze if there is a difference in the efficacy of esketamine depending on the type of cancer and depression disorder of the patient.\n\n6. To analyze the sustainability of the effect after each dose throughout the time of treatment.\n\nWe propose a single-center, randomized, placebo-controlled, phase II, double-blinded with two parallel groups, superiority trial. The primary endpoint will be the change in pain intensity measured by the NPRS18 (at week 0, 1, 2, 3, and 4, measured exactly in between doses).\n\n\nProtocol\n\nThis is version 2 of the protocol from the 22nd of December 2020.\n\nThe study will be conducted at a reference tertiary cancer hospital that will be selected based on willingness to participate and availability of qualified participants willing to take part in the study.\n\nWe plan to conduct a single hospital trial and randomization will be done by the OxMaR software. Allocation sequence will be concealed for everyone involved in the trial, except for the pharmacist and the nurse responsible for patient’s allocation. This allocation will remain hidden for health care professionals and participants until the final result of the study are obtained20.\n\nFirst approach will be performed by patient's physician and subsequent recruitment steps will be conducted by trained clinical staff. Written informed consent will be explained and signed. The recruiting physician will have access to a computer with OxMar software, the initial forms will be fulfilled and imputed personal data should be confirmed. The software will generate the allocation, which is automatically sent to the pharmacist that will provide the medication or placebo to the participant. The access to this allocation will be protected by Swordfish (http://www.swordfishsecurity.com/).\n\nThe pharmacist will be responsible for formulating and packaging (identically) the trial medicines (placebo and esketamine nasal spray). He or she will label and deliver them with corresponding randomized number received from OxMar records.\n\nAll investigators, patients, care providers, outcome assessors, and study statisticians will remain blinded with respect to the treatment allocation throughout the trial. The masking will be assured by delivering identical unlabeled nasal injectors with the same volume (200 μL) and color of active drug and placebo. Each device will deliver 28mg of esketamine or placebo in a 200 μL solution, so in order to achieve the 56mg dose two devices will be required21,22.\n\nEsketamine nasal spray has a certain taste often described as bitter, that can cause dysgeusia. Therefore, blinding of subjects will be further guaranteed by adding denatonium benzoate to the water-based placebo agent mimicking the taste of esketamine intranasal solution.\n\nOn each dosing day during the trial, participants will self-administer at 2 time points 1 spray of the study drug (esketamine or placebo) into each nostril. Each administration will be separated by 5 minutes. The administration will take place at the facility to ensure adherence and surveillance in regard to side effects.\n\nEmergency unblinding. In case of problems and safety concerns that cannot be solved with on-going blinded treatment, the participant’s allocated intervention will be revealed. Unblinding procedure will be performed at the request of the Data Monitoring Committee (DMC; see information below).\n\nDuring the period of treatment, the patient's status will be evaluated weekly by an experienced blinded anesthesiologist. Vital signs and oxygen saturation levels will be examined and recorded. Also a physical examination will take place to follow up the patients’ functional status.\n\nSealed envelopes will be stored safely and access will be available 24/7 if needed. Treating physicians and the responsible pharmaceutical staff will be instructed and encouraged to maintain blinding status if possible. Certainly, they are advised to maintain blinding towards other subjects and ensure disclosure towards corporate sponsors, office staff, data analysts and study personnel.\n\nThe Investigator must report all code breaks (with reason) as they occur on the corresponding electronic case report form page. Unblinding should not necessarily be a reason for study drug discontinuation.\n\nInclusion criteria:\n\nAge ≥ 18 years; Male or female patients.\n\nPatients with refractory cancer pain, defined as: multiple evidence-based biomedical therapies used in a clinically appropriate and acceptable fashion have failed to reach treatment goals that mainly include adequate pain reduction and/or improvement in daily living functioning activities; AND/OR patients’ functional activities do not allow a quality of life, which is acceptable and/or pharmaceutical therapies have resulted in intolerable adverse effects.\n\nPsychiatric disorders and psychosocial factors that could influence pain outcomes have been assessed and appropriately addressed23,24; in practice, when the patient with cancer pain responds positively to any of these two questions:\n\n1. Is your pain under the current drug medication intolerable for your quality of life?\n\n2. Do you have intolerable adverse effects with your current treatment?\n\nCancer pain classified as chronic (persistent or recurrent pain lasting longer than 3 months)25, and currently refractory despite optimized analgesic therapy including an opioid. [Optimized analgesic therapy is arbitrarily defined as: oral morphine equivalent of 60 mg/d or more11,26 (or another strong opioid at optimized dose) plus at least one adjuvant analgesic drug, for at least 2 weeks.]\n\nNo increase in baseline long acting opioid dose or addition of a new adjuvant analgesic drug within 2 weeks prior to study entry\n\nAbility to communicate the intensity of pain using the NPRS pain scale ranging from (0 as no pain to 10 with severe pain).\n\nAbility to give fully informed written consent.\n\nExpect survival more than 3 months.\n\nExclusion criteria, taking into consideration the safety profile of the drug22 were defined as:\n\nHistory of allergy or intolerance to esketamine or ketamine.\n\nHistory of allergy to disinfecting products containing quaternary ammonium, who might be susceptible to be allergic to denatonium benzoate.\n\nConcomitant use of xanthine derivatives (e.g. aminophylline, theophylline), ergometrine, or monoamine oxidase inhibitors.\n\nActive nasal/sinus dysfunction (e.g. allergic or infectious rhinitis) or presence of any lesion of the nasal mucosa.\n\nPregnancy, breastfeeding and women of childbearing potential not using a highly effective contraception method.\n\nUncontrolled hypertension, arrhythmia, heart failure, or untreated coronary artery disease. History of transient ischemic attacks, stroke, neurovascular disease, hemorrhage, severe head injury, hydrocephalus or elevated intracranial pressure within the last 3 months.\n\nHistory of primary or metastatic malignant brain lesions (uncontrolled or without previous treatment).\n\nKnown aneurysmal vascular disease (including thoracic and abdominal aorta, intracranial, and peripheral arterial vessels) or arteriovenous malformation\n\nUncontrolled psychiatric illness with psychosis/hallucination (e.g. schizophrenia, acute psychosis).\n\nAlcohol abuse, drug abuse/dependence within the past 6 months as self-reported.\n\nCirrhosis or severe hepatic impairment defined as 5-fold elevation of transaminases.\n\nUncontrolled hyperthyroidism.\n\nGlobe injuries or increased intraocular pressure (e.g. glaucoma).\n\nHistory of ulcerative or interstitial cystitis.\n\nSubjects scheduled to receive radiotherapy (RT) to a site of pain during the study period, or who have received RT to a site of pain within 2 weeks before study entry.\n\nSubjects scheduled to undergo surgical treatment during the study period likely to affect pain.\n\nSubjects on or starting chemotherapy if there is a significant expectation of that therapy affecting pain.\n\nSubjects who have not provided signed informed consent form.\n\nConcomitant use of drugs moderately or severely affecting cytochrome P450 activity.\n\nParticipants will be recruited at the oncology and pain clinics of a tertiary academic hospital by the attending physicians during routine appointments. Participants may also be identified via review of medical records from the oncology and pain clinics and contacted by telephone prior to the next scheduled appointment with information about the study, in order to allow more time for consideration. Screening will continue until the target sample size is achieved (120 subjects as calculated sample size, see below). Expected recruitment time will be two years.\n\nTo achieve adherence in the present clinical trial, the characteristics and main objective of the study will be exposed to the patients and their relatives who are in charge of their care during recruitment interviews and informed consent. An instructional video will be provided, which contains information about the correct drug administration, as well as the benefits of having a quick and easy application therapy. Adjuvant drugs currently used will be provided at each visit.\n\nIn the initial interview with the subject the following information will be given:\n\n- Objectives of the study\n\n- Mode of administration of interventions\n\n- Importance of continuing with regular treatment (opioids, antidepressants, etc.)\n\nFigure 1 demonstrates a timeline with concise description of scheduled visits and assessments.\n\nThis figure displays the scheduled timeline of the study. Included are randomization, application of study drug/placebo (nasal spray devices), clinical assessments and time point of data collection (results).\n\nThe patients will be using around the clock baseline analgesia with long-acting opioid, and morphine rescue dose for the management of breakthrough pain. In the case of neuropathic pain, patients should start the use of an antidepressant or anticonvulsant in an optimized dose. If the participant is already using any of these drugs, they should continue the medication1 .\n\nThe study drug will be provided in two disposable nasal spray devices containing either 28 mg of esketamine or placebo in a 200 uL of solution - equivalent to two sprays doses. The 28 mg esketamine nasal spray device labeled “SpravatoTM” and a similar placebo device should be provided by the manufacturer Janssen Pharmaceutical Company of Johnson & Johnson. The participants should initially self-administer the nasal spray device containing either active drug or placebo under the supervision of a health care provider in the researcher center. Health care providers (physicians or nurses) will be informed and trained prior to the beginning of recruitment. Training will be provided by the investigators. One spray must be administered into each nostril in a 45-degree reclination to deliver 28mg. After a 5 minute interval the second device is used to deliver the second 28mg dose or placebo22.\n\nThe nasal spray device contains an indicator with two green dots referring to the two spray doses. After one spray is delivered, one green dot will disappear. When the device empties after the two doses, all green dots will have disappeared. The participants will receive the intervention twice weekly for 4 consecutive weeks.\n\nThe primary outcome of this trial is a reduction in the 11-point Numeric Pain Rating Scale (NPRS). It will be used to assess pain intensity at enrollment and at each visit. Patients will be asked to rate their weekly pain on a scale from 0 to 10 where 0 equals “no pain” and 10 equals \"the worst pain they can imagine”27. NPRS will be taken for both during physical activity and at rest. The difference in NRPS after first dose of esketamine and baseline will be measured to analyze efficacy.\n\nThe use of a unidimensional measure is justified as the goal of this study is to evaluate the efficacy of the investigational drug as an analgesic complement for patients already consuming morphine. There are several tools to approach pain intensity and this choice (NRPS) was based on literature review on similar situations27–30.\n\nOncological patients suffering from chronic pain frequently present associated degrees of depression, which is why we chose a scale evaluating exclusively pain over other multidimensional questionnaires. This strategy enables to differentiate the analgesic action of esketamine from its role as antidepressant, as this itself might improve the patients’ overall well-being and disposition.\n\nSecondary outcomes. Sustainability of effect. NRPS will be used to identify the pattern of effect after each dose over the 4 weeks.\n\nRescue morphine use. The use of morphine rescue (whether it will be reduced, no change, or increased). If pain decreases or stabilizes during the trial. This will be monitored using either the Aircure artificial intelligence through a mobile application or personal diaries.\n\nPatients’ functional status and satisfaction. These will be measured by the change in Brief Pain Inventory (BPI). A multidimensional standardized tool as the BPI will allow the researchers to assess pain intensity in relation to its interference in functional activities31.\n\nDepression symptoms. Depression is frequently described in association with chronic pain and affects its threshold. On the other hand, esketamine has a proven antidepressive effect. We propose to examine a depression score using the Patient Health Questionnaire (PHQ9) at enrollment and at each visit32. We would also evaluate the effect of depression on the analgesic effect of esketamine.\n\nSide effects. Reported side effects in other studies include decline in mood, conscious perceptions and intellectual performance. It is well established that esketamine presents fewer side effects than the racemic ketamine33. In this protocol the dissociative effects of the drug will be monitored applying the Side Effect Rating Scale for Dissociative Anesthetics (SERSDA) at enrollment and at each visit34. It assesses fatigue, dizziness, headache, nausea, changes in vision and mood changes. The cognitive performance will be addressed using Speed Reading Test (where the time for reading 20 independent characters is noted)33.\n\nVital signs. Blood pressure, heart rate, respiratory rate and temperature, pulse oximetry and 12-lead-ECGs will be measured at enrollment and at each visit (prior, 40 minutes and 2 hours after intranasal application of the drug or placebo). In the occurrence of any abnormality, patients will be advised to stay at the institution for complementary evaluation. Furthermore, lower urinary tract symptoms will be asked and in case of positive answers, further examination will be performed35.\n\nA DMC will be established, which is independent of the trial sponsor and study investigators. The DMC will be composed by experts of the field related to the study36 (at least 1 oncologist, 1 anesthesiologist, 1 psychiatrist, 1 pharmacologist, 1 patient advocacy, and 1 statistician). All members of the DMC will have to declare competing interests. Besides, the DMC will also monitor recruitment, inclusion rate, baseline characteristics, participant adherence, adverse events and safety monitoring every 15 days confidentially37.\n\nSolicited and spontaneously reported adverse events will be collected by the outcome assessors at each planned visit for a clinical assessment. Details about efficacy and safety of the trial medication have been reported previously (see above).\n\nSample size is based on practical and clinical considerations for this phase II trial. Based on a previous study (case series) assessing the analgesic effect of nasal esketamine38 we obtained a standard deviation of 5.5 (that is the 75% upper limit of the confidence interval of the standard deviation). However, we would like to emphasize that power alone will not be relevant for the clinical interpretation of the study results, especially in a clinical trial39,40.\n\nConsidering the population of patients with refractory cancer pain (20%), to achieve a primary outcome reduction of 3 points on the 11 points NPRS with a power of 80% and an alpha error of 5%, a sample size of 108 subjects is needed. Based on previous studies in this patient population a dropout rate of 10% is estimated39,41, so for reasons of compensation, a total sample size of 120 participants will be recruited42,43.\n\nFor the purpose of descriptive statistics all numerical continuous data will be presented as means or medians with standard deviation (SD) and interquartile ranges (IQR) respectively, as measures of dispersion. Categorical data will be presented in the form of proportions, frequencies or percentages. For all primary and secondary analysis, the intervention arm (nasal esketamine plus opioid therapy plus adjuvant standard analgesia) will be compared with the control group (placebo plus opioid therapy plus adjuvant standard therapy).\n\nThe P-values will be reported with two decimal points; all our tests will be 2-sided p-values with a level of significance of <0.05 to consider statistical significance. The statistical software to be employed is an up-to-date version of STATA (StataCorp LLC, College Station, Texas, USA), managed by professional blinded statisticians. After the analysis of all data, if the effect of esketamine doesn't last until next dose, a clustered analysis might be performed using results from all visits to evaluate acute efficacy of the drug against placebo.\n\nMissing data. According to our sample size calculation, we are considering an estimation of missing data of 10% of the participants, which may leave the study for several reasons that can include death secondary to cancer, adverse reactions, lack of improvement, early recovery, and others44,45.\n\n\nEthical considerations\n\nResearch ethics committee/institutional review board (REC/IRB) approval will be obtained by the local ethics committee or institutional review board of the site center prior to begin of recruitment.\n\nProtocol amendments, if any, should be registered and approved by the local ethics committee. Accordingly, study protocol and if applicable informed consent form will be adjusted and investigators involved in the trial informed immediately.\n\n\nDissemination\n\nAfter completion of data acquisition and analysis, the results will be summarized and published. Accordingly, patients will be informed about allocation and trial results via telephone calls and optional visit in person. Manuscript writing will be performed by the investigators and there will be no public access granted to the raw data.\n\nPersonal information about enrolled patients will be collected by study nurses of the study site. Collected information will be saved in coded folders and investigators, outcome assessors and statisticians will receive a pseudononymized version of the data. The data and information will be stored at a previously defined location after completion of the trial for 5 additional years.\n\nOnly investigators and statisticians will have access to the final version of the data. To secure restricted access to the data, corresponding files will be coded.\n\nMeanwhile and after completion of the trial patients will be treated according to current guidelines and standards of care. Any harm associated with the participation of the trial is not expected as patients continue to receive standard of care.\n\n\nStudy status\n\nNot yet recruiting.\n\n\nData availability\n\nNo data is associated with this article.\n\nFigshare: SPIRIT and TIDieR checklist for ‘IMPORTANCE trial: a provisional study-design of a single-center, phase II, double-blinded, placebo-controlled, randomized, 4-week study to compare the efficacy and safety of intranasal esketamine in chronic opioid refractory pain’, https://doi.org/10.6084/m9.figshare.13435742.v246.\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).",
"appendix": "Acknowledgements\n\nThis study protocol was developed within the Principles and Practice in Clinical Research (PPCR) Course of Harvard T.H. Chan School of Public Health. Professor of Epidemiology at Harvard T.H. Chan School of Public Health: Felipe Fregni, MD, PhD, MMSc, MPH, Med\n\n\nReferences\n\nFallon M, Giusti R, Aielli F, et al.: Management of cancer pain in adult patients: ESMO Clinical Practice Guidelines. Ann Oncol. 2018; 29(Suppl 4): iv166–iv191. PubMed Abstract | Publisher Full Text\n\nSingh V, Gillespie TW, Harvey RD: Intranasal Ketamine and Its Potential Role in Cancer-Related Pain. Pharmacotherapy. 2018; 38(3): 390–401. PubMed Abstract | Publisher Full Text\n\nvan den Beuken-van Everdingen MHJ, de Rijke JM, Kessels AG, et al.: Prevalence of pain in patients with cancer: a systematic review of the past 40 years. Ann Oncol. 2007; 18(9): 1437–49. PubMed Abstract | Publisher Full Text\n\nHardy J, Quinn S, Fazekas B, et al.: Randomized, Double-Blind, Placebo-Controlled Study to Assess the Efficacy and Toxicity of Subcutaneous Ketamine in the Management of Cancer Pain. J Clin Oncol. 2012; 30(29): 3611–7. PubMed Abstract | Publisher Full Text\n\nBell RF, Eccleston C, Kalso EA: Ketamine as an adjuvant to opioids for cancer pain. Cochrane Database Syst Rev. 2017; 6(6): CD003351. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTolle SW, Tilden VP, Rosenfeld AG, et al.: Family Reports of Barriers to Optimal Care of the Dying. Nurs Res. 2000; 49(6): 310–7. PubMed Abstract | Publisher Full Text\n\nManji J, Thamboo A, Sunkaraneni V, et al.: The association of Leptospermum honey with cytokine expression in the sinonasal epithelium of chronic rhinosinusitis patients. World J Otorhinolaryngol Head Neck Surg. 2018; 5(1): 19–25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNCCN: NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines®) Genetic/Familial High-Risk Assessment: Breast and Ovarian. Version 3.2019. 2019. Reference Source\n\nSchatzberg AF: A Word to the Wise About Intranasal Esketamine. Am J Psychiatry. 2019; 176(6): 422–4. PubMed Abstract | Publisher Full Text\n\nKim JW, Huh G, Rhee CS, et al.: Unsupervised cluster analysis of chronic rhinosinusitis with nasal polyp using routinely available clinical markers and its implication in treatment outcomes. Int Forum Allergy Rhinol. 2019; 9(1): 79–86. PubMed Abstract | Publisher Full Text\n\nShteamer JW, Harvey RD, Spektor B, et al.: Safety of intranasal ketamine for reducing uncontrolled cancer-related pain: Protocol of a phase i/II clinical trial. JMIR Res Protoc. 2019; 8(4): e12125. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLapidus KAB, Levitch CF, Perez AM, et al.: A Randomized Controlled Trial of Intranasal Ketamine in Major Depressive Disorder. Biol Psychiatry. 2014; 76(12): 970–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRiediger C, Haschke M, Bitter C, et al.: The analgesic effect of combined treatment with intranasal S-ketamine and intranasal midazolam compared with morphine patient-controlled analgesia in spinal surgery patients: a pilot study. J Pain Res. 2015; 8: 87–94. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPeltoniemi MA, Hagelberg NM, Olkkola KT, et al.: Ketamine: A Review of Clinical Pharmacokinetics and Pharmacodynamics in Anesthesia and Pain Therapy. Clin Pharmacokinet. 2016; 55(9): 1059–77. PubMed Abstract | Publisher Full Text\n\nFregni F, Illigens BMW, editors: Critical Thinking in Clinical Research. Critical Thinking in Clinical Research. Oxford University Press. 2018. Reference Source\n\nMuller J, Pentyala S, Dilger J, et al.: Ketamine enantiomers in the rapid and sustained antidepressant effects. Ther Adv Psychopharmacol. 2016; 6(3): 185–92. PubMed Abstract | Publisher Full Text | Free Full Text\n\nScholl L, Seth P, Kariisa M, et al.: Drug and Opioid-Involved Overdose Deaths — United States, 2013–2017. MMWR Morb Mortal Wkly Rep. 2018; 67(5152): 1419–1427. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCaraceni A, Cherny N, Fainsinger R, et al.: Pain Measurement Tools and Methods in Clinical Research in Palliative Care: recommendations of an Expert Working Group of the European Association of Palliative Care. J Pain Symptom Manage. 2002; 23(3): 239–55. PubMed Abstract | Publisher Full Text\n\nShrestha R, Pant S, Shrestha A, et al.: Intranasal ketamine for the treatment of patients with acute pain in the emergency department. World J Emerg Med. 2016; 7(1): 19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nO’Callaghan CA: OxMaR: Open Source Free Software for Online Minimization and Randomization for Clinical Trials. Añel JA, editor. PLoS One. 2014; 9(10): e110761. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDaly EJ, Singh JB, Fedgchin M, et al.: Efficacy and safety of intranasal esketamine adjunctive to oral antidepressant therapy in treatment-resistant depression: A randomized clinical trial. JAMA Psychiatry. 2018; 75(2): 139–48. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJanssen Pharmaceuticals: DailyMed - SPRAVATO- esketamine hydrochloride solution. DailyMed - U.S. National Library of Medicine. 2019. Reference Source\n\nDeer TR, Caraway DL, Wallace MS: A Definition of Refractory Pain to Help Determine Suitability for Device Implantation. Neuromodulation. 2014; 17(8): 711–5. PubMed Abstract | Publisher Full Text\n\nWorld Health Organization: WHO guidelines for the pharmacological and radiotherapeutic management of cancer pain in adults and adolescents. World Health Organization. 2018. Reference Source\n\nScholz J, Finnerup NB, Attal N, et al.: The IASP classification of chronic pain for ICD-11: chronic neuropathic pain. Pain. 2019; 160(1): 53–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDowell D, Haegerich TM, Chou R: CDC Guideline for Prescribing Opioids for Chronic Pain--United States, 2016. JAMA. 2016; 315(15): 1624–45. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDansie EJ, Turk DC: Assessment of patients with chronic pain. Br J Anaesth. 2013; 111(1): 19–25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHjermstad MJ, Fayers PM, Haugen DF, et al.: Studies Comparing Numerical Rating Scales, Verbal Rating Scales, and Visual Analogue Scales for Assessment of Pain Intensity in Adults: A Systematic Literature Review. J Pain Symptom Manage. 2011; 41(6): 1073–93. PubMed Abstract | Publisher Full Text\n\nHawker GA, Mian S, Kendzerska T, et al.: Measures of adult pain: Visual Analog Scale for Pain (VAS Pain), Numeric Rating Scale for Pain (NRS Pain), McGill Pain Questionnaire (MPQ), Short-Form McGill Pain Questionnaire (SF-MPQ), Chronic Pain Grade Scale (CPGS), Short Form-36 Bodily Pain Scale (SF-36 BPS), and Measure of Intermittent and Constant Osteoarthritis Pain (ICOAP). Arthritis Care Res (Hoboken). 2011; 63(Suppl 11): S240–52. PubMed Abstract | Publisher Full Text\n\nDworkin RH, Turk DC, Farrar JT, et al.: Core outcome measures for chronic pain clinical trials: IMMPACT recommendations. Pain. 2005; 113(1-2): 9–19. PubMed Abstract | Publisher Full Text\n\nShahid A, Wilkinson K, Marcu S, et al.: Brief Pain Inventory (BPI). In: STOP, THAT and One Hundred Other Sleep Scales. New York, NY: Springer New York. 2011; 81–8. Publisher Full Text\n\nShreders A, Niazi S, Hodge D, et al.: Universal screening for depression in cancer patients and its impact on management patterns. J Clin Oncol. 2016; 34(26_suppl): 232–232. Publisher Full Text\n\nPfenninger EG, Durieux ME, Himmelseher S: Cognitive Impairment after Small-dose Ketamine Isomers in Comparison to Equianalgesic Racemic Ketamine in Human Volunteers. Anesthesiology. 2002; 96(2): 357–66. PubMed Abstract | Publisher Full Text\n\nAhern TL, Herring AA, Stone MB, et al.: Effective analgesia with low-dose ketamine and reduced dose hydromorphone in ED patients with severe pain. Am J Emerg Med. 2013; 31(5): 847–51. PubMed Abstract | Publisher Full Text\n\nKim J, Farchione T, Potter A, et al.: Esketamine for Treatment-Resistant Depression - First FDA-Approved Antidepressant in a New Class. N Engl J Med. 2019; 381(1): 1–4. PubMed Abstract | Publisher Full Text\n\nEllenberg SS: Independent data monitoring committees: rationale, operations and controversies. Stat Med. 2001; 20(17-18): 2573–83. PubMed Abstract | Publisher Full Text\n\nLin JY, Lu Y: Establishing a data monitoring committee for clinical trials. Shanghai Arch Psychiatry. 2014; 26(1): 54–56. PubMed Abstract | Free Full Text\n\nJohansson J, Sjöberg J, Nordgren M, et al.: Prehospital analgesia using nasal administration of S-ketamine -- a case series. Scand J Trauma Resusc Emerg Med. 2013; 21(1): 38. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBurstein HJ, Manola J, Younger J, et al.: Docetaxel Administered on a Weekly Basis for Metastatic Breast Cancer. J Clin Oncol. 2000; 18(6): 1212–9. PubMed Abstract | Publisher Full Text\n\nBacchetti P: Current sample size conventions: Flaws, harms, and alternatives. BMC Med. 2010; 8(1): 17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMathibe LJ: Drop-out rates of cancer patients participating in longitudinal RCTs. Contemp Clin Trials. 2007; 28(4): 340–2. PubMed Abstract | Publisher Full Text\n\nKhan I, Sarker SJ, Hackshaw A: Smaller sample sizes for phase II trials based on exact tests with actual error rates by trading-off their nominal levels of significance and power. Br J Cancer. 2012; 107(11): 1801–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRubinstein LV, Korn EL, Freidlin B, et al.: Design Issues of Randomized Phase II Trials and a Proposal for Phase II Screening Trials. J Clin Oncol. 2005; 23(28): 7199–206. PubMed Abstract | Publisher Full Text\n\nSterne JAC, White IR, Carlin JB, et al.: Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ. 2009; 338(jun29 1): b2393–b2393. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim Y: Missing Data Handling in Chronic Pain Trials. J Biopharm Stat. 2011; 21(2): 311–25. PubMed Abstract | Publisher Full Text\n\nDoldi P: TIDieR/SPIRIT Checklist. figshare. Online resource. 2020. http://www.doi.org/10.6084/m9.figshare.13435742.v2"
}
|
[
{
"id": "77932",
"date": "05 Feb 2021",
"name": "Vinita Singh",
"expertise": [
"Reviewer Expertise Intranasal ketamine for cancer and chronic neuropathic pain",
"cancer pain",
"neuropathic pain",
"non-opioid and non-pharmacological treatment for cancer pain"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a well-written description of a prospective clinical trial that will test the efficacy of intranasal esketamine in the reduction of chronic, refractory, cancer pain. A total of 120 patients will be recruited and randomized to receive placebo or intranasal esketamine 56 mg twice a week for 4 weeks. The primary outcome is a reduction in pain score, after the first dose of medication, on Numerical Pain Rating Scale. There is an appropriate data safety monitoring plan.\nFew critiques below:\nPHQ-9 is a screening tool for depression, but not measuring changes in depression.\n\nThere is a section for missing data, however, how missing data will be handled is not specified. Also, there's can be several reasons for missing data besides patients leaving the study.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes",
"responses": []
},
{
"id": "79031",
"date": "15 Feb 2021",
"name": "Danny Kupka",
"expertise": [
"Reviewer Expertise Cardiology",
"Oncology",
"Immunology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article represents the protocol of the IMPORTANCE trial. The IMPORTANCE trial (phase II, randomized and double-blind) evaluates the efficacy and safety of intranasal administration of S-Ketamine in chronic opioid refractory cancer pain. The primary outcome is the reduction in the Numeric Pain Rating Scale.\n\nThe study is well designed with appropriate questionnaires to determine primary and secondary outcomes. For me, sample size calculation based on ref. 38 is possible but addresses a different patient set. Maybe the authors comment on that briefly. Moreover, I would comment on the fact that no midazolam is used in combination with S-Ketamin. However, this would imply 3 study groups (S-Ketamine + Midazolam i.n. versus Midazolam i.n. versus placebo).\n\nThe manuscript sounds scientific. I have no additional comments on this issue.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Not applicable",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-42
|
https://f1000research.com/articles/9-1239/v1
|
15 Oct 20
|
{
"type": "Software Tool Article",
"title": "netDx: Software for building interpretable patient classifiers by multi-'omic data integration using patient similarity networks",
"authors": [
"Shraddha Pai",
"Philipp Weber",
"Ruth Isserlin",
"Hussam Kaka",
"Shirley Hui",
"Muhammad Ahmad Shah",
"Luca Giudice",
"Rosalba Giugno",
"Anne Krogh Nøhr",
"Jan Baumbach",
"Gary D. Bader",
"Philipp Weber",
"Ruth Isserlin",
"Hussam Kaka",
"Shirley Hui",
"Muhammad Ahmad Shah",
"Luca Giudice",
"Rosalba Giugno",
"Anne Krogh Nøhr",
"Jan Baumbach"
],
"abstract": "Patient classification based on clinical and genomic data will further the goal of precision medicine. Interpretability is of particular relevance for models based on genomic data, where sample sizes are relatively small (in the hundreds), increasing overfitting risk netDx is a machine learning method to integrate multi-modal patient data and build a patient classifier. Patient data are converted into networks of patient similarity, which is intuitive to clinicians who also use patient similarity for medical diagnosis. Features passing selection are integrated, and new patients are assigned to the class with the greatest profile similarity. netDx has excellent performance, outperforming most machine-learning methods in binary cancer survival prediction. It handles missing data – a common problem in real-world data – without requiring imputation. netDx also has excellent interpretability, with native support to group genes into pathways for mechanistic insight into predictive features. The netDx Bioconductor package provides multiple workflows for users to build custom patient classifiers. It provides turnkey functions for one-step predictor generation from multi-modal data, including feature selection over multiple train/test data splits. Workflows offer versatility with custom feature design, choice of similarity metric; speed is improved by parallel execution. Built-in functions and examples allow users to compute model performance metrics such as AUROC, AUPR, and accuracy. netDx uses RCy3 to visualize top-scoring pathways and the final integrated patient network in Cytoscape. Advanced users can build more complex predictor designs with functional building blocks used in the default design. Finally, the netDx Bioconductor package provides a novel workflow for pathway-based patient classification from sparse genetic data.",
"keywords": [
"precision medicine",
"networks",
"classification",
"supervised learning",
"genomics",
"data integration"
],
"content": "Introduction\n\nSupervised learning methods are useful in clinical genomics for disease diagnosis, risk stratification for prognosis, and evaluating treatment response. Machine learning is a powerful analytic approach that can identify patterns separating patient groups, but interpreting models remains an active area of research1. Interpretability is desirable to better understand biological mechanism underlying the phenotype and for rational treatment design. It is also important for genomic applications, where most contemporary datasets have fewer than a thousand samples, increasing the risk of overfit models that do not independently replicate. Separately, most machine learning methods do not handle missing data – a common feature of real-world datasets – without prior data imputation or filtering. netDx is a supervised learning algorithm that classifies patients by integrating multimodal patient data2. It is notable among machine learning methods for handling missing data without imputation, and excels at interpretability by enabling users to create biologically-meaningful grouping of features, such as grouping genes into pathway-level features. netDx integrates multi-modal data by converting each layer into a patient similarity network and then integrating these networks (Figure 1a).\n\n(a) Conceptual visualization of patient similarity networks. Nodes are patients and edge weights measure pairwise similarity. The example shows a two-class problem (high and low risk patients), with four features shown as patient similarity networks: similarity for clinical (red), gene expression (green), metabolomic (blue) and mutation (orange) data. (b) Conceptual workflow for netDx predictor. Samples are split into train and test samples, and training samples are subjected to feature selection (blue flow). Feature selection uses regularized regression to integrate networks, such that networks with non-zero regression weights have their feature score increased. This process is repeated with different subsamples of training data, for a user-provided maximum of times (featScoreMax). This process is repeated for each patient label. Features passing a user-specified threshold are used to classify held-out samples. Test patients are classified by highest similarity. Patient networks that combine training and test patients are then integrated; only networks from features passing selection are used for this step. Label propagation is used to compute similarity of each test patient to training samples from each label; a given patient is assigned to the class with highest similarity. Average model performance is computed by running this whole process over several train/test splits. Features with consistent high scores can be used to classify an independent validation set.\n\nThis paper provides an introduction to the R-based software implementation of netDx for the Bioconductor system3 and showcases common use cases. The details of the netDx algorithm and performance have been previously pubished2, though we provide a brief conceptual summary here (Figure 1b). As input, the user provides multiple sets of data measured on the same set of labelled patients. The user additionally provides functions to compute pairwise patient similarity and optionally, rules to group measures from each data type into features (feature design). For example, gene expression measures could be grouped into features representing known biological pathways. As with other machine learning methods, the user specifies parameters for training the model, such as the threshold scores for feature selection. Consider an application to predict good or poor patient survival, using tumour-derived gene expression, DNA methylation and proteomic data. In this scenario, netDx is provided with three data tables, one per ‘omic data type, a table with patient identifiers and known labels, and the grouping rule that one feature is to be created per data layer. Patients are then automatically split into training and test samples and feature selection is performed using training samples. netDx uses the given feature processing rules to convert data from different modalities into a common space of patient similarity networks1,2. Feature selection is performed once per patient label, and features passing selection are used to classify patients from the held-out test data. Performance robustness is evaluated by repeating this feature selection and classification exercise for multiple train/test splits. The final model is created by choosing features that consistently score highly. A feature may comprise an entire data layer, a single variable, or specified groupings; one example of the last is grouping gene-level measures into pathways, so that each pathway is a separate feature. Interpretability is aided by the pathway-level predictor design, which identifies cellular processes with predictive value.\n\n\nMethods\n\nnetDx3 is integrated into the Bioconductor system, a high-quality computational biology software framework for genomic data analysis in the statistical programming language R4. Figure 2 shows the workflow for building a model using the netDx software package; Table 1 describes major function calls. netDx uses Bioconductor data structures and mechanisms for fetching and storing data, and representation of input data.\n\nThe yellow box shows data provided to netDx to build the predictor. See use cases for examples. Patient data is provided as a MultiAssayExperiment object, with the patient metadata in the colData slot. Variable grouping rules, used for feature design, are provided as a list of lists. The outer list corresponds to a given assay; each entry in the corresponding list corresponds to one group of measures and is used to create a single feature. For instance, in a pathway-based design, one entry in the outer list would be \"gene expression\", with each inner list containing genes grouped by pathways. In this scenario, each of these gene groupings generates a single pathway-level feature. Pathway definitions can be automatically fetched using the fetchPathwayDefinitions() function, or custom definitions can be provided in the GMT file format. For the workflow using sparse genetic data (see Use Case 3), such as CNVs, patient CNVs are provided to netDx as a GRanges object. In this instance, pathways are provided in GRangesList format.\n\nIntermediate functions used to prepare data are not shown, but are illustrated in the use cases. The third column shows outdated names for these functions from the software release that accompanied the paper describing the methods.\n\nRunning netDx requires a machine with one or more cores running Intel processors (1.7 GHz i7 or later) or equivalent, a minimum of 1Gb RAM per thread, and 1Gb disk space. Feature selection is an embarrassingly parallel problem, and we highly recommend running the software on a multi-core machine to reduce compute time. netDx is currently supported on machines running OS X or Unix-based operating systems. The software requires the Java executable (v1.8 or higher) to be available on the system path, and will not work on recent Windows-based operating systems that lack this type of installation. netDx v1.1.4 requires R>=3.6 and BioConductor release 3.11 or higher. When building a predictor, patient data is provided to netDx as a MultiAssayExperiment object, a Bioconductor data structure used to represent multi-’omics experiments associated with a given set of samples. In our usage, data types are collected in the assays slot of the object; the sole exception is clinical data, which is provided as part of the sample metadata, using the colData slot. Grouping rules are provided as a nested list object - or list-of-lists (groupList). The outer list consists of one entry per data type, with corresponding groupings in the inner list. Assays names must be identical in the assays slot and in groupList.\n\nThe easiest way to build classifiers is to use the wrapper function, buildPredictor(). This function runs feature selection and classification over a specified number of train/test splits, and returns all associated feature scores and detailed classification results in a list object. Advanced users can create custom predictor designs by combining the individual steps used in buildPredictor() (Table 1).\n\n\nUse cases\n\nThis section describes four use cases for building predictors with netDx. The first uses pathway-level features based on gene expression data to generate a binary classifier of breast cancer subtype. The second performs three-way classification of breast cancer subtype by integrating gene expression, DNA methylation and proteomic assays. The third builds a binary classifier of autism spectrum disorder diagnosis from sparse genetic mutations. The fourth involves prediction of tumour stage from somatic mutations that have been desparsified using prior knowledge of gene interaction networks.\n\nIntroduction. In this example, we will build a binary breast tumour Luminal A subtype classifier from clinical data and gene expression data. We will use different rules to create features for each assay. Specifically:\n\nClinical measures (e.g. age, stage): Features are defined at the level of variables; similarity is defined as normalized difference.\n\nGene expression: Features are defined at the level of pathways; similarity is defined by pairwise Pearson correlation.\n\nFeature scoring is automatically performed over multiple random splits of the data into train and blind test partitions. Feature selected networks are those that consistently score highly across the multiple splits (e.g. those that score 9 out of 10 in ≥70% of splits).\n\nConceptually, this is what the higher-level logic looks like for building a predictor over multiple random splits of samples into training and test groups. In the example below, the predictor runs for 100 train/test splits. Within a split, features are scored from 0 to 10. Features scoring ≥9 are used to predict labels on the held-out test set (20%). The example shows pseudocode, not actual netDx function calls:\n\n\n\nSetup\n\n\n\nData. In this example, we use curated data from The Cancer Genome Atlas, through the Bioconductor curatedTCGAData package. The goal is to classify a breast tumour into either a Luminal A subtype or otherwise. The predictor integrates clinical variables selected by the user, along with gene expression data.\n\nHere we load the required packages and download clinical and gene expression data.\n\n\n\nList the available data without downloading any:\n\n\n\nWe will work only with the gene expression data in this example:\n\n\n\nThis next code block prepares the TCGA data. In practice you would do this once, and save the data before running netDx, but we run it here in full to see an end-to-end example.\n\n\n\nThe predictor will look for columns named ID and STATUS columns in the sample metadata table. netDx uses these to get the patient identifiers and labels, respectively.\n\n\n\nDesign custom patient similarity networks (features). netDx provides a set of default functions to compute patient similarity, including Pearson correlation, normalized difference, and scaled Euclidean distance. However, users may choose to define a custom function that takes patient data and variable groupings as input, and returns a set of patient similarity networks (PSN) as output. The user can customize what datatypes are used, how they are grouped, and what defines patient similarity for a given datatype.\n\nWhen running the predictor (next section), the user simply passes this custom function as an input variable; i.e. the makeNetFunc parameter when calling buildPredictor().\n\nNote: While netDx supports flexible experimental design, the user must ensure that the design, i.e. the similarity metric and variable groupings are appropriate for a given application. Domain knowledge is recommended to support good design.\n\nnetDx requires that the makeNetFunc function take some generic parameters as input. These include:\n\ndataList: the patient data, provided as a MultiAssayExperiment object. Refer to online tutorials for MultiAssayExperiment to see how to construct those objects from data.\n\ngroupList: sets of input data that will define individual networks (e.g. genes grouped into pathways)\n\nnetDir: the directory where the resulting patient similarity networks will be stored.\n\ndataList\n\nIn this example, the breast cancer data is already provided to us as a MultiAssayExperiment object:\n\n\n\ngroupList\n\nThis object tells the predictor how to group units when constructing a network. For example, genes may be grouped into a patient similarity network representing a pathway. This object is a list; the names match those of dataList while each value is itself a list and reflects a potential network.\n\n\n\nSo the groupList variable has one entry per data layer:\n\n\n\nEach entry contains a list, with one entry per feature. Here we have three pathway-level features for mRNA and two variable-level features for clinical data.\n\nFor example, here are the networks to be created with RNA data. Genes corresponding to pathways are to be grouped into individual network. Such a groupList would create pathway-level networks:\n\n\n\nFor clinical data, we will define each variable as its own network:\n\n\n\nDefine patient similarity measure for each network. This function is defined by the user and tells the predictor how to create networks from the provided input data.\n\nThis function requires dataList, groupList, and netDir as input variables. The residual ... parameter is to pass additional variables to makePSN_NamedMatrix(), notably numCores (number of parallel jobs).\n\nIn this example, the custom similarity function does the following:\n\n1. Creates pathway-level networks from RNA data using the default Pearson correlation measure makePSN_NamedMatrix(writeProfiles=TRUE, ...)\n\n2. Creates variable-level networks from clinical data using a custom similarity function of normalized difference: makePSN_NamedMatrix(writeProfiles=FALSE, simMetric=\"custom\", customFunc=normDiff).\n\n\n\nNote: dataList and groupList are generic containers that can contain whatever object the user requires to create a PSN. The custom function supports flexible feature design.\n\nBuild predictor. Finally, we call the function that runs the netDx predictor. We provide:\n\n• number of train/test splits over which to collect feature scores and average performance (numSplits),\n\n• maximum score for features in one round of feature selection (featScoreMax)\n\n• threshold to call feature-selected networks for each train/test split (featSelCutoff); only features scoring this value or higher will be used to classify test patients, and\n\n• the information to create the PSN, including patient data (dataList), how variables are to be grouped into networks (groupList) and the custom function to generate features (makeNetFunc).\n\nChange numCores to match the number of cores available on your machine for parallel processing.\n\nThe call below runs two train/test splits. Within each split, it:\n\n• splits data into train/test using the default split of 80:20\n\n• scores networks between 0 to 2 (i.e. featScoreMax=2)\n\n• uses networks that score ≥1 out of 2 (featSelCutoff) to classify test samples for that split.\n\nThese are unrealistically low values set so the example will run fast. In practice a good starting point is featScoreMax=10, featSelCutoff=9 and numSplits=100, but these parameters may need to be tuned to the sample sizes in the dataset and heterogeneity of the samples. Datasets with high levels of heterogeneity or small sample sizes may benefit from increased sampling – i.e. higher numSplits value. Increasing this setting increases the time to train the model but identifies generalizable patterns over a larger set of random subsamples.\n\n\n\nExamine output\n\nThe results are stored in the list object returned by the buildPredictor() call. This list contains:\n\ninputNets: all input networks that the model started with.\n\nSplit<i>: a list with results for each train-test split\n\n– predictions: real and predicted labels for test patients\n\n– accuracy: percent accuracy of predictions\n\n– featureScores: feature scores for each label (list with g entries, where g is number of patient labels). Each entry contains the feature selection scores for the corresponding label.\n\n– featureSelected: vector of features that pass feature selection. List of length g, with one entry per label.\n\n\n\nReformat results for further analysis\n\nThis code collects different components of model output to examine the results.\n\n\n\nCompute model performance\n\nAfter compiling the data above, plot accuracy for each train/test split:\n\n\n\nCreate a ROC curve, a precision-recall curve, and plot average AUROC and AUPR (Figure 3):\n\n\n\nClockwise from top-left: Mean area under ROC curve across train/test splits; mean area under precision-recall curve; precision-recall curve (average in blue; individual splits in grey); ROC curves (average in blue; individual splits in grey).\n\nExamine feature scores and consistently high-scoring features. Use getNetConsensus() to convert the list data structure into a single table, one per patient label. The rows show train/test splits and the columns show features that consistently perform well.\n\nWe then use callFeatSel() to identify features that consistently perform well across the various train/test splits. Because this is a toy example, we set the bar low to get some features. Here we accept a feature if it scores 1 or higher (fsCutoff=1) in even one split (fsPctPass=0.05), setting the latter to a low positive fraction.\n\n\n\nWhere features are scored out of 10, a reasonable setting is fsCutoff=9 and fsPctPass=0.7. This setting gives us features that score a minimum of 9 in at least 70% of the train/test splits.\n\n\n\nVisualize pathway features as an enrichment map. An enrichment map is a network-based visualization of pathway connectivity and is used in netDx to visualize themes in predictive pathway-based features5. It is used in conjunction with the AutoAnnotate Cytoscape app to identify clusters, and apply auto-generated labels to these6.\n\nUse getEMapInput_many() to create the input that helps generate the enrichment map in Cytoscape.\n\n\n\nWrite the results to files that Cytoscape can read in:\n\n\n\nFinally, plot the enrichment map. This step requires Cytoscape to be installed, along with the EnrichmentMap and AutoAnnotate apps. It also requires the Cytoscape application to be open and running on the machine running the code. This block is commented out for automatic builds on Bioconductor, but a screenshot of the intended result is shown below (Figure 4).\n\n\n\nThe small number of nodes reflects the limited number of pathways provided to the toy example model, and also reduced parameter values for model building. See Figure 5 for an example of a more informative enrichment map produced by running a real-world example.\n\nThis example enrichment map isn’t terribly exciting because of the low number of pathway features permitted, the upper bound on feature selection scores and low number of train/test splits in the demonstration example.\n\nHere is an example of an enrichment map generated by running the above predictor with more real-world parameter values, and all available pathways (Figure 5):\n\nThis network is generated by running the plotEmap() function, which uses the RCy3 Bioconductor package to programmatically call Cytoscape network visualization software from within R, to run the EnrichmentMap app5–7. Nodes show pathways features that scored a minimum of 9 out of 10 in feature selection, in at least 70% of train/test splits; node fill indicates feature score. Edges connect pathways with shared genes. The larger yellow bubbles are auto-generated by the AutoAnnotate Cytoscape app6,8; these thematically group top pathways, by clustering and word-frequency based cluster annotation.\n\nVisualize integrated patient similarity network based on top features. We apply a threshold to define the most predictive features, and integrate these into a single patient similarity network. Such a network is useful for downstream operations such as ascertaining whether or not classes are significantly separated, and for visualization of results.\n\nHere we define predictive features as those scoring 2 out of 2 in all train/test splits.\n\n\n\nWe next examine the features:\n\n\n\nCreate a new groupList limited to top features:\n\n\n\nWe plot the integrated patient network based on the features selected above.\n\nIn the example below, the networks are integrated by taking the mean of the edge weights (aggFun=\"MEAN\"). For plotting we retain only the top 5% strongest edges (topX=0.05).\n\nBy setting calcShortestPath=TRUE, the function will also compute the pairwise shortest path for within- and across-group nodes. The result is shown as a set of violin plots and a one-sided Wilcoxon-Mann-Whitney test is used to assign significance.\n\nAs with plotEMap(), this method must be run on a computer with Cytoscape installed and running. To bypass plotting the PSN in Cytoscape, set plotCytoscape to FALSE. This function call computes shortest-path distances within- and among clusters (Figure 6) and plots the integrated PSN (Figure 7). The resulting network is shown below (Figure 7).\n\n\n\nThis visualization and statistic are useful to ascertain whether or not patients of the same label are more similar in the integrated network; having within-class distance be significantly smaller than across-class distance is indicative of good class separation. This graph is generated using the plotIntegratedPatientNetwork() function. From left to right, it shows pairwise patient shortest distances: within patients of class \"LumA\"; between the two class labels; within patients of the residual class \"nonLumA\"; and between all patients in the network.\n\nThis network is generated by calling plotIntegratedPatientNetwork() and uses RCy3 to programmatically generate the network in Cytoscape7,8. This network uses features that scored 2 out of 2 in all train-test splits. For visualization, only the top 8% most-distant edges are shown. Nodes are patients, and edges weights show average similarity across all features passing feature selection. Node fills indicate patient label, with “LumA” in green and “nonLumA” in orange.\n\nThe integrated PSN can also be visualized as a tSNE plot:\n\nThe integrated PSN can also be visualized as a tSNE plot:\n\n\n\n\n\nIntroduction. In this example, we will use clinical data and three types of ’omic data - gene expression, DNA methylation and proteomic data - to classify breast tumours as being one of three types: Luminal A, Luminal B, or Basal. This example is an extension of the one used to build a binary classifier (see Use Case 1).\n\nWe also use several strategies and definitions of similarity to create features:\n\n• Clinical variables: Each variable (e.g. age) is its own feature; similarity is defined as normalized difference.\n\n• Gene expression: Features are defined at the level of pathways; i.e. a feature groups genes within the pathway. Similarity is defined as pairwise Pearson correlation.\n\n• Proteomic and methylation data: Features are defined at the level of the entire data layer; a single feature is created for all of proteomic data, and the same for methylation. Similarity is defined as pairwise Pearson correlation.\n\nSetup. Load the netDx package.\n\n\n\nData. For this example, we download data from The Cancer Genome Atlas through the Bioconductor curatedTCGAData package. The fetch command automatically creates a MultiAssayExperiment object containing the data.\n\n\n\nWe use the curatedTCGAData() command to explore available data types in the breast cancer dataset.\n\n\n\nIn this call we fetch only the gene expression, proteomic and methylation data; setting dry.run=FALSE initiates the fetching of the data.\n\n\n\nThis next code block prepares the TCGA data. In practice this is performed once, and the resulting data is saved before running netDx, but we run it here to see an end-to-end example.\n\n\n\nThe important thing is to create ID and STATUS columns in the sample metadata slot. netDx uses these to get the patient identifiers and labels, respectively.\n\n\n\nRules to create features (patient similarity networks). We will group gene expression data by pathways and clinical data by single variables. We will treat methylation and proteomic data each as a single feature, so each of those groups will contain the entire input table for those corresponding data types.\n\nIn the code below, we fetch pathway definitions from January 2018 from a source that auto-compiles these from curated pathway databases (http://download.baderlab.org/EM_Genesets). We choose the January 2018 source to be consistent with earlier published work, but normally the latest source would be downloaded. We group gene expression measures by pathways.\n\nGrouping rules are accordingly created for the clinical, methylation and proteomic data.\n\n\n\nDefine patient similarity for each network. We provide netDx with a custom function to generate similarity networks (i.e. features). The first block tells netDx to generate correlation-based networks using everything but the clinical data. This is achieved by the call:\n\n\n\nTo make features from single measures using clinical data, the second block makes a slightly-modified call to makePSN_NamedMatrix(), this time requesting the use of the normalized difference similarity metric. This is achieved by calling:\n\n\n\nnormDiff is a function provided in the netDx package, but the user may define custom similarity functions in this block of code and pass those to makePSN_NamedMatrix(), using the customFunc parameter.\n\n\n\nBuild predictor. Finally, we make the call to build the predictor.\n\n\n\nCompute accuracy for three-way classification:\n\n\n\nOn examining the confusion matrix above, we can see that the model perfectly classifies basal tumours, but performs poorly in distinguishing between the two types of luminal tumours. This performance is unsurprising because luminal and basal tumours have different molecular characteristics, with the latter being ER- tumours; in contrast, both Luminal A and B are both types of ER+ tumours9.\n\n\n\nnetDx natively handles missing data, making it suitable to build predictors with sparse genetic data such as somatic DNA mutations, frequently seen in cancer, and from DNA copy number variations (CNVs). This example demonstrates how to use netDx to build a predictor from sparse genetic data. Here we build a case/control classifier for autism spectrum disorder (ASD) diagnosis, starting from rare CNVs; for this, we use data from Pinto et al.10. The design for this predictor is shown in Figure 9.\n\nCNVs are grouped into pathway-level features and patient similarity is binary; i.e. two patients have similarity of one if they share CNVs in genes from the same pathway. Feature selection is iteratively performed on independent thirds of the sample set. This design uses an additional label enrichment step that precedes feature selection. Label enrichment filters out networks with insufficient bias towards case-case edges, using a label-based permutation approach. Networks with significant label enrichment are used in feature selection. Scores from all three feature-selection splits are added to get a final score for each feature, with a maximum attainable score of 30. Test patients are classified as cases if they carry a CNV in a pathway that passes feature selection.\n\nDesign and adapting the algorithm for sparse event data. In this design, we group CNVs by pathways. The logic behind the grouping is prior evidence showing that genetic events in diseases tend to converge on cellular processes of relevance to the pathophysiology of the disease10.\n\nBinary similarity and label enrichment\n\nIn this design, similarity is defined as a binary function, a strategy that has advantages and drawbacks. In plain terms, if two patients share a mutation in a pathway, their similarity for that pathway is 1; otherwise it is zero. This binary definition, while conceptually intuitive, increases the false positive rate in the netDx feature selection step. That is, networks with even a single case sample will get a high feature score, regardless of whether that network is enriched for case samples.\n\nTo counter this problem, we introduce a label-enrichment step in the feature selection. A bias measure is first computed for each network, such that a network with only cases scores +1; one with only controls scores -1; and one with an equal number of both has a score of zero. Label-enrichment compares the bias in each real network, to the bias in that network in label-permuted data. It then assigns an empirical p-value for the proportion of times a label-permuted network has a bias as high as the real network. Only networks with a p-value below a user-assigned threshold (default: 0.07) pass label-enrichment, and feature selection is limited to these networks. In netDx, label-enrichment is enabled by setting enrichLabels=TRUE in the call to buildPredictor_sparseGenetic().\n\nCumulative feature scoring\n\nThe other difference between this design and those with non-sparse data, is the method of scoring features (Figure 9). The user specifies a parameter which indicates the number of times to split the data and run feature selection. The algorithm then runs feature selection numSplits times, each time leaving 1/numSplits of the samples out. In each split, features are scored between zero and featScoreMax, using the same approach as is used for continuous-valued input. Feature scores are then added across the splits so that a feature can score as high as numSplits*featScoreMax.\n\nEvaluating model performance\n\nFor a given cutoff for features, a patient is called a “case” if they have a genetic event in pathways that pass feature selection at that cutoff; otherwise, at that cutoff, they are labelled a “control”. These calls are used to generate the false positive and true positive rates across the various cutoffs, which ultimately generates a ROC curve.\n\nSetup\n\n\n\nData. CNV coordinates are read in, and converted into a GRanges object. As always, the sample metadata table, here the pheno object, must have ID and STATUS columns.\n\n\n\nGroup CNVs by pathways. The fetchPathwayDefinitions() function downloads pathway definitions from baderlab.org but users may provide custom .gmt files as well. We use the BiocFileCache package to download gene definitions for the hg18 genome build, and convert these a GRanges object. The function mapNamedRangesToSets() is used to group this GRanges object into pathway-level sets.\n\n\n\nGroup gene extents into pathway-based sets, which effectively creates grouping rules for netDx. The function mapNamedRangesToSets() does this grouping, generating a GRangesList object.\n\n\n\nRun predictor. Once the phenotype matrix and grouping rules are set up, the predictor is called using buildPredictor_sparseGenetic(). Note that unlike with non-sparse data, the user does not provide a custom similarity function in this application; currently, the only option available is the binary similarity defined above. As discussed above, setting enrichLabels=TRUE to enable label-enrichment is highly recommended to reduce false positive rate.\n\n\n\nPlot results. Feature selection identifies pathways that are consistently enriched for the label of interest; here, “case” status. From the diagnostic point of view, a patient with a genetic event in a selected feature - here, a CNV in a feature-selected pathway - is labelled a “case”. “True positives” are therefore cases with CNVs in feature-selected pathways, while “false positives” are controls with CNVs in feature-selected pathways. These definitions are used to compute the ROC curve below (Figure 10).\n\nIn this design, patients are classified as cases if they carry a CNV in pathways passing feature selection, and controls otherwise. Each dot in the graph shows the sensitivity/specificity for a given cutoff for feature selection.\n\n\n\nWe can also compute the AUROC and AUPR.\n\n\n\nThis predictor performs outperforms previous CNV-based classifiers11; in a real-world scenario this model would need to be validated on an independent dataset. In our experience, using a combination of sparse genetic data and binary similarity makes classifiers prone to overfitting. Measures commonly used to mitigate overfitting include training the model on larger datasets, and larger number of train/test splits are advised.\n\nPathway scores are also added across the splits, for a total of 9 across the 3 splits (3 + 3 + 3).\n\n\n\nAs before, running the predictor with all possible pathway-related features and realistic training parameters, such as numPermsEnrich=200L, featScoreMax=10L, numSplits=3L identifies a much richer set of themes related to synaptic transmission and cell proliferation, consistent with the known biology of ASD as well as those identified in the original publication12.\n\nNodes show pathway features cumulatively scoring 13 or higher out of 30, while edges connect pathways with common member genes. Node fill indicates pathway score, with yellow for the lowest and red for the highest.\n\nThe nodes in Figure 11 have been reorganized to group clusters sharing a broader theme. Terms related to neurotransmission and synaptic plasticity are in the bottom left, those related to the cell cycle and proliferation are in the top-right, and those related to immune function are in the bottom right.\n\nThe dynamic range of feature scores is much larger as well, here ranging from 0 to 30. The resulting ROC curve in Figure 12 accordingly has 30 cutoffs at which specificity and sensitivity are evaluated, evidenced by 30 datapoints in that curve. This is in contrast to 9 cutoffs in the ROC curve shown in Figure 10.\n\nLegend identical to Figure 9, except that here the graph is comprised of 30 measures because features are scored out of 30, rather than in Figure 9, where features are scored out of 9.\n\nnetDx provides the option of reducing the sparsity of mutation data by inferring \"indirect mutations\" using prior knowledge of gene-gene interaction networks. Conceptually, the logic is that if a patient has a mutation in a given gene, the mutation indirectly impacts interacting genes. Indirect mutation is inferred by label propagating patient mutations over a gene-gene interaction network onto neighbours. The resulting smoothed network is then used for downstream applications. This network-based smoothing improved mutation-based tumour class discovery in four types of cancer13. For label propagation, we use an R-based implementation of random walk with restart, a popular strategy in bioinformatic applications13–16. The result of using this strategy on a patient’s binary somatic mutation profile is a non-sparse profile in which genes are assigned a continuous score between zero and one, that reflects its network proximity to patient mutations. This propagation value is then ranked and binarized, with the top-ranked fraction set to one; this fraction defaults to 3% and is tunable. The binarization serves to limit inferred mutation to genes closest to the known mutations. For instance, genes distant from the patient's mutation would get a low propagation value, and would be thresholded to zero, i.e. not considered to be mutated. The result of this step is a less sparse binary matrix, which serves as input data to the predictor.\n\nIn this example, we use direct and inferred somatic mutations to classify Testicular Germ Cell Tumours (TGCT)17 by binarized pathologic stage. As with the previous use case, we create pathway-level features to reflect that cancer progression occurs by a combination of genes acting in molecular networks corresponding to cancer hallmark processes such as cell proliferation and apoptosis14,18. As in Use Case 3, similarity used is the binary function. If two patients share a mutation in a pathway, their similarity for that pathway is one; otherwise it is zero.\n\nSetup\n\n\n\nData. Clinical and genetic data are downloaded using the Bioconductor package curatedTCGAData. Mutations are converted to a binary matrix format where rows represent genes, columns represent patients; entry [i,j] is set to one if gene i has a somatic mutation, and zero otherwise.\n\n\n\nSmooth mutations over a gene interaction network. The gene-gene interaction network used in this example contains high-confidence cancer-specific interactions19. This specific network effectively clusters tumour samples of patients, distinguishing them by tumour type and time of survival. This is a binary symmetric network.\n\n\n\nsmoothMutations_LabelProp() is used to smooth the mutations using the provided interaction network, by using label propagation. The output of this method is a continuous-valued network which reflects the network proximity of the non-zero values to the original mutations.\n\n\n\nFinally, the smoothed matrix is binarized. Genes with a propagation value greater than a specified cutoff are set to one, with the rest set to zero. This step ensures that genes which get a low propagation value are not used. Genes with lower smoothed values reflect those farther from the original mutation, and setting these to zero signifies a lack of confidence that these were impacted.\n\n\n\nCreate pathway-level features with binary patient similarity. Smoothed mutations are now grouped at the level of biological pathways. As with other examples, pathways are downloaded from a compilation of curated pathway databases (GMT format). Thereafter, we define pathway-level patient similarity to be binary; i.e. if two patients share a mutation in genes from the same pathway, their mutual similarity is one; else it is zero. Individual steps below use identical functions to those used in the first use case above.\n\n\n\nNow we define functions for patient similarity:\n\n\n\nBuild predictor. Finally, we compile all the data into a MultiAssayExperiment object and as before, run the predictor.\n\n\n\nThe predictor call is essentially the same as with other simpler designs:\n\n\n\nExamine output. This code collects different components of model output to examine the results.\n\n\n\nPlot the AUROC and AUPR curves (Figure 13):\n\n\n\nExamine features with the highest scores. Here, these are pathways with somatic mutations that best predict vital status:\n\n\n\n\nSoftware updates\n\nnetDx v1.1.4 has several updates relative to the version released with the netDx methods report (v1.0.23)2. The new netDx package supports OS X and Unix platforms. It also supports Windows systems, with the exception of those that do not have the Java executable available in the system search path. The companion R package netDx-examples, previously used to store example data, is now deprecated. All examples are now either contained within the netDx package or are fetched from Bioconductor using local file-caching via the FileCache package. Major functions have been renamed to reflect their role rather than implementation, making their usage more intuitive (Table 1). The current version of netDx includes a novel workflow of building a classifier from sparse genetic data (see Use Case 3), using the function buildPredictor_sparseGenetic(). We also added the functionality to generate an integrated patient similarity network from features passing selection. The plotIntegratedPatientNetwork() function generates this network, computes statistics on pairwise shortest distance measures (Dijkstra distance) within and across labels, and automatically generates a network visualization in Cytoscape.\n\nA number of software updates were made as part of Bioconductor integration. Unlike the previous version where all user output was written to a specific output directory, all predictor output is now returned to users as R objects, and intermediate work is written to temporary directories by default. The turnkey predictor-building function no longer automatically generates a log file; rather, users are required to create their own log files using the R sink() function. Functions computing model performance and plotting no longer assume a directory structure created by the model-building step. Users now set random number generator seeds at the outset, instead of providing a seed as an input parameter to various functions. Automated network visualization in Cytoscape now uses RCy3, for programmatic access of Cytosape from R.\n\nMemory improvements were made to the underlying GeneMANIA network integration algorithm Java implementation20,21, creating a modified version specifically for netDx. netDx incurs a relatively higher memory footprint because each feature in netDx internally generates a similarity network with pairwise similarity measures. Network integration, a step in feature selection, requires keeping all these networks in memory. Certain grouping rules also incur a greater memory footprint than others. Notably, a model with pathway-level features converts one gene expression data matrix into ~2,000 pathway-level patient similarity networks; such a design is less scalable in the number of nodes, than one which creates a single feature based on all gene expression. We optimized netDx memory usage by customizing the underlying GeneMANIA Java application used for network integration. netDx uses a modified version of the GeneMANIA implementation, which bypasses steps not required for the netDx pipeline, such as the identifier conversion and steps involving file input/output. Memory and computational time improvements were benchmarked by building binary classifiers for breast tumours and schizophrenia case-control classification. The CommonMind Consortium22 dataset (downloaded from Synapse: syn5607607) included 279 controls and 258 cases, with a total of 537 patients, with gene expression data from the prefrontal cortex organized into pathway level features (1,735 pathways). The breast cancer data was part of the TCGA project9, with tumour gene-expression for 348 patients, including 154 Luminal A and 194 tumours of other subtypes, also organized into pathway-level features (1,706 pathways). In the benchmark, an approximately 70:30 split of samples was used for cross validation. We measured training time for the predictor using the 70% of samples of a single subtype. All tests were performed on an Intel Xeon @ 2.6GHz machine with 126 GB of available RAM and 12 cores. During benchmarking, threads had a fixed amount of RAM available, with discrete steps of 4 GB, 6 GB and 8 GB. Here each predictor was built using only a single core. Benchmarking runs were parallelized using GNU parallel23, where the performance was averaged over four runs of the 10 queries for each datasets. Following improvements, memory use dropped to one-third of the original amount. With the updated software, the CommonMind dataset also required two-thirds of the time to build the predictor, as compared to with the original version (Table 2).\n\nComputation times are averaged over four runs of the same ten queries for feature-scoring one patient label, while limiting the executable to a single core. All tests were performed on an Intel Xeon @ 2.6GHz machine with 126GB of available RAM and 12 cores.\n\nFinally, the feature selection step now provides the option of using a Monte Carlo resampling strategy for selecting samples for iterative feature scoring. The previous version of the software required a fraction of samples to be held out, the fraction being directly related to the maximum feature score. The Monte Carlo resampling approach should allow users to increase the upper-bound of feature scores, even in smaller samples.\n\n\nConclusions\n\nThe updated netDx software provides an improved user experience for clinical research applications pertaining to risk stratification, treatment response, and to identify biomarkers associated with patient subtypes. Method extensions will be required for further feature additions, such as the ability to predict continuous outcome. Classification of borderline patients could be better controlled, perhaps by a user-specified margin. Similar to pathway-level grouping for gene expression data, other grouping strategies will be required for other types of genomic data, such as miRNA, single nucleotide polymorphisms, and brain imaging data.\n\n\nData availability\n\nData for the autism case/control classification5 is provided as part of the netDx package.\n\nData for the breast cancer example is from The Cancer Genome Atlas9,24. They are fetched from the curatedTCGAData package which is maintained by the Bioconductor repository.\n\nnetDx vignettes are available at: https://bioconductor.org/packages/release/bioc/html/netDx.html\n\n\nSoftware availability\n\nnetDx is available from Bioconductor: http://bioconductor.org/packages/devel/bioc/html/netDx.html\n\nSource code available from: https://github.com/BaderLab/netDx.\n\nArchived source code at time of publication: http://doi.org/10.5281/zenodo.40488523.\n\nIssue tracker: https://github.com/BaderLab/netDx/issues\n\nLicense: MIT License",
"appendix": "Acknowledgements\n\nWe greatly appreciate the input from Marcel Ramos and the Bioconductor core development team in guiding the software development for integration of netDx with Bioconductor. CommonMind data were generated as part of the CommonMind Consortium supported by funding from Takeda Pharmaceuticals Company Limited, F. Hoffmann-La Roche Ltd and NIH grants R01MH085542, R01MH093725, P50MH066392, P50MH080405, R01MH097276, RO1-MH-075916, P50M096891, P50MH084053S1, R37MH057881, AG02219, AG05138, MH06692, R01MH110921, R01MH109677, R01MH109897, U01MH103392, and contract HHSN271201300031C through IRP NIMH. Brain tissue for the study was obtained from the following brain bank collections: the Mount Sinai NIH Brain and Tissue Repository, the University of Pennsylvania Alzheimer’s Disease Core Center, the University of Pittsburgh NeuroBioBank and Brain and Tissue Repositories, and the NIMH Human Brain Collection Core. CMC Leadership: Panos Roussos, Joseph Buxbaum, Andrew Chess, Schahram Akbarian, Vahram Haroutunian (Icahn School of Medicine at Mount Sinai), Bernie Devlin, David Lewis (University of Pittsburgh), Raquel Gur, Chang-Gyu Hahn (University of Pennsylvania), Enrico Domenici (University of Trento), Mette A. Peters, Solveig Sieberts (Sage Bionetworks), Thomas Lehner, Stefano Marenco, Barbara K. Lipska (NIMH).\n\n\nReferences\n\nPai S, Bader GD: Patient Similarity Networks for Precision Medicine. J Mol Biol. 2018; 430(18 Pt A): 2924–2938. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPai S, Hui S, Isserlin R, et al.: netDx: interpretable patient classification using integrated patient similarity networks. Mol Syst Biol. 2019; 15(3): e8497. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPai S, Weber P, Giudice L, et al.: BaderLab/netDx: Freeze of code for netDx software manuscript (Version v1.1.4). Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4048852\n\nHuber W, Carey VJ, Gentleman R, et al.: Orchestrating high-throughput genomic analysis with Bioconductor. Nat Methods. 2015; 12(2): 115–121. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMerico D, Isserlin R, Stueker O, et al.: Enrichment map: a network-based method for gene-set enrichment visualization and interpretation. PLoS One. 2010; 5(11): e13984. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKucera M, Isserlin R, Arkhangorodsky A, et al.: AutoAnnotate: A Cytoscape app for summarizing networks with semantic annotations [version 1; peer review: 2 approved]. F1000Res. 2016; 5: 1717. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGustavsen JA, Pai S, Isserlin R, et al.: RCy3: Network biology using Cytoscape from within R [version 2; peer review: 3 approved]. F1000Res. 2019; 8: 1774. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShannon P, Markiel A, Ozier O, et al.: Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003; 13(11): 2498–2504. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCancer Genome Atlas Network: Comprehensive molecular portraits of human breast tumours. Nature. 2012; 490(7418): 61–70. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPinto D, Delaby E, Merico D, et al.: Convergence of genes and cellular pathways dysregulated in autism spectrum disorders. Am J Hum Genet. 2014; 94(5): 677–694. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEngchuan W, Dhindsa K, Lionel AC, et al.: Performance of case-control rare copy number variation annotation in classification of autism. BMC Med Genomics. 2015; 8 Suppl 1(Suppl 1): S7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPinto D, Delaby Elsa, Merico Daniele, et al.: Convergence of genes and cellular pathways dysregulated in autism spectrum disorders. Am J Hum Genet. 2014; 94(5): 677–94. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHofree M, Shen JP, Carter H, et al.: Network-based stratification of tumor mutations. Nat Methods. 2013; 10(11): 1108–1115. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKreeger PK, Lauffenburger DA: Cancer systems biology: a network modeling perspective. Carcinogenesis. 2010; 31(1): 2–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRonen J, Akalin A: netSmooth: Network-smoothing based imputation for single cell RNA-seq [version 3; peer review: 2 approved]. F1000Res. 2018; 7: 8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVanunu O, Magger O, Ruppin E, et al.: Associating genes and protein complexes with disease via network propagation. PLoS Comput Biol. 2010; 6(1): e1000641. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShen H, Shen H, Shen H, et al.: Integrated Molecular Characterization of Testicular Germ Cell Tumors. Cell Rep. 2018; 23(11): 3392–3406. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHanahan D, Weinberg RA: Hallmarks of cancer: the next generation. Cell. 2011; 144(5): 646–674. PubMed Abstract | Publisher Full Text\n\nHuang JK, Jia T, Carlin DE, et al.: pyNBS: a Python implementation for network-based stratification of tumor mutations. Bioinformatics. 2018; 34(16): 2859–2861. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWarde-Farley D, Donaldson SL, Comes O, et al.: The GeneMANIA prediction server: biological network integration for gene prioritization and predicting gene function. Nucleic Acids Res. 2010; 38(Web Server issue): W214–220. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZuberi K, Franz M, Rodriguez H, et al.: GeneMANIA prediction server 2013 update. Nucleic Acids Res. 2013; 41(Web Server issue): W115–122. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFromer M, Roussos P, Sieberts SK, et al.: Gene expression elucidates functional impact of polygenic risk for schizophrenia. Nat Neurosci. 2016; 19(11): 1442–1453. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTange O: GNU Parallel - The Command-Line Power Tool. The USENIX Magazine. 2011; 42–47. Publisher Full Text\n\nCiriello G, Gatza ML, Beck AH, et al.: Comprehensive Molecular Portraits of Invasive Lobular Breast Cancer. Cell. 2015; 163(2): 506–519. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "73048",
"date": "29 Oct 2020",
"name": "Kim-Anh Lê Cao",
"expertise": [
"Reviewer Expertise Computational statistics",
"multi omics integration",
"R software development."
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors illustrate full R workflows to apply netDx using four case studies with various ranges of classification difficulty and data analysis settings. The netDx package proposes various ways of grouping features, for example using known biological pathways, and several graphical outputs. I appreciated that netDx builds on MultiAssayExperiment for easier handling of multi omics data sets. The manuscript will be useful for readers eager to get started with netDx. Below are some suggestions for improvement of the manuscript.\nMethodological aspects:\nI acknowledge that the original algorithm has been detailed in reference [2], however, the present manuscript gives some emphasis on the ability of netDx to handle missing values. A short statement describing how this is done would be helpful.\n\nI suggest rewriting the sentence 'The final model is created by choosing features that consistently score highly.' in the introduction. On a first read, it appeared as if there was selection bias during the process.\n\nI have some reservations regarding the representation of SEM in the AUROC figures, why not using SD?\nImplementation aspects:\nWhile much effort and improvements have been done in the netDx package v1.1.4, I believe that additional functions could be created to be user friendly. For example, many customs functions (e.g. makeNets, the code proposed to reformat the results and calculate the accuracy or to extract the results to Cytoscape, to list a few) could be recoded into more generic functions. I would encourage the authors to revisit the code they propose in these workflows and improve when possible.\n\nI am not sure Table 1 column 3 (function name in v1.0.23) and part of the software update paragraph is useful here. Presumably this could appear on the GitHub page and the NEWS files, unless the objective of this manuscript is also to update the users of the latest changes.\nMinor typos:\n'published' in the Introduction.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "6228",
"date": "22 Jan 2021",
"name": "Shraddha Pai",
"role": "Author Response",
"response": "Responses are shown in bold under corresponding reviewer comments shown in italics. --- The authors illustrate full R workflows to apply netDx using four case studies with various ranges of classification difficulty and data analysis settings. The netDx package proposes various ways of grouping features, for example using known biological pathways, and several graphical outputs. I appreciated that netDx builds on MultiAssayExperiment for easier handling of multi omics data sets. The manuscript will be useful for readers eager to get started with netDx. We thank the reviewer for their time, appreciative comments, and feedback. We also hope the manuscript helps new users get started with netDx for classification and data integration. Below are some suggestions for improvement of the manuscript. Methodological aspects: I acknowledge that the original algorithm has been detailed in reference [2], however, the present manuscript gives some emphasis on the ability of netDx to handle missing values. A short statement describing how this is done would be helpful. Response: Text added in introduction of use case 3: “netDx handles missing data at two levels. First, netDx uses patient similarity networks, not input data, as its features. Missing data can be handled by the similarity metric used to make this conversion. e.g. If similarity is defined as the Pearson correlation between gene expression measures at the pathway level, then omitting missing genes from the correlation calculation still allows the correlations, and thus the pathway-level network,, to be computed. Where patients are missing a particular feature, the network integration step uses what information it has. For example, in a scenario where the data consist of transcriptomic and proteomic measures, if a patient is missing transcriptomic data, the integration step will use only the proteomic data (edges) for that patient (network edges) for that patient.” I suggest rewriting the sentence 'The final model is created by choosing features that consistently score highly.' in the introduction. On a first read, it appeared as if there was selection bias during the process. Response: We altered the sentence in the manuscript to: “The final model is created from features that scored highly in feature selection, a step that uses only training samples”. I have some reservations regarding the representation of SEM in the AUROC figures, why not using SD? Response: We have now changed the function that plots the AUROC curve to use standard deviation as default, and have provided the user the option of using SEM. The corresponding figures have been updated in the manuscript to show the change (Figure 3 and Figure 13). Implementation aspects: While much effort and improvements have been done in the netDx package v1.1.4, I believe that additional functions could be created to be user friendly. For example, many customs functions (e.g. makeNets, the code proposed to reformat the results and calculate the accuracy or to extract the results to Cytoscape, to list a few) could be recoded into more generic functions. I would encourage the authors to revisit the code they propose in these workflows and improve when possible. Response: We agree and will continue to create useful utility functions to make it easier for new users to use in future releases of netDx. In order to avoid potentially over complicating or overengineering our API, we are waiting for user feedback before making these additions. I am not sure Table 1 column 3 (function name in v1.0.23) and part of the software update paragraph is useful here. Presumably this could appear on the GitHub page and the NEWS files, unless the objective of this manuscript is also to update the users of the latest changes. Response: We agree and have removed this column from Table 1. Minor typos: 'published' in the Introduction. Response: Corrected."
}
]
},
{
"id": "73050",
"date": "25 Nov 2020",
"name": "Anais Baudot",
"expertise": [
"Reviewer Expertise Systems and Network Biology",
"Bioinformatics",
"Computational Biology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper entitled “netDx: Software for building interpretable patient classifiers by multi-'omic data integration using patient similarity networks” is a companion paper to the article “netDx: interpretable patient classification using integrated patient similarity networks“ published in Plos Comp Bio in 2019. Its goal is to present an updated version of the R software implementation of netDx as a Bioconductor package.\nThe netDx tool proposes an approach to building a patient classifier from heterogeneous patient data, from clinical to omics. The availability as an R package is of interest to the community. In addition, the manuscript details 4 different uses cases that could help interest readers to apply the tools. We however had difficulties running the code provided in the use cases and obtained different errors and warnings, so have been in contact with the authors to try to solve the problems. However, debugging code necessitate a lot of exchanges, and hundreds of lines of error outputs cannot go in a peer review. These problems are not solved yet.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "6229",
"date": "22 Jan 2021",
"name": "Shraddha Pai",
"role": "Author Response",
"response": "Responses are shown in bold beneath each reviewer comment, the latter shown in italics. --- The paper entitled “netDx: Software for building interpretable patient classifiers by multi-'omic data integration using patient similarity networks” is a companion paper to the article “netDx: interpretable patient classification using integrated patient similarity networks“ published in Plos Comp Bio in 2019. Its goal is to present an updated version of the R software implementation of netDx as a Bioconductor package. The netDx tool proposes an approach to building a patient classifier from heterogeneous patient data, from clinical to omics. The availability as an R package is of interest to the community. In addition, the manuscript details 4 different uses cases that could help interest readers to apply the tools. We however had difficulties running the code provided in the use cases and obtained different errors and warnings, so have been in contact with the authors to try to solve the problems. However, debugging code necessitate a lot of exchanges, and hundreds of lines of error outputs cannot go in a peer review. These problems are not solved yet. Response: We thank the reviewers for taking the time and effort to work with us to resolve these issues. There were three main sources of errors that the reviewers encountered. The resolution for each is described below. To better support Windows users we now provide Docker images of working environments with the latest version of netDx. The following text has been added to the “Operation” section: “Windows users can access netDx via a Docker image provided at https://hub.docker.com/repository/docker/shraddhapai/netdx.” Incompatibility with French locale: The reviewers tested netDx on a system with a French locale, which identified an unforeseen international incompatibility. netDx uses a Java-based network integration software during feature selection. Parts of this software would break when provided with numbers using a comma for a decimal separator; as such they were incompatible with several non-English locales. We have now fixed the issue in netDx v1.3.1 to ensure that all files passed from R to Java are forced to use a period as decimal separator. We are now able to successfully run all vignettes in a French locale; please see https://hub.docker.com/repository/docker/shraddhapai/netdx/general (Tag: v1.3.1_french). Note that French locale users do not have to download this specific version of netDx; the Docker image is provided as a contained environment with a French locale, where we have demonstrated that netDx now works. Windows incompatibility: netDx is currently not supported on some versions of the Windows operating system because of variation in how Java is invoked by different Java versions. In particular, newer Windows systems do not have a java executable available on the search path by default. We have noted this in the current version of the manuscript. Therefore for now we will continue to support netDx on OS X and Unix systems in BioConductor, and will provide a Docker container for Windows users. A working Docker image, supporting Windows and other operating systems that support Docker, is available on Docker hub: https://hub.docker.com/repository/docker/shraddhapai/netdx/general (Tag: v1.3.1) Use case 4 had outdated function calls. That has been amended in the current version of the manuscript. We also supplied the reviewers with the updated vignette when we learnt about the error. We hope that the resolution of the above three issues is satisfactory. We are working with the reviewers to ensure that they are able to run the software given the changes above."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1239
|
https://f1000research.com/articles/10-40/v1
|
21 Jan 21
|
{
"type": "Systematic Review",
"title": "Anosmia and dysgeusia in SARS-CoV-2 infection: incidence and effects on COVID-19 severity and mortality, and the possible pathobiology mechanisms - a systematic review and meta-analysis",
"authors": [
"Endang Mutiawati",
"Marhami Fahriani",
"Sukamto S. Mamada",
"Jonny Karunia Fajar",
"Andri Frediansyah",
"Helnida Anggun Maliga",
"Muhammad Ilmawan",
"Talha Bin Emran",
"Youdiil Ophinni",
"Ichsan Ichsan",
"Nasrul Musadir",
"Ali A. Rabaan",
"Kuldeep Dhama",
"Syahrul Syahrul",
"Firzan Nainu",
"Harapan Harapan",
"Marhami Fahriani",
"Sukamto S. Mamada",
"Jonny Karunia Fajar",
"Andri Frediansyah",
"Helnida Anggun Maliga",
"Muhammad Ilmawan",
"Talha Bin Emran",
"Youdiil Ophinni",
"Ichsan Ichsan",
"Nasrul Musadir",
"Ali A. Rabaan",
"Kuldeep Dhama",
"Syahrul Syahrul",
"Firzan Nainu",
"Harapan Harapan"
],
"abstract": "Background: The present study aimed to determine the global prevalence of anosmia and dysgeusia in coronavirus disease 2019 (COVID-19) patients and to assess their association with severity and mortality of COVID-19. Moreover, this study aimed to discuss the possible pathobiological mechanisms of anosmia and dysgeusia in COVID-19. Methods: Available articles from PubMed, Scopus, Web of Science, and preprint databases (MedRxiv, BioRxiv, and Researchsquare) were searched on November 10th, 2020. Data on the characteristics of the study (anosmia, dysgeusia, and COVID-19) were extracted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline. Newcastle–Ottawa scale was used to assess research quality. Moreover, the pooled prevalence of anosmia and dysgeusia were calculated, and the association between anosmia and dysgeusia in presence of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was assessed using the Z test. Results: Out of 32,142 COVID-19 patients from 107 studies, anosmia was reported in 12,038 patients with a prevalence of 38.2% (95% CI: 36.5%, 47.2%); whereas, dysgeusia was reported in 11,337 patients out of 30,901 COVID-19 patients from 101 studies, with prevalence of 36.6% (95% CI: 35.2%, 45.2%), worldwide. Furthermore, the prevalence of anosmia was 10.2-fold higher (OR: 10.21; 95% CI: 6.53, 15.96, p < 0.001) and that of dysgeusia was 8.6-fold higher (OR: 8.61; 95% CI: 5.26, 14.11, p < 0.001) in COVID-19 patients compared to those with other respiratory infections or COVID-19 like illness. To date, no study has assessed the association of anosmia and dysgeusia with severity and mortality of COVID-19. Conclusion: Anosmia and dysgeusia are prevalent in COVID-19 patients compared to those with the other non-COVID-19 respiratory infections. Several possible mechanisms have been hypothesized; however, future studies are warranted to elucidate the definitive mechanisms of anosmia and dysgeusia in COVID-19. Protocol registration: PROSPERO CRD42020223204.",
"keywords": [
"anosmia",
"COVID-19",
"dysgeusia",
"predictor",
"SARS-CoV-2"
],
"content": "Introduction\n\nCoronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), was initially identified in late December 2019 in Wuhan, Hubei Province, Republic of China1,2. This viral pandemic rapidly spread worldwide, infecting more than 60 million people, causing more than 1 million deaths3, and severely affecting the global healthcare system4,5. Several drugs have been repurposed for treating COVID-195-9; however, no drug has been recommended or approved by the World Health Organization (WHO). The common symptoms of COVID-19 include dry cough, fever, dyspnea, fatigue, anorexia, diarrhea, chest pain, headache, and muscle ache10,11. In particular, two manifestations have been increasingly identified among asymptomatic people that later tested positive for the presence of SARS-CoV-2: anosmia and dysgeusia12. Remarkably, previous studies reported that these olfactory issues were reported in 11.8% of COVID-19 cases before other symptoms occured13-15.\n\nAnosmia, a severe condition of hyposmia, is a part of olfactory dysfunction where the person is unable to sense smell or detect odor16. Dysgeusia is a sensory dysfunction where the individual loses the perception of taste17. The British Association of Otorhinolaryngology reported that both dysfunctions varied from 3-20% among COVID-19 patients18. A previous study among 42 patients revealed that more than a third presented anosmia and dysgeusia19. A higher percentage of anosmia and dysgeusia cases were also reported20. Furthermore, another study reported that anosmia in COVID-19 is related to the enlargement of bilateral olfactory bulb edema21.\n\nThis evidence may be crucial in the present COVID-19 pandemic. As the real-time reverse transcriptase polymerase chain reaction (RT-PCR) test has certain limitations for screening, the manifestation of anosmia and dysgeusia could be used as an early warning for practitioners or clinicians to build a rationale to reach a firm conclusion on patients with SARS-CoV-2 infection22,23. Additionally, a recent study reported that anosmia and dysgeusia are among the earliest symptoms observed in COVID-19 patients24; however, in-depth analysis of this dysfunction and its relation to the pathogenesis, severity, and mortality of COVID-19 is missing from the literature. Thus, the present study aimed to summarize the global evidence of anosmia and dysgeusia among COVID-19 patients, in order to assess their association with the severity and mortality of the disease, and provide a comprehensive review related to the possible pathogenesis of anosmia and dysgeusia in SARS-CoV-2 infection.\n\n\nMethods\n\nTo comprehensively calculate the cumulative prevalence of anosmia and dysgeusia in SARS-CoV-2 infection worldwide, a systematic review was conducted following guideline of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)25. The protocol of this systematic review has been registered at PROSPERO (CRD42020223204).\n\nAll articles reporting anosmia and dysgeusia as the symptom of COVID-19 were included. COVID-19 case was defined by a positive RT-PCR for SARS-CoV-2 from either nasopharyngeal swab, oropharyngeal swab, bronchoalveolar lavage, or cerebrospinal fluid. All cross-sectional, retrospective, and prospective studies that randomly sampled COVID-19 cases from community or hospitals were considered eligible; whereas case reports and case series, including all editorials, reviews, and commentaries, were excluded. Studies targeting specific groups such as pregnant females, children, and other groups, were excluded. Only articles written in English during 2019-2020 were included.\n\nThree bibliographical databases (PubMed, Scopus, and Web of Science) and three preprint databases (MedRxiv, BioRxiv, and Researchsquare) were used to identify the potential articles (as of November 10th, 2020). The search criteria were as follows. PubMed ([Title] “SARS-CoV-2” OR “COVID-19” OR “Wuhan coronavirus” OR “Wuhan virus” OR “novel coronavirus” OR “nCoV” OR “severe acute respiratory syndrome coronavirus 2” OR “coronavirus disease 2019 virus” OR “2019-nCoV” OR “2019 novel coronavirus” OR “severe acute respiratory syndrome coronavirus 2” OR “coronavirus” OR “coronaviruses” OR “SARS 2” OR “2019-nCoV acute respiratory disease” OR “novel coronavirus pneumonia” OR “COVID”) AND ([All] “Anosmia” OR “smell loss” OR “smell dysfunction” OR “smell impairment” OR “hyposmia” OR “dysosmia” OR “olfactory dysfunction” OR “olfactory disorder”) AND (“dysgeusia” OR “taste loss” OR “taste dysfunction” OR “taste impairment” OR “gustatory dysfunction” OR “gustatory disorder” OR “hypogeusia” OR “ageusia”). Scopus ([Title] “SARS-CoV-2” OR “COVID-19” OR “Wuhan coronavirus” OR “Wuhan virus” OR “novel coronavirus” OR “nCoV” OR “severe acute respiratory syndrome coronavirus 2” OR “coronavirus disease 2019 virus” OR “2019-nCoV” OR “2019 novel coronavirus” OR “severe acute respiratory syndrome coronavirus 2” OR “coronavirus” OR “coronaviruses” OR “SARS 2” OR “2019-nCoV acute respiratory disease” OR “novel coronavirus pneumonia” OR “COVID”) AND ([All] “Anosmia” OR “smell loss” OR “smell dysfunction” OR “smell impairment” OR “hyposmia” OR “dysosmia” OR “olfactory dysfunction” OR “olfactory disorder”) AND (“dysgeusia” OR “taste loss” OR “taste dysfunction” OR “taste impairment” OR “gustatory dysfunction” OR “gustatory disorder” OR “hypogeusia” OR “ageusia”). Web of Science ([Title] “SARS-CoV-2” OR “COVID-19” OR “Wuhan coronavirus” OR “Wuhan virus” OR “novel coronavirus” OR “nCoV” OR “severe acute respiratory syndrome coronavirus 2” OR “coronavirus disease 2019 virus” OR “2019-nCoV” OR “2019 novel coronavirus” OR “severe acute respiratory syndrome coronavirus 2” OR “coronavirus” OR “coronaviruses” OR “SARS 2” OR “2019-nCoV acute respiratory disease” OR “novel coronavirus pneumonia” OR “COVID”) AND ([All] “Anosmia” OR “smell loss” OR “smell dysfunction” OR “smell impairment” OR “hyposmia” OR “dysosmia” OR “olfactory dysfunction” OR “olfactory disorder”) AND ([All] “dysgeusia” OR “taste loss” OR “taste dysfunction” OR “taste impairment” OR “gustatory dysfunction” OR “gustatory disorder” OR “hypogeusia” OR “ageusia”).\n\nMoreover, we searched the preprint servers MedRxiv, BioRxiv, and Researchsquare for non-peer-reviewed articles. Data were extracted from the articles as well as supplementary materials. Reference lists from the eligible articles were retrieved for further relevant studies.\n\nThe information of identified articles was imported into EndNote X9 (Thompson Reuters, Philadelphia, PA, USA). Duplicates between databases were removed. To identify eligible studies, the retrieved articles were screened based on title and abstract. The potentially eligible studies were then fully reviewed by two authors (MF and JKF). After reviewing the full texts, the eligibility of each study was decided.\n\nInformation of study characteristics, study site, study design, number of patients with anosmia, number of patients with dysgeusia, and COVID-19 characteristics such as number of patients, severity, and outcome were collected.\n\nThe primary outcomes were: (a) the global incidence of anosmia in COVID-19 patients; (b) the global incidence of dysgeusia in COVID-19 patients; (c) the association of anosmia with the severity of COVID-19; (d) the association of dysgeusia with the severity of COVID-19; (e) the association of anosmia with mortality of COVID-19; and (f) the association of dysgeusia with mortality of COVID-19. Moreover, this review was conducted to provide the possible pathogenesis of anosmia and dysgeusia in SARS-CoV-2 infection.\n\nThe cumulative prevalence rate of anosmia and dysgeusia was calculated for COVID-19 cases by dividing the number of COVID-19 cases with anosmia by the total number of COVID-19 cases with and without anosmia, and was expressed as a percentage (%) with 95% confidence intervals (95% CI). Pooled odds ratios (OR) and 95% CI were calculated to assess the association of anosmia and the occurrence of SARS-CoV-2 compared to non-SARS-CoV-2 respiratory infections. The same method was used for dysgeusia. The pooled OR and 95% CI were presented in a forest plot.\n\nCritical assessment was conducted for the study setting and diagnosis of SARS-CoV-2 to reduce the bias. The Newcastle-Ottawa scale (NOS)26 was used as critical appraisals to assess the quality of eligible studies. Prior to analysis, gathered data from studies were evaluated for heterogeneity and potential publication bias.\n\nTo assess the association between anosmia or dysgeusia and the presence of SARS-CoV-2, Z test was performed (p < 0.05 was considered statistically significant). Q test was used to evaluate the heterogeneity among studies, and the data with heterogeneity was analyzed using a random effect model. The reporting and publication bias were assessed using Egger’s test and a funnel plot (p < 0.05 was considered having potential for publication bias). The data were analyzed using Review Manager version 5.327.\n\n\nResults\n\nIn total, 691 articles (660 reviewed articles and 31 preprint articles) were identified through the databases; of these, 182 articles were removed as duplicates. An additional 287 articles were excluded following a screening process of the titles and abstracts due to irrelevant studies, leaving 222 references (Figure 1). Full-texts of the remaining 222 references were retrieved and screened for eligibility, and this process excluded an additional 115 references as the inclusion criteria was not met. This exclusion included articles with no access28,29, RT-PCR not clearly stated in the text30-45, case reports46-97, case seriess98-113, repeated datasetss114-118, and studies in specific groups119-130. A complete assessment was conducted for 107 references.\n\nThe meta-analysis included 107 studies to calculate the prevalence of anosmia in COVID-19 patients. Additionally, 6 studies were excluded while calculating the prevalence of dysgeusia in COVID-19, thus leaving 101 eligible studies. In total, 20 and 16 studies were included to assess the association of anosmia and dysgeusia with the COVID-19 occurrence, respectively.\n\nTo calculate the prevalence of anosmia in COVID-19 cases, 107 studies were included comprising 32,142 COVID-19 patients, and anosmia was reported in 12,038 patients with a global pooled prevalence of 38.2% (95% CI: 36.5%, 47.2%). The list of the studies and the prevalence of anosmia in each study are presented in Table 1.\n\nIn total, 30,901 COVID-19 patients from 101 studies were included to calculate the prevalence of dysgeusia in COVID-19. Dysgeusia was identified in 11,337 out of 30,901 COVID-19 patients resulting in a cumulative prevalence of 36.6% (95% CI: 35.2%, 45.2%). The individual studies and the prevalence of dysgeusia from each study are listed in Table 2.\n\nThe prevalence of anosmia among COVID-19 patients around the globe.\n\nThe prevalence of dysgeusia among COVID-19 patients around the globe.\n\nIn total, 20 studies comprising 1,213 COVID-19 cases with anosmia and 2,735 non-COVID-19 patients (mostly COVID-19-like symptoms with negative RT-PCR for SARS-CoV-2) were analyzed to investigate the association between anosmia and the occurrence of COVID-19. Data suggested that anosmia was 10.2-fold more prevalent in patients with COVID-19 compared to those with COVID-19 like illness, OR 10.21 (95% CI: 6.53, 15.96) with p < 0.001 (Figure 2).\n\nIn total, 16 studies comprising 1,342 COVID-19 cases with dysgeusia and 1,990 patients with other respiratory illness (COVID-19 like illness with negative RT-PCR for SARS-CoV-2) were included to assess the association between dysgeusia and the occurrence of COVID-19. Data suggested that dysgeusia was 8.6-fold more prevalent in patients with COVID-19 compared to those with other respiratory illness, with OR 8.61 (95% CI: 5.26, 14.11) and p<0.001 (Figure 3).\n\nLimited studies have assessed the association between anosmia and dysgeusia and the severity and mortality of COVID-19 cases. One study linked anosmia with a lower fatality rate and a lower ICU admission240.\n\n\nDiscussion\n\nThe pooled prevalence of anosmia in our systematic review was 38.2% of 32,142 COVID-19 cases. This result was almost thrice the initial prevalence reported from Wuhan, China174,208. This suggests that anosmia is a potential indicator of SARS-CoV-2 infection, and may be useful for screening and early identification of COIVD-19 patients, particularly asymptomatics241. Some countries, such as the UK and US have used anosmia as an indicator for preventive measure, wherein COVID-19 patient with anosmia should commence self-isolation242-244.\n\nAnosmia is not only present in COVID-19 patients, but also in patients with other respiratory diseases such as influenza, parainfluenza, Eipstein Barr virus, picornavirus, and rhinovirus245-248. However, our study demonstrated that the prevalence of anosmia was 10.2-fold higher in COVID-19 patient than that in non-COVID-19 patient. During the previous pandemics, such as severe acute respiratory syndrome (SARS) and Middle East respiratory syndrome (MERS), anosmia was rarely reported249. Only one study reported persistent anosmia after 2 years of recovery from SARS250. Another study reported that anosmia in COVID-19 patients varied based on ethnicity; anosmia in Caucasian is three times more prevalent than in Asian population251.\n\nDysgeusia was initially reported in 11.7% of patients who were discharged from Wuhan hospital, which persisted for at least four weeks. This result was lower than ours (36.6% out of 30,901 COVID-19 cases), which might be attributable to either lower dysgeusia prevalence in China or underestimation of this symptom itself208. Moreover, the prevalence of dysgeusia in COVID-19 patients was 8.6-fold higher than that in non-COVID-19 like illness. Herpes zoster and HIV have also been linked to gustatory dysfunction252,253. Furthermore, another study reported that anosmia and dysgeusia have 82.5% predictive value for positive SARS-CoV-2 RT-PCR254.\n\nSeveral mechanisms have been proposed to explain the emergence of anosmia in COVID-19 patients.\n\na. Obstruction in the nasal airway\n\nAs several viral infections in the respiratory system display blockage of nasal airway or nasal congestion, this hypothesis was initially proposed. According to this mechanism, the interaction between the odorants and olfactory receptors is inhibited by certain obstructions, thereby impairing the subsequent smelling processes255. This condition results in anosmia. The obstruction could be caused by nasal discharge or by inflammation occurring in the nasal cavity256; however, this hypothesis can be presumably ruled out. Moreover, several studies reported that anosmia is more prevalent than nasal congestion in COVID-19 patients188,256-259. Interestingly, the incidence of rhinorrhea and nasal obstruction in SARS-CoV-2 infection is lower than other coronaviruses such as SARS-CoV and MERS-CoV260.\n\nFurthermore, presumably, nasal obstruction is a secondary mechanism by which anosmia is induced in COVID-19 patients as the obstruction in viral infection typically occurs as a subsequent event after damage in the mucociliary system, thereby inhibiting the nasal discharge and leading to nasal obstruction. In certain viral infections, the mucociliary system operated by ciliated cells is impaired. A previous study reported that human coronavirus (HCoV) disrupted the nasal ciliated respiratory epithelium leading to impaired mucociliary escalator system261.\n\nb. Damage in olfactory sensory neurons\n\nSmelling processes commence when the odorants bind to the olfactory sensory neurons (OSNs) in the olfactory epithelium located in the nasal cavity, which subsequently transmits this information through their axons to the olfactory bulb in the brain262. According to this concept, a viral attack on the receptor neurons eventually creates disturbances in the sense of smell; however, this hypothesis remains under debate as several recent studies reported the absence of angiotensin-converting enzyme 2 (ACE2) and transmembrane protease serine 2 (TMPRSS), the key factors for the virus to enter the cell263, in the OSNs264-267. These findings are supported by another study carried out by Bryche et al., who demonstrated that SARS-CoV-2 was not detected in the OSNs of hamsters266.\n\nMoreover, after comparing the duration between anosmia incidence in COVID-19 patients and the normal cellular regeneration process, this proposed mechanism should be reconsidered. Several studies reported that COVID-19-related anosmia disappeared within 1-2 weeks, whereas regeneration of dead OSNs requires more than 2 week time period188,206,255,262,268. This discrepancy results in a temporary conclusion that COVID-19-related anosmia is not directly associated with the impairment of the OSNs.\n\nc. Olfactory center damage in the brain\n\nThe aforementioned dysfunction of OSNs and the mechanism by which SARS-CoV-2 directly affects the olfactory center via axonal transport of the neuron remains unclear, as the OSN lacks ACE2 and TMPRSS2 which hinders viral entry into the cell264-267. Nevertheless, the possibility of olfactory center disruption caused by SARS-CoV-2 should not be overlooked as the cause of anosmia, since a previous study concluded that human ACE2 (hACE2)-transgenic mice suffered from brain infection after intranasal inoculation with SARS-CoV269. The study found that the brain infection commenced from the olfactory bulb, which is the axonal trajectory pathway of the OSNs269. This finding suggests that SARS-CoV-2 might also first utilize another structure in the nasal cavity before it is transported into the OSNs.\n\nd. Olfactory supporting cells dysfunction\n\nAs OSN does not express ACE2 and TMPRSS2, the virus should use another pathway to infect the olfactory system. Numerous studies have established the expression of these SARS-CoV-2 entry proteins in several supporting cells in olfactory epithelium, that is, Bowman’s gland cells, horizontal basal cells, olfactory bulb pericytes, mitral cells, sustentacular cells, and microvillar cells264-267. Of these supporting cells, the sustentacular cells have gained immense attention as the initial site of SARS-CoV-2 infection in the olfactory epithelium. In addition to their higher expression of ACE2 and TMPRSS2 than the others, sustentacular cells are located on the surface of the nasal cavity making them vulnerable to exposure to the external environment264,267.\n\nNotably, sustentacular cells act as supporting cells and promote olfactory neuron in the olfactory system. These cells detoxify harmful odorants, promote odorant-receptor binding, and provide nutritional substances to support the action of olfactory receptor neurons255,264. Considerably, it is plausible to suggest that any damage occurring in sustentacular cells will in turn affect the olfactory epithelium and produce anosmia.\n\nThe corresponding regeneration time to the recovery of anosmia also supports the notion that sustentacular cell damage relates to anosmia caused by SARS-CoV-2. As the replenishment of dead OSNs does not correspond to the duration of COVID-19-related anosmia within 1-2 weeks, the regeneration of sustentacular cells seems to be in line with that time frame264,266,268.\n\nFurthermore, this hypothesis is supported by a recent study conducted by Bryche et al., who reported that SARS-CoV-2 was accumulated in sustentacular cells but not in the OSNs266. The olfactory epithelial damage and sustentacular cell loss occurred 2 days after instilling SARS-CoV-2 intranasally in golden Syrian hamsters266.\n\ne. Inflammation-related olfactory epithelium dysfunction\n\nIt is worth noting that the cytokine storm in COVID-19 is strongly associated with organ dysfunctions, including OSNs232. The dysfunction in this structure can lead to disturbance in the sense of smell270. Torabi et al. suggested that proinflammatory cytokines, particularly tumor necrosis factor a (TNF-α), may lead to COVID-19-induced anosmia271. Another proinflammatory cytokine, interleukin-6 (IL-6), increased in cases presenting with anosmia232,272.\n\nThe mechanism used by these cytokines, in particular IL-6, to produce anosmia is not fully understood. Cazzolla et al. suggested that this effect can be caused by either peripheral or central action of the cytokines232. In the periphery, IL-6 may induce apoptosis of ciliary neuronal cells in the olfactory epithelium272, whereas in its central action, the olfactory center in the brain is attacked by the cytokine as a result of virus infection232.\n\nAlthough gustatory impairment is always displayed concomitantly with olfactory dysfunction, this symptom has a relatively different mechanism and is often distantly linked to the latter symptom. Several hypotheses have been proposed to explain the mechanism behind the emergence of dysgeusia in COVID-19 patients.\n\na. The subsequent effect of cranial nerves dysfunction\n\nConsidering the close relationship between the olfactory and gustatory system both peripherally and centrally, smell and taste dysfunction in COVID-19 often occurs concomitantly256,273. This hypothesis describes dysgeusia as a secondary event of olfactory dysfunction274; however, several studies revealed that the percentage of dysgeusia in COVID-19 patients is higher than symptoms related to olfactory dysfunction188,275. Based on this finding, another mechanism may be involved in inducing SARS-CoV-2-related dysgeusia. Furthermore, COVID-19-induced dysgeusia could also occur when there is certain damage in the cranial nerves responsible for gustatory transmission (cranial nerve VII, IX, and X)276. Among these nerves, SARS-CoV-2 exposure to cranial nerve VII has gained immense attention. Based on this hypothesis, the virus initially colonizes the nasopharynx structure, then moves to the Eustachian tube, and eventually reaches the middle ear where the virus gets access to chorda tympani and causes dysgeusia276.\n\nb. Zinc deficiency\n\nAnother interesting hypothesis underlying dysgeusia in COVID-19 is related to zinc deficiency276. This hypothesis was developed as zinc is an important mineral in carbonic anhydrase, which is pivotal in maintaining taste sensation277. Interestingly, one study reported that zinc level in patients with SARS-CoV-2 infection was significantly lower compared to that in the healthy control groups278. Alterations in the sense of taste after being treated with certain treatments, such as irradiation in cancer patients279,280, could be prevented by zinc supplementation. Moreover, dengue fever virus and human immunodeficiency virus replication could be inhibited by zinc chelation281,282. Furthermore, pharmacological agents influencing ACE2 activity are associated with taste disturbances283,284.\n\nNevertheless, this effect does not relate to zinc deficiency as these drugs do not influence both serum and salivary zinc concentrations284. Further investigation needs to be carried out to reveal the role of zinc in dysgeusia associated with COVID-19.\n\nc. SARS-CoV-2-bound sialic acid\n\nSARS-CoV-2 may produce dysgeusia via interaction with sialic acid receptors232,274,285. Sialic acid plays a pivotal role in the taste processing pathway as it is a component of the normal salivary composition286. Moreover, reduced amount of sialic acid impairs the ability to taste287. An in silico study revealed that SARS-CoV-2 could interact with the sialic acid receptor through its spike protein288. Previously, MERS-CoV was also reported to interact with this receptor289. Following this occupancy, the gustatory threshold increases, while gustatory particles degrade at a higher rate274,287.\n\nd. Direct attack on several oral sites\n\nA previous study investigated the expression of ACE2 in various tissues in the oral cavity and found that the tongue had higher ACE2 expression in comparison to other tissues, such as buccal and gingival tissues290. This finding raised a hypothesis that SARS-CoV-2 could directly attack the taste buds in the tongue, initiating inflammatory responses, and would eventually alter the sense of taste276. It is proposed that the Toll-like receptor-mediated cascade and apoptosis are the subsequent events that could lead to taste dysfunction276,291.\n\nA previous study investigating SARS-CoV infection in rhesus macaques revealed that, initially, the salivary gland was attacked by the virus292. As the human salivary gland expresses a high level of ACE2293, it is reasonable to pay more attention to the vulnerability of this gland against SARS-CoV-2 exposure. Disruption in the activity of the salivary gland would produce either imbalance in salivary composition or impairment of salivary flow, which could ultimately result in dysgeusia276.\n\n\nConclusions\n\nOut of 32,142 and 30,901 COVID-19 cases studied for anosmia and dysgeusia, respectively, the prevalence of anosmia was approximately 38.2%, whereas that of dysgeusia was 36.6%. Both of these symptoms were more common in COVID-19 compared to other respiratory infections (approximately 10 and 9 times, respectively). Several mechanisms have been proposed to explain the emergence of anosmia in COVID-19 patients including nasal airway obstruction, damage in OSNs, olfactory center damage in the brain, dysfunction of olfactory supporting cells, and inflammation-related olfactory epithelium dysfunction. Furthermore, some possible pathogenesis of dysgeusia in SARS-CoV-2 infection has been proposed including cranial nerve dysfunction, zinc deficiency, virion interaction, and direct attack of the virus to several oral sites.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\nFigshare: PRISMA checklist for ‘Anosmia and dysgeusia in SARS-CoV-2 infection: Incidence, effects on COVID-19 severity and mortality, and the possible pathobiology mechanisms - A systematic review and meta-analysis’, https://doi.org/10.6084/m9.figshare.13323080.v1294.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Acknowledgement\n\nAuthors would like to thank HT Editorial Services in assisting the writing process.\n\n\nReferences\n\nDhama K, Patel SK, Pathak M, et al.: An update on SARS-CoV-2/COVID-19 with particular reference to its clinical pathology, pathogenesis, immunopathology and mitigation strategies. Travel Med Infect Dis 2020 Sep - Oct; 37: 101755. Epub 2020/06/02. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDhama K, Khan S, Tiwari R, et al.: Coronavirus Disease 2019-COVID-19. Clin Microbiol Rev 2020 Sep 16; 33(4). Epub 2020/06/26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorldometers: COVID-19 coronavirus pandemic.Reference Source [cited 2020 November 14]\n\nGreene CJ, Burleson SL, Crosby JC, et al.: Coronavirus disease 2019: International public health considerations. JACEP Open 2020; 1(2): 70–77. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFrediansyah A, Tiwari R, Sharun K, et al.: Antivirals for COVID-19: a critical review. Clin Epidemiol Glob Health 2020. Publisher Full Text\n\nFrediansyah A, Nainu F, Dhama K, et al.: Remdesivir and its antiviral activity against COVID-19: A systematic review. Clin Epidemiol Glob Health 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMudatsir M, Yufika A, Nainu F, et al.: Antiviral Activity of Ivermectin Against SARS-CoV-2: An Old-Fashioned Dog with a New Trick—A Literature Review. Scientia Pharmaceutica 2020; 88(3): 36. Publisher Full Text\n\nRabaan AA, Al-Ahmed SH, Sah R, et al.: Recent advances in vaccine and immunotherapy for COVID-19. Hum Vaccin Immunother 2020 Nov 6: 1–12. Epub 2020/11/07 PubMed Abstract | Publisher Full Text | Free Full Text\n\nKeam S, Megawati D, Patel SK, et al.: Immunopathology and immunotherapeutic strategies in severe acute respiratory syndrome coronavirus 2 infection. Rev Med Virol 2020 Sep; 30(5): e2123. Epub 2020/07/11 PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang Y, Wang Y, Chen Y, et al.: Unique epidemiological and clinical features of the emerging 2019 novel coronavirus pneumonia (COVID-19) implicate special control measures. New Microbiol 2020; 92(6): 568–576. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRodriguez-Morales AJ, Cardona-Ospina JA, Gutiérrez-Ocampo E, et al.: Clinical, laboratory and imaging features of COVID-19: A systematic review and meta-analysis. Travel Med. Infect. Dis. 2020: 101623. Publisher Full Text\n\nMarinosci A, Landis BN, Calmy A: Possible link between anosmia and COVID-19: sniffing out the truth. Eur Arch Otorhinolaryngol Suppl 2020: 1–2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKaye R, Chang CD, Kazahaya K, et al.: COVID-19 anosmia reporting tool: initial findings. Otolaryngol Head Neck Surg 2020: 0194599820922992. PubMed Abstract | Publisher Full Text\n\nBoscolo-Rizzo P, Borsetto D, Spinato G, et al.: New onset of loss of smell or taste in household contacts of home-isolated SARS-CoV-2-positive subjects. Eur Arch Otorhinolaryngol Suppl 2020: 1–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaderno A, Schreiber A, Grammatica A, et al.: Smell and taste alterations in Covid-19: a cross-sectional analysis of different cohorts Int Forum Allergy Rhinol Wiley Online Library; 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuynh PP, Ishii LE, Ishii M: What is anosmia? Jama 2020; 324(2): 206. Publisher Full Text\n\nSamaranayake L, Fakhruddin KS, Panduwawala C: Loss of taste and smell. Br Dent J 2020; 228(11): 813. Publisher Full Text\n\nPallanti S: Importance of SARs-Cov-2 anosmia: From phenomenology to neurobiology. Comp Psych 2020: 152184. Publisher Full Text\n\nLevinson R, Elbaz M, Ben-Ami R, et al.: Anosmia and dysgeusia in patients with mild SARS-CoV-2 infection. medRxiv 2020. Publisher Full Text\n\nKlopfenstein T, Toko L, Royer PY, et al.: Features of anosmia in COVID-19. Med Mal Infect 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLaurendon T, Radulesco T, Mugnier J, et al.: Bilateral transient olfactory bulb edema during COVID-19-related anosmia. Neurology 2020; 95(5): 224–225. PubMed Abstract | Publisher Full Text\n\nLee DJ, Lockwood J, Das P, et al.: Self-reported anosmia and dysgeusia as key symptoms of COVID-19. CJEM 2020: 1–19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZayet S, Klopfenstein T, Mercier J, et al.: Contribution of anosmia and dysgeusia for diagnostic of COVID-19 in outpatients. Infection 2020: 1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHjelmesæth J, Skaare D: Loss of smell or taste as the only symptom of COVID-19. Tidsskr Nor Laegeforen 2020. PubMed Abstract | Publisher Full Text\n\nMoher D, Liberati A, Tetzlaff J, et al.: Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. Plos Med 2009 Jul; 6(7): e1000097. ISI:000268452400005. English. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStang A: Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 2010 Sep; 25(9): 603–5. Epub 2010/07/24 PubMed Abstract | Publisher Full Text\n\nCochrane T: Review Manager (RevMan) 5.3. Copenhagen: The Nordic Cochrane Centre.2008: 373.\n\nIran-Pour E, Tavabi AA, Seifi A: Presentation with Anosmia and Ageusia: Possible Hidden Carriers of COVID-19. South Med J 2020 Aug; 113(8): 399–400. Epub 2020/08/05. eng. PubMed Abstract | Publisher Full Text\n\nKanchana A, Varuvel VN: A study of COVID-19 symptoms using fuzzy cognitive map. Int J Pharm Res 2020; 12(4): 1068–1072.\n\nLiu H, Li W, Zhang L, et al.: Clinical and CT manifestations of coronavirus disease 2019 (COVID-19): comparison of suspected cases of COVID-19 in isolation and non-COVID-19 pneumonia in a single-center study conducted in Beijing, China. Res Sq 2020 2020/11/19. Publisher Full Text\n\nBagheri SH, Asghari A, Farhadi M, et al.: Coincidence of COVID-19 epidemic and olfactory dysfunction outbreak in Iran. Med J Islam Repub Iran 2020; 34(1). PubMed Abstract | Publisher Full Text | Free Full Text\n\nTung-Chen Y, Martí de Gracia M, Díez-Tascón A, et al.: Correlation between Chest Computed Tomography and Lung Ultrasonography in Patients with Coronavirus Disease 2019 (COVID-19). Ultrasound Med Biol 2020; 46(11): 2918–2926. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGilani S, Roditi R, Naraghi M: COVID-19 and anosmia in Tehran, Iran. Med Hypotheses 2020: 141. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrisca G, Ferretti M, Sartoris G, et al.: The early experiences of a single tertiary Italian emergency department treating COVID-19 in children. Acta Paediatr 2020; 109(10): 2155–2156. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDenis F, Galmiche S, Dinh A, et al.: Epidemiological Observations on the Association between Anosmia and COVID-19 Infection: Analysis of Data from a Self-Assessment Web Application. J Med Internet Res 2020; 22(6). PubMed Abstract | Publisher Full Text | Free Full Text\n\nAhmed MAM, Colebunders R, Siewe Fodjo JN: Evidence for significant COVID-19 community transmission in Somalia using a clinical case definition. Int J Infect Dis 2020; 98: 206–207. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPoncet-Megemont L, Paris P, Tronchere A, et al.: High Prevalence of Headaches During Covid-19 Infection: A Retrospective Cohort Study. Headache 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLindahl JF, Hoffman T, Esmaeilzadeh M, et al.: High seroprevalence of SARS-CoV-2 in elderly care employees in Sweden. Infect Ecol Epidemiol 2020; 10(1). PubMed Abstract | Publisher Full Text | Free Full Text\n\nGerbaud L, Guiguet-Auclair C, Breysse F, et al.: Hospital and population-based evidence for covid-19 early circulation in the east of france. Int J Environ Res Public Health 2020; 17(19): 1–17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFernandes DE, Ferreira PRA, Kirsztajn GM: Pre-exposure prophylaxis during the SARS-CoV-2 pandemic: can PrEP prevent COVID-19-related symptoms? Epidemiol. Infect. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFjaeldstad AW: Prolonged complaints of chemosensory loss after covid-19. Dan. Med. J. 2020; 67(8): 1–11.\n\nMakaronidis J, Mok J, Balogun N, et al.: Seroprevalence of SARS-CoV-2 antibodies in people with an acute loss in their sense of smell and/or taste in a community-based population in London, UK: An observational cohort study. PLoS Med 2020 Oct; 17(10): e1003358. Epub 2020/10/02. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCoelho DH, Kons ZA, Costanzo RM, et al.: Subjective Changes in Smell and Taste During the COVID-19 Pandemic: A National Survey—Preliminary Results. Otolaryngol Head Neck Surg 2020; 163(2): 302–306. PubMed Abstract | Publisher Full Text\n\nReiter ER, Coelho DH, Kons ZA, et al.: Subjective smell and taste changes during the COVID-19 pandemic: Short term recovery. Am J Otolaryngol Head Neck Med Surg 2020; 41(6). PubMed Abstract | Publisher Full Text | Free Full Text\n\nSpadera L, Viola P, Pisani D, et al.: Sudden olfactory loss as an early marker of COVID-19: a nationwide Italian survey. Eur. Arch. Otorhinolaryngol. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAguila-Gordo D, Flores-Barragan JM, Ferragut-Lloret F, et al.: Acute myelitis and SARS-CoV-2 infection. A new etiology of myelitis? J. Clin. Neurosci. 2020 Oct; 80: 280–281. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSelvaraj V, Sacchetti D, Finn A, et al.: Acute Vision Loss in a Patient with COVID-19. R I Med J (2013) 2020 Jun 10; 103(6): 37–8. Epub 2020/06/18. eng. PubMed Abstract\n\nMak PQ, Chung KS, Wong JSC, et al.: Anosmia and ageusia: Not an uncommon presentation of COVID-19 infection in children and adolescents. Pediatr. Infect. Dis. J. 2020; 39(8): E199–E200. PubMed Abstract | Publisher Full Text\n\nLi CW, Syue LS, Tsai YS, et al.: Anosmia and olfactory tract neuropathy in a case of COVID-19. J. Microbiol. Immunol. Infect. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAragão MFVV, Leal MC, Cartaxo Filho OQ, et al.: Anosmia in COVID-19 associated with injury to the olfactory bulbs evident on MRI. Am. J. Neuroradiol. 2020; 41(9): 1703–1706. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLaurendon T, Radulesco T, Mugnier J, et al.: Bilateral transient olfactory bulb edema during COVID-19-related anosmia. Neurology 2020; 95(5): 224–225. PubMed Abstract | Publisher Full Text\n\nKadono Y, Nakamura Y, Ogawa Y, et al.: A case of COVID-19 infection presenting with a seizure following severe brain edema. Seizure 2020; 80: 53–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nManiaci A, Iannella G, Vicini C, et al.: A case of covid-19 with late-onset rash and transient loss of taste and smell in a 15-year-old boy. Am J Case Rep 2020; 21: 1–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHo BE, Ho AP, Ho MA, et al.: Case report of familial COVID-19 cluster associated with High prevalence of anosmia, ageusia, and gastrointestinal symptoms. IDCases 2020: 22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCaviezel C, Weiss L, Haessig G, et al.: Case report of sequential bilateral spontaneous pneumothorax in a never-ventilated, lung-healthy COVID-19-patient. Int. J. Surg. Case Rep. 2020; 75: 441–445. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPapa A, Di Dato MT, Buonavolonta P, et al.: Clinical management of il-6 driven cytokine storm related to covid-19 in a patient with recent spinal cord stimulator implants: A case report. Reg. Anesth. Pain Med. 2020; 10(4): 1–3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLockey RF, Hudey SN: Coronavirus disease 2019-associated urticaria with angioedema in a morbidly obese man successfully treated with glucocorticoids. Reg. Anesth. Pain Med. 2020; 125(3): 359–360. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZito A, Alfonsi E, Franciotta D, et al.: COVID-19 and Guillain-Barré Syndrome: A Case Report and Review of Literature. Front. Neurol. 2020: 11. Publisher Full Text\n\nElzein F, Alsherbeeni N, Almatrafi K, et al.: COVID-19 co-infection in a patient with brucella bacteremia. Respir. Med. Case. Rep. 2020: 31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFitsiori A, Pugin D, Thieffry C, et al.: COVID-19 is Associated with an Unusual Pattern of Brain Microbleeds in Critically Ill Patients. J. Neuroimaging 2020; 30(5): 593–597. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMauro DD Md, Smimmo A, Meschini C, et al.: COVID-19 Pandemic: management of patients affected by SARSCoV-2 in Rome Covid Hospital 2 Trauma Center and safety of our surgical equipe. Res Squ 2020 2020/11/19. Publisher Full Text\n\nAksan F, Nelson EA, Swedish KA: A COVID-19 patient with intense burning pain. J Neurovirol. 2020; 26(5): 800–801. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAl-olama M, Rashid A, Garozzo D: COVID-19-associated meningoencephalitis complicated with intracranial hemorrhage: a case report. Acta Neurochir 2020; 162(7): 1495–1499. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHaroon A, Alnassani M, Aljurf M, et al.: COVID-19 post hematopoietic cell transplant, a report of 11 cases from a single center. Mediterr J Hematol Infect Dis 2020; 12(1). PubMed Abstract | Publisher Full Text | Free Full Text\n\nCoco D, Leanza S: CT scan bilateral interstitial pneumonia caused by SARS-CoV 2. Pan Afr. Med. J. 2020; 35(2): 1. Publisher Full Text\n\nTran TA, Cezar R, Frandon J, et al.: CT scan does not make a diagnosis of Covid-19: A cautionary case report. Int. J. Infect. Dis. 2020; 100: 182–183. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRein N, Haham N, Orenbuch-Harroch E, et al.: Description of 3 patients with myasthenia gravis and COVID-19. J. Neurol. Sci. 2020: 417. PubMed Abstract | Publisher Full Text\n\nRubin ES, Sansone SA, Hirshberg A, et al.: Detection of COVID-19 in a Vulvar Lesion. Am. J. Perinatol. 2020; 37(11): 1183–1184. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOrtac EE: Determination of diagnosis and disease severity, hospital and intensive care unit admission criteria in COVID-19. J Crit Intensive Care 2020; 11: 4–7. Publisher Full Text\n\nHorowitz RI, Freeman PR, Bruzzese J: Efficacy of glutathione therapy in relieving dyspnea associated with COVID-19 pneumonia: A report of 2 cases. Respir. Med. Case. Rep. 2020; 30. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDemirci Otluoglu G, Yener U, Demir MK, et al.: Encephalomyelitis associated with Covid-19 infection: case report. Br. J. Neurosurg. 2020: 1–3. PubMed Abstract | Publisher Full Text\n\nHernandez A, Muñoz P, Rojas JC, et al.: Epidemiological Chronicle of the First Recovered Coronavirus Disease Patient From Panama: Evidence of Early Cluster Transmission in a High School of Panama City. Front. Public Health 2020: 8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nValencia-Sanchez C, Wingerchuk DM: A fine balance: Immunosuppression and immunotherapy in a patient with multiple sclerosis and COVID-19. Mult. Scler. Relat. Disord. 2020: 42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nScheidl E, Canseco DD, Hadji-Naumov A, et al.: Guillain-Barré syndrome during SARS-CoV-2 pandemic: A case report and review of recent literature. J. Peripher. Nerv. Syst. 2020; 25(2): 204–207. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDalakas MC: Guillain-Barre syndrome: The first documented COVID-19-triggered autoimmune neurologic disease: More to come with myositis in the offing. Neurol Neuroimmunol Neuroinflamm 2020 Sep; 7(5). PubMed Abstract | Publisher Full Text | Free Full Text\n\nde Oliveira FAA, Palmeira DCC, Rocha-Filho PAS: Headache and pleocytosis in CSF associated with COVID-19: case report. Neurol. Sci. 2020; 41(11): 3021–2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMelley LE, Bress E, Polan E: Hypogeusia as the initial presenting symptom of COVID-19. BMJ Case Reports 2020; 13(5). PubMed Abstract | Publisher Full Text | Free Full Text\n\nPallanti S: Importance of SARs-Cov-2 anosmia: From phenomenology to neurobiology. Compr Psychiatry 2020 Jul; 100: 152184. Epub 2020/05/19. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNaz S, Hanif M, Haider MA, et al.: Meningitis as an Initial Presentation of COVID-19: A Case Report. Front. Public Health 2020: 8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRay A: Miller Fisher syndrome and COVID-19: Is there a link. BMJ Case Reports 2020; 13(8). PubMed Abstract | Publisher Full Text | Free Full Text\n\nGutiérrez-Ortiz C, Méndez-Guerrero A, Rodrigo-Rey S, et al.: Miller Fisher syndrome and polyneuritis cranialis in COVID-19. Neurology 2020; 95(5): e601–e605. PubMed Abstract | Publisher Full Text\n\nPalao M, Fernández-Díaz E, Gracia-Gil J, et al.: Multiple sclerosis following SARS-CoV-2 infection. Mult Scler Relat Disord 2020 Oct; 45: 102377. Epub 2020/07/23. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLim WS, Liang CK, Assantachai P, et al.: COVID-19 and Older People in Asia: AWGS Calls to Actions. Geriatr Gerontol Int 2020 May 4. Epub 2020/05/05. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHatipoglu N, Mine Yazici Z, Palabiyik F, et al.: Olfactory bulb magnetic resonance imaging in SARS-CoV-2-induced anosmia in pediatric cases. Int. J. Pediatr. Otorhinolaryngol. 2020: 139. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSinadinos A, Shelswell J: Oral ulceration and blistering in patients with COVID-19. Evid. Based Dent. 2020; 21(2): 49. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuber M, Rogozinski S, Puppe W, et al.: Postinfectious Onset of Myasthenia Gravis in a COVID-19 Patient. Front. Neurol. 2020: 11. Publisher Full Text\n\nBaba H, Kanamori H, Oshima K, et al.: Prolonged presence of SARS-CoV-2 in a COVID-19 case with rheumatoid arthritis taking iguratimod treated with ciclesonide. J. Infect. Chemother. 2020; 26(10): 1100–1103. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZanin L, Saraceno G, Panciani PP, et al.: SARS-CoV-2 can induce brain and spine demyelinating lesions. Acta Neurochirurgica 2020; 162(7): 1491–1494. PubMed Abstract | Publisher Full Text | Free Full Text\n\nde Brito CAA, Lima PMA, de Brito MCM, et al.: Second episode of COVID-19 in health professionals: Report of two cases. International Medical Case Reports Journal 2020; 13: 471–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKasuga Y, Nishimura K, Go H, et al.: Severe olfactory and gustatory dysfunctions in a Japanese pediatric patient with coronavirus disease (COVID-19). J. Infect. Chemother. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen C, Chen M, Cheng C, et al.: A special symptom of olfactory dysfunction in coronavirus disease 2019: report of three cases. J Neurovirol. 2020; 26(3): 456–458. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaoli D, Pallotti F, Colangelo S, et al.: Study of SARS-CoV-2 in semen and urine samples of a volunteer with positive naso-pharyngeal swab. J. Endocrinol. Investig. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGiné C, Laín A, García L, et al.: Thoracoscopic bullectomy for persistent air leak in a 14-year-old child with COVID-19 bilateral pulmonary disease. J. Laparoendosc. Adv. Surg. Tech. A 2020; 30(8): 935–938. PubMed Abstract | Publisher Full Text\n\nRivas-Pollmar MI, Álvarez-Román MT, Butta-Coll NV, et al.: Thromboprophylaxis in a patient with COVID-19 and severe hemophilia A on emicizumab prophylaxis. J. Thromb. Haemost. 2020; 18(9): 2202–2204. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarchand L, Pecquet M, Luyton C: Type 1 diabetes onset triggered by COVID-19. Acta Diabetol. 2020; 57(10): 1265–1266. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGlick LR, Fogel AL, Ramachandran S, et al.: Unilateral laterothoracic exanthem in association with coronavirus disease 2019. JAAD Case Rep 2020; 6(9): 900–901. PubMed Abstract | Publisher Full Text | Free Full Text\n\nde Sousa GC, de Sousa TC, Sakiyama MAK, et al.: Vasculitis-related stroke in young as a presenting feature of novel coronavirus disease (COVID19) - Case report. J. Clin. Neurosci. 2020; 79: 169–71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVargas-Gandica J, Winter D, Schnippe R, et al.: Ageusia and anosmia, a common sign of COVID-19? A case series from four countries. J Neurovirol. 2020; 26(5): 785–789. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLechner M, Chandrasekharan D, Jumani K, et al.: Anosmia as a presenting symptom of SARS-CoV-2 infection in healthcare workers - A systematic review of the literature, case series, and recommendations for clinical assessment and management. Rhinology 2020; 58(4): 1–9. PubMed Abstract | Publisher Full Text\n\nToptan T, Aktan Ç, Basari A, et al.: Case Series of Headache Characteristics in COVID-19: Headache Can Be an Isolated Symptom. Headache 2020; 60(8): 1788–1792. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKilic O, Kalcioglu MT, Cag Y, et al.: Could sudden sensorineural hearing loss be the sole manifestation of COVID-19? An investigation into SARS-COV-2 in the etiology of sudden sensorineural hearing loss. Int. J. Infect. Dis. 2020; 97: 208–211. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBeach SR, Praschan NC, Hogan C, et al.: Delirium in COVID-19: A case series and exploration of potential mechanisms for central nervous system involvement. Gen. Hosp. Psychiatry 2020; 65: 47–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRichtmann R, Torloni MR, Oyamada Otani AR, et al.: Fetal deaths in pregnancies with SARS-CoV-2 infection in Brazil: A case series. Case Rep Womens Health 2020: 27. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLechien JR, Cabaraux P, Chiesa-Estomba CM, et al.: Objective olfactory evaluation of self-reported loss of smell in a case series of 86 COVID-19 patients. Head Neck 2020; 42(7): 1583–1590. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrandão TB, Gueiros LA, Melo TS, et al.: Oral lesions in patients with SARS-CoV-2 infection: could the oral cavity be a target organ? Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFerrero P, Piazza I, Bonino C, et al.: Patterns of myocardial involvement in children during COVID-19 pandemic: Early experience from northern Italy. Ann. Pediatr. Cardiol. 2020; 13(3): 230–3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu PY, Jiang RS: Prognosis of olfactory and gustatory dysfunctions in COVID-19 patients: A case series. Clin Case Rep 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSisó-Almirall A, Kostov B, Mas-Heredia M, et al.: Prognostic factors in Spanish COVID-19 patients: A case series from Barcelona. PLoS ONE 2020; 15(8 August 2020). PubMed Abstract | Publisher Full Text | Free Full Text\n\nRudberg AS, Havervall S, Månberg A, et al.: SARS-CoV-2 exposure, symptoms and seroprevalence in healthcare workers in Sweden. Nat Commun 2020 Oct 8; 11(1): 5064. Epub 2020/10/10. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZurita MF, Iglesias Arreaga A, Luzuriaga Chavez AA, et al.: SARS-CoV-2 Infection and COVID-19 in 5 Patients in Ecuador After Prior Treatment with Hydroxychloroquine for Systemic Lupus Erythematosus. Am J Case Rep 2020 Sep 26; 21: e927304. Epub 2020/09/27. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaggiolo F, Zoboli F, Arosio M, et al.: SARS-CoV-2 infection in persons living with HIV: A single center prospective cohort. J. Med. Virol. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSeo MY, Seok H, Hwang SJ, et al.: Trend of Olfactory and Gustatory Dysfunction in COVID-19 Patients in a Quarantine Facility. J. Korean Med. Sci. 2020; 35(41): e375. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarzano AV, Genovese G, Fabbrocini G, et al.: Varicella-like exanthem as a specific COVID-19-associated skin manifestation: Multicenter case series of 22 patients. J. Am. Acad. Dermatol. 2020; 83(1): 280–285. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVaira LA, Lechien JR, Salzano G, et al.: Gustatory Dysfunction: A Highly Specific and Smell-Independent Symptom of COVID-19. Indian J Otolaryngol Head Neck Surg 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoscolo-Rizzo P, Borsetto D, Spinato G, et al.: New onset of loss of smell or taste in household contacts of home-isolated SARS-CoV-2 positive subjects. Res Sq 2020 2020/11/19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVaira LA, Hopkins C, Salzano G, et al.: Olfactory and gustatory function impairment in COVID-19 patients: Italian objective multicenter-study. Head Neck 2020 Jul; 42(7): 1560–9. WOS:000534373200001 PubMed Abstract | Publisher Full Text | Free Full Text\n\nPetrocelli M, Ruggiero F, Baietti AM, et al.: Remote psychophysical evaluation of olfactory and gustatory functions in early-stage coronavirus disease 2019 patients: The Bologna experience of 300 cases. J. Laryngol. Otol. 2020; 134(7): 571–576. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVaira LA, Salzano G, Petrocelli M, et al.: Validation of a self-administered olfactory and gustatory test for the remotely evaluation of COVID-19 patients in home quarantine. Head Neck 2020; 42(7): 1570–1576. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSahin D, Tanacan A, Erol SA, et al.: A pandemic center’s experience of managing pregnant women with COVID-19 infection in Turkey: A prospective cohort study. Int. J. Gynaecol. Obstet. 2020; 151(1): 74–82. PubMed Abstract | Publisher Full Text\n\nPouletty M, Borocco C, Ouldali N, et al.: Paediatric multisystem inflammatory syndrome temporally associated with SARS-CoV-2 mimicking Kawasaki disease (Kawa-COVID-19): A multicentre cohort. Ann. Rheum. Dis. 2020; 79(8): 999–1006. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaolo G: Does COVID-19 cause permanent damage to olfactory and gustatory function? Med Hypotheses 2020 Oct; 143: 110086. Epub 2020/07/30. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKarimi-Galougahi M, Safavi Naini A, Ghorbani J, et al.: Emergence and Evolution of Olfactory and Gustatory Symptoms in Patients with COVID-19 in the Outpatient Setting. Indian J Otolaryngol Head Neck Surg 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPapa ND, Sambataro G, Minniti A, et al.: Impact of COVID-19 outbreak in an Italian cohort of patients with systemic sclerosis. Ther Adv Musculoskelet Dis 2020: 12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKosugi EM, Lavinsky J, Romano FR, et al.: Incomplete and late recovery of sudden olfactory dysfunction in COVID-19. Braz. J. Otorhinolaryngol. 2020; 86(4): 490–496. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIslamoglu Y, Gemcioglu E, Ates I: Objective evaluation of the nasal mucosal secretion in COVID-19 patients with anosmia. Ir. J. Med. Sci. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLechien J, Cabaraux P, Chiesa-Estomba C, et al.: Objective olfactory testing in patients presenting with sudden onset olfactory dysfunction as the first manifestation of confirmed COVID-19 infection. medRxiv 2020: 2020.04.15.20066472. Publisher Full Text\n\nKandemirli SG, Altundag A, Yildirim D, et al.: Olfactory Bulb MRI and Paranasal Sinus CT Findings in Persistent COVID-19 Anosmia. Acad. Radiol. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStrauss SB, Lantos JE, Heier LA, et al.: Olfactory bulb signal abnormality in patients with COVID-19 who present with neurologic symptoms. Am. J. Neuroradiol. 2020; 41(10): 1882–1887. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNaeini AS, Karimi-Galougahi M, Raad N, et al.: Paranasal sinuses computed tomography findings in anosmia of COVID-19. Am J Otolaryngol Head Neck Med Surg 2020; 41(6). PubMed Abstract | Publisher Full Text\n\nPieruzzini R, Ayala C, Navas J, et al.: PREDICTIVE VALUE OF SMELL AND TASTE TEST VS PCR-RT SARS-COV-2 AND RAPID DIAGNOSTIC TESTS IN THE DIAGNOSIS OF INFECTION BY COVID-19. A PROSPECTIVE MULTI-CENTRIC STUDY. medRxiv 2020: 2020.08.31.20185298. Publisher Full Text\n\nTan JY, Sim XYJ, Wee LE, et al.: A comparative study on the clinical features of COVID-19 with non-SARS-CoV-2 respiratory viral infections. J. Med. Virol. 2020. PubMed Abstract | Publisher Full Text\n\nPeyrony O, Marbeuf-Gueye C, Truong V, et al.: Accuracy of Emergency Department Clinical Findings for Diagnosis of Coronavirus Disease 2019. Ann. Emerg. Med. 2020; 76(4): 405–412. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBeltrán-Corbellini Á, Chico-García JL, Martínez-Poles J, et al.: Acute-onset smell and taste disorders in the context of COVID-19: a pilot multicentre polymerase chain reaction based case-control study. Eur. J. Neurol. 2020; 27(9): 1738–1741. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu JY, Chen TJ, Hwang SJ: Analysis of imported cases of covid-19 in taiwan: A nationwide study. Int. J. Environ. Res. Public Health 2020; 17(9). PubMed Abstract | Publisher Full Text | Free Full Text\n\nCarignan A, Valiquette L, Grenier C, et al.: Anosmia and dysgeusia associated with SARS-CoV-2 infection: an age-matched case-control study. Cmaj 2020 Jun 29; 192(26): E702–e7. Epub 2020/05/29. eng. PubMed Abstract | Publisher Full Text\n\nChen A, Agarwal A, Ravindran N, et al.: Are Gastrointestinal Symptoms Specific for Coronavirus 2019 Infection? A Prospective Case-Control Study From the United States. Gastroenterology 2020; 159(3): 1161–3.e2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLuigetti M, Iorio R, Bentivoglio AR, et al.: Assessment of neurological manifestations in hospitalized patients with COVID-19. Eur. J. Neurol. 2020; 27(11): 2322–2328. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYan CH, Faraji F, Prajapati DP, et al.: Association of chemosensory dysfunction and COVID-19 in patients presenting with influenza-like symptoms. Int Forum Allergy Rhinol 2020; 10(7): 806–813. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRojas-Lechuga MJ, Izquierdo-Domínguez A, Chiesa-Estomba C, et al.: Chemosensory dysfunction in COVID-19 out-patients. Eur. Arch. Otorhinolaryngol. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNeto DB, Fornazieri MA, Dib C, et al.: Chemosensory Dysfunction in COVID-19: Prevalences, Recovery Rates, and Clinical Associations on a Large Brazilian Sample. Otolaryngol. Head Neck Surg. WOS:000565269500001. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDurrani M, Haq IU, Kalsoom U, et al.: Chest x-rays findings in covid 19 patients at a university teaching hospital-a descriptive study. Pak J Med Sci 2020; 36(COVID19-S4): S22–S6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGalván Casas C, Català A, Carretero Hernández G, et al.: Classification of the cutaneous manifestations of COVID-19: a rapid prospective nationwide consensus study in Spain with 375 cases. Br. J. Dermatol. 2020; 183(1): 71–77. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLechien JR, Chiesa-Estomba CM, Place S, et al.: Clinical and epidemiological characteristics of 1420 European patients with mild-to-moderate coronavirus disease 2019. J. Intern. Med. 2020; 288(3): 335–344. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim GU, Kim MJ, Ra SH, et al.: Clinical characteristics of asymptomatic and symptomatic patients with mild COVID-19. Clin. Microbiol. Infect. 2020 Jul; 26(7). WOS:000544273300033. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLapostolle F, Schneider E, Vianu I, et al.: Clinical features of 1487 COVID-19 patients with outpatient management in the Greater Paris: the COVID-call study. Intern. Emerg. Med. 2020; 15(5): 813–817. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCorsini Campioli C, Cano Cevallos E, Assi M, et al.: Clinical predictors and timing of cessation of viral RNA shedding in patients with COVID-19. J Clin Virol 2020 Sep; 130: 104577. Epub 2020/08/11. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFerreli F, Gaino F, Russo E, et al.: Clinical presentation at the onset of COVID-19 and allergic rhinoconjunctivitis. J Allergy Clin Immunol Pract 2020; 8(10): 3587–3589. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRajkumar I, Anand KH, Revathishree K, et al.: Contemporary Analysis of Olfactory Dysfunction in Mild to Moderate Covid 19 Patients in A Tertiary Health Care Centre. Indian J Otolaryngol Head Neck Surg 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang TZ, Sell J, Weiss D, et al.: COVID-19 presenting as anosmia and dysgeusia in New York City emergency departments, March - April, 2020. medRxiv 2020: 2020.07.06.20147751. Publisher Full Text\n\nCho RHW, To ZWH, Yeung ZWC, et al.: COVID-19 Viral Load in the Severity of and Recovery From Olfactory and Gustatory Dysfunction. Laryngoscope 2020; 130(11): 2680–2685. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKadiane-Oussou NJ, Klopfenstein T, Royer PY, et al.: COVID-19: comparative clinical features and outcome in 114 patients with or without pneumonia (Nord Franche-Comte Hospital, France). Microbes Infect. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNakanishi H, Suzuki M, Maeda H, et al.: Differential Diagnosis of COVID-19: Importance of Measuring Blood Lymphocytes, Serum Electrolytes, and Olfactory and Taste Functions. Tohoku J. Exp. Med. 2020; 252(2): 109–119. PubMed Abstract | Publisher Full Text\n\nSheng WH, Liu WD, Wang JT, et al.: Dysosmia and dysgeusia in patients with COVID-19 in northern Taiwan. J. Formos. Med. Assoc. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSakalli E, Temirbekov D, Bayri E, et al.: Ear nose throat-related symptoms with a focus on loss of smell and/or taste in COVID-19 patients. Am J Otolaryngol Head Neck Med Surg 2020; 41(6). PubMed Abstract | Publisher Full Text | Free Full Text\n\nLagi F, Piccica M, Graziani L, et al.: Early experience of an infectious and tropical diseases unit during the coronavirus disease (COVID-19) pandemic, Florence, Italy, February to March 2020. Euro Surveill. 2020; 25(17). PubMed Abstract | Publisher Full Text | Free Full Text\n\nVacchiano V, Riguzzi P, Volpi L, et al.: Early neurological manifestations of hospitalized COVID-19 patients. Neurol. Sci. 2020; 41(8): 2029–2031. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAmer MA, Elsherif HS, Abdel-Hamid AS, et al.: Early recovery patterns of olfactory disorders in COVID-19 patients; a clinical cohort study. Am J Otolaryngol Head Neck Med Surg 2020; 41(6). PubMed Abstract | Publisher Full Text | Free Full Text\n\nOmbajo LA, Mutono N, Sudi P, et al.: Epidemiological And Clinical Characteristics Of Covid-19 Patients In Kenya. medRxiv 2020: 2020.11.09.20228106. Publisher Full Text\n\nMaechler F, Gertler M, Hermes J, et al.: Epidemiological and clinical characteristics of SARS-CoV-2 infections at a testing site in Berlin, Germany, March and April 2020—a cross-sectional study. Clin. Microbiol. Infect. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHussain MH, Mair M, Rea P: Epistaxis as a marker for severe acute respiratory syndrome coronavirus-2 status - A prospective study. J. Laryngol. Otol. 2020; 134(8): 717–720. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShah NN, Hussain RT, Mustafa H, et al.: Evaluation of Olfactory Acuity in Patients with Coronavirus Disease 2019 (COVID-19). Indian J of Otolaryngol Head Neck Surg 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGorzkowski V, Bevilacqua S, Charmillon A, et al.: Evolution of Olfactory Disorders in COVID-19 Patients. Laryngoscope 2020; 130(11): 2667–2673. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKlopfenstein T, Kadiane-Oussou NJ, Toko L, et al.: Features of anosmia in COVID-19. Med. Mal. Infect. 2020; 50(5): 436–439. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJalessi M, Barati M, Rohani M, et al.: Frequency and outcome of olfactory impairment and sinonasal involvement in hospitalized patients with COVID-19. Neurol. Sci. 2020; 41(9): 2331–2338. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiotta EM, Batra A, Clark JR, et al.: Frequent neurologic manifestations and encephalopathy-associated morbidity in Covid-19 patients. Ann. Clin. Transl. Neurol. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMembrilla JA, de Lorenzo Í, Sastre M: Headache as a Cardinal Symptom of Coronavirus Disease 2019: A Cross-Sectional Study. Headache 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRocha-Filho PAS, Magalhães JE: Headache associated with COVID-19: Frequency, characteristics and association with anosmia and ageusia. Cephalalgia 2020; 40(13): 1443–1451. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUygun O, Ertas M, Ekizoglu E, et al.: Headache characteristics in COVID-19 pandemic-a survey study. J. Headache Pain 2020 Oct; 21(1). WOS:000579589000001. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAssaad S, Avrillon V, Fournier ML, et al.: High mortality rate in cancer patients with symptoms of COVID-19 with or without detectable SARS-COV-2 on RT-PCR. Eur. J. Cancer 2020; 135: 251–259. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMenni C, Valdes A, Freydin MB, et al.: Loss of smell and taste in combination with other symptoms is a strong predictor of COVID-19 infection. medRxiv 2020: 2020.04.05.20048421. Publisher Full Text\n\nMohamud MFY, Mohamed YG, Ali AM, et al.: Loss of taste and smell are common clinical characteristics of patients with COVID-19 in somalia: A retrospective double centre study. Infect Drug Resist 2020; 13: 2631–2635. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDawson P, Rabold EM, Laws RL, et al.: Loss of Taste and Smell as Distinguishing Symptoms of COVID-19. medRxiv 2020: 2020.05.13.20101006. PubMed Abstract | Publisher Full Text | Free Full Text\n\nÇalica Utku A, Budak G, Karabay O, et al.: Main symptoms in patients presenting in the COVID-19 period. Scott. Med. J. 2020; 65(4): 127–132. PubMed Abstract | Publisher Full Text\n\nMao L, Jin HJ, Wang MD, et al.: Neurologic Manifestations of Hospitalized Patients With Coronavirus Disease 2019 in Wuhan, China. JAMA Neurol. 2020 Jun; 77(6): 683–90. WOS:000542138800006. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStudart-Neto A, Guedes BF, de Luca e Tuma R, et al.: Neurological consultations and diagnoses in a large, dedicated COVID-19 university hospital. Arq. Neuropsiquiatr. 2020; 78(8): 494–500. PubMed Abstract | Publisher Full Text\n\nPinna P, Grewal P, Hall JP, et al.: Neurological manifestations and COVID-19: Experiences from a tertiary care center at the Frontline. J. Neurol. Sci. 2020: 415. PubMed Abstract | Publisher Full Text\n\nGarg R, Jain R, Sodani A, et al.: Neurological symptoms as initial manifestation of Covid-19-An observational study. Ann. Indian Acad. Neurol. 2020; 23(4): 482–486. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiang YJ, Xu JB, Chu M, et al.: Neurosensory dysfunction: A diagnostic marker of early COVID-19. Int. J. Infect. Dis. 2020 Sep; 98: 347–52. WOS:000569065400020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKlopfenstein T, Zahra H, Kadiane-Oussou NJ, et al.: New loss of smell and taste: Uncommon symptoms in COVID-19 patients in Nord Franche-Comte cluster, France. Int J Infect Dis 2020; 100: 117–122. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoscolo-Rizzo P, Borsetto D, Spinato G, et al.: New onset of loss of smell or taste in household contacts of home-isolated SARS-CoV-2-positive subjects. Eur. Arch. Otorhinolaryngol. 2020; 277(9): 2637–2640. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPatel A, Charani E, Ariyanayagam D, et al.: New-onset anosmia and ageusia in adult patients diagnosed with SARS-CoV-2 infection. Clin. Microbiol. Infect. 2020; 26(9): 1236–1241. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVaira LA, Deiana G, Fois AG, et al.: Objective evaluation of anosmia and ageusia in COVID-19 patients: Single-center experience on 72 cases. Head Neck 2020 Jun; 42(6): 1252–8. WOS:000529255000001. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLechien JR, Ducarme M, Place S, et al.: Objective olfactory findings in hospitalized severe COVID-19 patients. Pathogens 2020; 9(8): 1–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShemer A, Einan-Lifshitz A, Itah A, et al.: Ocular involvement in coronavirus disease 2019 (COVID-19): a clinical and molecular analysis. Int. Ophthalmol. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAltin F, Cingi C, Uzun T, et al.: Olfactory and gustatory abnormalities in COVID-19 cases. Eur. Arch. Otorhinolaryngol. 2020; 277(10): 2775–2781. PubMed Abstract | Publisher Full Text | Free Full Text\n\nQiu CH, Cui C, Hautefort C, et al.: Olfactory and Gustatory Dysfunction as an Early Identifier of COVID-19 in Adults and Children: An International Multicenter Study. Otolaryngol. Head Neck Surg. 2020 Oct; 163(4): 714–21. WOS:000542267700001. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRamasamy K, Saniasiaya J, Abdul GN: Olfactory and Gustatory Dysfunctions as a Clinical Manifestation of Coronavirus Disease 2019 in a Malaysian Tertiary Center Ann. Otol. Rhinol. Laryngol. 2020. PubMed Abstract | Publisher Full Text\n\nLechien JR, Chiesa-Estomba CM, De Siati DR, et al.: Olfactory and gustatory dysfunctions as a clinical presentation of mild-to-moderate forms of the coronavirus disease (COVID-19): a multicenter European study. Eur. Arch. Otorhinolaryngol. 2020 2020/08/01; 277(8): 2251–61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMeini S, Suardi LR, Busoni M, et al.: Olfactory and gustatory dysfunctions in 100 patients hospitalized for COVID-19: sex differences and recovery time in real-life. Eur. Arch. Otorhinolaryngol. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaderno A, Mattavelli D, Rampinelli V, et al.: Olfactory and Gustatory Outcomes in COVID-19: A Prospective Evaluation in Nonhospitalized Subjects. Otolaryngol. Head Neck Surg. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSpeth MM, Singer-Cornelius T, Oberle M, et al.: Olfactory Dysfunction and Sinonasal Symptomatology in COVID-19: Prevalence, Severity, Timing, and Associated Characteristics. Otolaryngol. Head Neck Surg. 2020; 163(1): 114–120. PubMed Abstract | Publisher Full Text | Free Full Text\n\nD’Ascanio L, Pandolfini M, Cingolani C, et al.: Olfactory Dysfunction in COVID-19 Patients: Prevalence and Prognosis for Recovering Sense of Smell. Otolaryngol. Head Neck Surg. 2020. PubMed Abstract | Publisher Full Text\n\nOtte MS, Eckel HNC, Poluschkin L, et al.: Olfactory dysfunction in patients after recovering from COVID-19. Acta Otolaryngol. 2020. PubMed Abstract | Publisher Full Text\n\nKlein H, Asseo K, Karni N, et al.: Onset, duration, and persistence of taste and smell changes and other COVID-19 symptoms: longitudinal study in Israeli patients. medRxiv 2020: 2020.09.25.20201343. PubMed Abstract\n\nPanda S, Mohamed A, Sikka K, et al.: Otolaryngologic Manifestation and Long-Term Outcome in Mild COVID-19: Experience from a Tertiary Care Centre in India. Indian J Otolaryngol Head Neck Surg 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nÖzçelik Korkmaz M, Egilmez OK, Özçelik MA, et al.: Otolaryngological manifestations of hospitalised patients with confirmed COVID-19 infection. Eur. Arch. Otorhinolaryngol. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nElibol E: Otolaryngological symptoms in COVID-19. Eur. Arch. Otorhinolaryngol. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLagier JC, Million M, Gautret P, et al.: Outcomes of 3,737 COVID-19 patients treated with hydroxychloroquine/azithromycin and other regimens in Marseille, France: A retrospective analysis. Travel Med. Infect. Dis. 2020: 36. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUhm JS, Ahn JY, Hyun J, et al.: Patterns of viral clearance in the natural course of asymptomatic COVID-19: Comparison with symptomatic non-severe COVID-19. Int. J. Infect. Dis. 2020; 99: 279–285. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYan CH, Prajapati DP, Ritter ML, et al.: Persistent Smell Loss Following Undetectable SARS-CoV-2. Otolaryngol. Head Neck Surg. 2020; 163(5): 923–925. PubMed Abstract | Publisher Full Text\n\nLópez de la Iglesia J, Fernández-Villa T, Rivero A, et al.: Predictive factors of COVID-19 in patients with negative RT-qPCR. Semergen 2020; 46: 6–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHaehner A, Draf J, Dräger S, et al.: Predictive Value of Sudden Olfactory Loss in the Diagnosis of COVID-19. ORL 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nO’Keefe JB, Tong EJ, Datoo O’Keefe GA, et al.: Predictors of disease duration and symptom course of outpatients with acute covid-19: a retrospective cohort study. medRxiv 2020: 2020.06.05.20123471. Publisher Full Text\n\nPinato DJ, Lee AJX, Biello F, et al.: Presenting features and early mortality from SARS-CoV-2 infection in cancer patients during the initial stage of the COVID-19 pandemic in Europe. Cancers 2020; 12(7): 1–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRedd WD, Zhou JC, Hathorn KE, et al.: Prevalence and Characteristics of Gastrointestinal Symptoms in Patients With Severe Acute Respiratory Syndrome Coronavirus 2 Infection in the United States: A Multicenter Cohort Study. Gastroenterology 2020; 159(2): 765–7.e2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee Y, Min P, Lee S, et al.: Prevalence and Duration of Acute Loss of Smell or Taste in COVID-19 Patients. J Korean Med Sci 2020 May 11; 35(18): e174. Epub 2020/05/10. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChary E, Carsuzaa F, Trijolet JP, et al.: Prevalence and Recovery From Olfactory and Gustatory Dysfunctions in Covid-19 Infection: A Prospective Multicenter Study. Am. J. Rhinol. Allergy 2020; 34(5): 686–693. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLv H, Zhang W, Zhu Z, et al.: Prevalence and recovery time of olfactory and gustatory dysfunction in hospitalized patients with COVID-19 in Wuhan. China. Int. J. Infect. Dis. 2020; 100: 507–512. Publisher Full Text\n\nMoein ST, Hashemian SM, Tabarsi P, et al.: Prevalence and reversibility of smell dysfunction measured psychophysically in a cohort of COVID-19 patients. Int Forum Allergy Rhinol 2020; 10(10): 1127–1135. Publisher Full Text\n\nAl-Ani RM, Acharya D: Prevalence of Anosmia and Ageusia in Patients with COVID-19 at a Primary Health Center, Doha, Qatar. Indian J Otolaryngol Head Neck Surg 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVivek Kumar P, Rohit S, Pradeepti N, et al.: Prevalence Of Anosmia And Dysgeusia In Patients Of COVID-19 In A Dedicated Covid Hospital. Res Sq 2020 2020/11/19. Publisher Full Text\n\nNouchi A, Chastang J, Miyara M, et al.: Prevalence of hyposmia and hypogeusia in 390 COVID-19 hospitalized patients and outpatients: a cross-sectional study. Eur. J. Clin. Microbiol. Infect. Dis. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMishra P, Gowda V, Dixit S, et al.: Prevalence of New Onset Anosmia in COVID-19 Patients: Is The Trend Different Between European and Indian Population? Indian J Otolaryngol Head Neck Surg 2020: 72(4): 484–487. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAvcl H, Karabulut B, Farasoglu A, et al.: Relationship between anosmia and hospitalisation in patients with coronavirus disease 2019: An otolaryngological perspective. J. Laryngol. Otol. 2020; 134(8): 710–716. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKarni N, Klein H, Asseo K, et al.: Self-rated smell ability enables highly specific predictors of COVID-19 status: a case control study in Israel. medRxiv 2020: 2020.07.30.20164327. Publisher Full Text\n\nLee DJ, Lockwood J, Das P, et al.: Self-reported anosmia and dysgeusia as key symptoms of coronavirus disease 2019. Cjem 2020 Sep; 22(5): 595–602. Epub 2020/06/09. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYan CRH, Faraji F, Prajapati DP, et al.: Self-reported olfactory loss associates with outpatient clinical course in COVID-19. Int Forum Allergy Rhinol 2020 Jul; 10(7): 821–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSong J, Deng Y-K, Wang H, et al.: Self-reported taste and smell disorders in patients with COVID-19: distinct features in China. medRxiv 2020: 2020.06.12.20128298. Publisher Full Text\n\nDeBiasi RL, Song X, Delaney M, et al.: Severe Coronavirus Disease-2019 in Children and Young Adults in the Washington, DC, Metropolitan Region. J. Pediatr. 2020; 223: 199–203.e1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKonstantinidis I, Delides A, Tsakiropoulou E, et al.: Short-term follow-up of self-isolated covid-19 patients with smell and taste dysfunction in Greece: Two phenotypes of recovery. ORL 2020. Publisher Full Text\n\nAlshami A, Alattas R, Anan H, et al.: Silent disease and loss of taste and smell are common manifestations of SARS-COV-2 infection in a quarantine facility: Saudi Arabia. PLoS One 2020; 15(10): e0241258. Epub 2020/10/31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaderno A, Schreiber A, Grammatica A, et al.: Smell and taste alterations in COVID-19: a cross-sectional analysis of different cohorts. Int Forum Allergy Rhinol 2020; 10(8): 955–962. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDell’Era V, Farri F, Garzaro G, et al.: Smell and taste disorders during COVID-19 outbreak: Cross-sectional study on 355 patients. Head Neck 2020; 42(7): 1591–1596. Publisher Full Text\n\nBarón-Sánchez J, Santiago C, Goizueta-San Martín G, et al.: Smell and taste disorders in Spanish patients with mild COVID-19. Neurologia 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIzquierdo-Domínguez A, Rojas-Lechuga MJ, Chiesa-Estomba C, et al.: Smell and taste dysfunction in covid-19 is associated with younger age in ambulatory settings: A multicenter cross-sectional study. J. Investig. Allergol. Clin. Immunol. 2020; 30(5): 346–357. PubMed Abstract | Publisher Full Text\n\nVaira LA, Hopkins C, Petrocelli M, et al.: Smell and taste recovery in coronavirus disease 2019 patients: A 60-day objective and prospective study. J. Laryngol. Otol. 2020; 134(8): 703–709. Publisher Full Text\n\nLima MA, Silva MTT, Oliveira RV, et al.: Smell dysfunction in COVID-19 patients: More than a yes-no question. J. Neurosurg. Sci. 2020: 418. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoein ST, Hashemian SM, Mansourafshar B, et al.: Smell dysfunction: a biomarker for COVID-19. Int Forum Allergy Rhinol 2020; 10(8): 944–950. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFoster KJ, Jauregui E, Tajudeen B, et al.: Smell loss is a prognostic factor for lower severity of coronavirus disease 2019. Ann. Allergy Asthma Immunol. 2020; 125(4): 481–483. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFreni F, Meduri A, Gazia F, et al.: Symptomatology in head and neck district in coronavirus disease (COVID-19): A possible neuroinvasive action of SARS-CoV-2. Am J Otolaryngol 2020 Sep-Oct; 41(5): 102612. Epub 2020/06/24. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSalepci E, Turk B, Ozcan SN, et al.: Symptomatology of COVID-19 from the otorhinolaryngology perspective: a survey of 223 SARS-CoV-2 RNA-positive patients. Eur. Arch. Otorhinolaryngol. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCazzolla AP, Lovero R, Lo Muzio L, et al.: Taste and Smell Disorders in COVID-19 Patients: Role of Interleukin-6. ACS Chem Neurosci 2020; 11(17): 2774–81. Epub 2020/08/19. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBidkar V, Mishra M, Selvaraj K, et al.: Testing Olfactory and Gustatory Dysfunctions among Quarantine COVID-19 Suspects. Indian J Otolaryngol Head Neck Surg 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJoffily L, Ungierowicz A, David AG, et al.: The close relationship between sudden loss of smell and COVID-19. Braz. J. Otorhinolaryngol. 2020; 86(5): 632–638. Publisher Full Text\n\nMoro E, Priori A, Beghi E, et al.: The international European Academy of Neurology survey on neurological symptoms in patients with COVID-19 infection. Eur. J. Neurol. 2020; 27(9): 1727–1737. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFantozzi PJ, Pampena E, Di Vanna D, et al.: Xerostomia, gustatory and olfactory dysfunctions in patients with COVID-19. Am J Otolaryngol Head Neck Med Surg 2020; 41(6). PubMed Abstract | Publisher Full Text | Free Full Text\n\nMao L, Wang M, Chen S, et al.: Neurological Manifestations of Hospitalized Patients with COVID-19 in Wuhan, China: a retrospective case series study. medRxiv 2020: 2020.02.22.20026500. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCabaraux P, Lechien JR, Saussez S, et al.: Objective Olfactory Evaluation of Self-reported Olfactory Dysfunction in a Case Series of 86 COVID-19 Patients. medRxiv 2020: 2020.05.03.20088526. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChiesa-Estomba CM, Lechien JR, Radulesco T, et al.: Patterns of smell recovery in 751 patients affected by the COVID-19 outbreak. Eur. J. Neurol. 2020; 27(11): 2318–2321. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTalavera B, García-Azorín D, Martínez-Pías E, et al.: Anosmia is associated with lower in-hospital mortality in COVID-19. J. Neurol. Sci. 2020: 419. PubMed Abstract | . PubMed Abstract | Publisher Full Text | Free Full Text\n\nGiorli A, Ferretti F, Biagini C, et al.: A Literature Systematic Review with Meta-Analysis of Symptoms Prevalence in Covid-19: the Relevance of Olfactory Symptoms in Infection Not Requiring Hospitalization. Curr. Treat. Options Neurol. 2020; 22(10): 36. Epub 2020/08/28. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDepartment of Health and Social Care: Statement from the UK Chief Medical Officers on an update to coronavirus symptoms: 18 May 2020.2020.\n\nEar N Throat Society of the United Kingdom (ENTUK): Loss of sense of smell as marker of COVID-19 infection.2020.\n\nKaye R, Chang CWD, Kazahaya K, et al.: COVID-19 Anosmia Reporting Tool: Initial Findings. Otolaryngol. Head Neck Surg. 2020 2020/07/01; 163(1): 132–4. PubMed Abstract | Publisher Full Text\n\nSuzuki M, Saito K, Min W-P, et al.: Identification of Viruses in Patients With Postviral Olfactory Dysfunction. Laryngoscope 2007 2007/02/01; 117(2): 272–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKondo Y, Miyazaki S, Yamashita R, et al.: Coinfection with SARS-CoV-2 and influenza A virus. BMJ Case Rep 2020; 13(7). PubMed Abstract | Publisher Full Text | Free Full Text\n\nLangerbeins P, Fürstenau M, Gruell H, et al.: COVID-19 complicated by parainfluenza co-infection in a patient with chronic lymphocytic leukemia. Eur. J. Haematol. 2020; 105(4): 508–511. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZayet S, Kadiane-Oussou NJ, Lepiller Q, et al.: Clinical features of COVID-19 and influenza: a comparative study on Nord Franche-Comte cluster. Microbes Infect. 2020; 22(9): 481–488. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlshebri MS, Alshouimi RA, Alhumidi HA, et al.: Neurological Complications of SARS-CoV, MERS-CoV, and COVID-19. SN Compr Clin Med 2020 2020/11/01; 2(11): 2037–47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHwang CS: Olfactory neuropathy in severe acute respiratory syndrome: report of A case. Acta Neurol. Taiwan. 2006 Mar; 15(1): 26–8. Epub 2006/04/08. eng. PubMed Abstract\n\nvon Bartheld CS, Hagen MM, Butowt R: Prevalence of Chemosensory Dysfunction in COVID-19 Patients: A Systematic Review and Meta-analysis Reveals Significant Ethnic Differences. ACS Chem. Neurosci. 2020 2020/10/07; 11(19): 2944–61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHeymans F, Lacroix J-S, Terzic A, et al.: Gustatory dysfunction after mandibular zoster. Neurol. Sci. 2011 2011/06/01; 32(3): 461–4. PubMed Abstract | Publisher Full Text\n\nGraham CS, Graham BG, Bartlett JA, et al.: Taste and smell losses in HIV infected patients. Physiol. Behav. 1995 1995/08/01/; 58(2): 287–93. PubMed Abstract | Publisher Full Text\n\nZayet S, Klopfenstein T, Mercier J, et al.: Contribution of anosmia and dysgeusia for diagnostic of COVID-19 in outpatients. Infection 2020: 1–5. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nButowt R, von Bartheld CS: Anosmia in COVID-19: Underlying Mechanisms and Assessment of an Olfactory Route to Brain Infection. Neuroscientist 2020 Sep 11: 1073858420956905. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAltin F, Cingi C, Uzun T, et al.: Olfactory and gustatory abnormalities in COVID-19 cases. Eur. Arch. Otorhinolaryngol. 2020; 277(10): 2775–81. Epub 2020/06/23. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEl-Anwar MW, Elzayat S, Fouad YA: ENT manifestation in COVID-19 patients. Auris Nasus Larynx 2020; 47(4): 559–64. Epub 2020/06/15. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGovindarajalu P, Shah R, Parsana M: Otorhinolaryngological manifestations of COVID-19-A systematic review. Res Sq 2020.\n\nTong JY, Wong A, Zhu D, et al.: The Prevalence of Olfactory and Gustatory Dysfunction in COVID-19 Patients: A Systematic Review and Meta-analysis. Otolaryngol. Head Neck Surg. 2020 Jul; 163(1): 3–11. Epub 2020/05/06. eng. PubMed Abstract | Publisher Full Text\n\nLehrich BM, Goshtasbi K, Raad RA, et al.: Aggregate Prevalence of Chemosensory and Sinonasal Dysfunction in SARS-CoV-2 and Related Coronaviruses. Otolaryngol Head Neck Surg 2020 Jul; 163(1): 156–61. Epub 2020/05/20. eng. PubMed Abstract | Free Full Text\n\nChilvers MA, McKean M, Rutman A, et al.: The effects of coronavirus on human nasal ciliated respiratory epithelium. Eur Respir J 2001 Dec; 18(6): 965–70. Epub 2002/02/07. eng. PubMed Abstract | Publisher Full Text\n\nBrann JH, Firestein SJ: A lifetime of neurogenesis in the olfactory system. Front Neurosci 2014; 8: 182. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHoffmann M, Kleine-Weber H, Schroeder S, et al.: SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor. Cell 2020; 181(2): 271–80.e8. Epub 2020/03/05. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBilinska K, Jakubowska P, Von Bartheld CS, et al.: Expression of the SARS-CoV-2 Entry Proteins, ACE2 and TMPRSS2, in Cells of the Olfactory Epithelium: Identification of Cell Types and Trends with Age. ACS Chem Neurosci 2020; 11(11): 1555–62. Epub 2020/05/19. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrann DH, Tsukahara T, Weinreb C, et al.: Non-neural expression of SARS-CoV-2 entry genes in the olfactory epithelium suggests mechanisms underlying anosmia in COVID-19 patients. bioRxiv 2020: 2020.03.25.009084. Publisher Full Text\n\nBryche B, St Albin A, Murri S, et al.: Massive transient damage of the olfactory epithelium associated with infection of sustentacular cells by SARS-CoV-2 in golden Syrian hamsters. Brain Behav Immun 2020; 89: 579–86. Epub 2020/07/03. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFodoulian L, Tuberosa J, Rossier D, et al.: SARS-CoV-2 receptor and entry genes are expressed by sustentacular cells in the human olfactory neuroepithelium. bioRxiv 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchwob JE: Neural regeneration and the peripheral olfactory system. Anat Rec 2002 Feb 15; 269(1): 33–49. Epub 2002/03/14. eng. PubMed Abstract | PubMed Abstract | Publisher Full Text\n\nNetland J, Meyerholz DK, Moore S, et al.: Severe acute respiratory syndrome coronavirus infection causes neuronal death in the absence of encephalitis in mice transgenic for human ACE2. J Virol 2008 Aug; 82(15): 7264–75. Epub 2008/05/23. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRodriguez S, Cao L, Rickenbacher GT, et al.: Innate immune signaling in the olfactory epithelium reduces odorant receptor levels: modeling transient smell loss in COVID-19 patients. medRxiv 2020: 2020.06.14.20131128. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTorabi A, Mohammadbagheri E, Akbari Dilmaghani N, et al.: Proinflammatory Cytokines in the Olfactory Mucosa Result in COVID-19 Induced Anosmia. ACS Chem Neurosci 2020; 11(13):1909–13. Epub 2020/06/11. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHenkin RI, Schmidt L, Velicu I: Interleukin 6 in hyposmia. JAMA Otolaryngol Head Neck Surg 2013 Jul; 139(7): 728–34. Epub 2013/07/23. eng. PubMed Abstract | Publisher Full Text\n\nSmall DM, Prescott J: Odor/taste integration and the perception of flavor. Exp Brain Res 2005 Oct; 166(3–4): 345–57. Epub 2005/07/20. eng. PubMed Abstract | Publisher Full Text\n\nVaira LA, Salzano G, Fois AG, et al.: Potential pathogenesis of ageusia and anosmia in COVID-19 patients. Int Forum Allergy Rhinol 2020 Sep; 10(9): 1103–4. Epub 2020/04/29. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGiacomelli A, Pezzati L, Conti F, et al.: Self-reported Olfactory and Taste Disorders in Patients With Severe Acute Respiratory Coronavirus 2 Infection: A Cross-sectional Study. Clin Infect Dis 2020 Jul 28; 71(15): 889–90. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLozada-Nur F, Chainani-Wu N, Fortuna G, et al.: Dysgeusia in COVID-19: Possible Mechanisms and Implications. Oral Surg Oral Med Oral Pathol Oral Radiol 2020; 130(3): 344–6. Epub 2020/06/27. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKomai M, Goto T, Suzuki H, et al.: Zinc deficiency and taste dysfunction; contribution of carbonic anhydrase, a zinc-metalloenzyme, to normal taste sensation. Biofactors 2000; 12(1-4): 65–70. Epub 2001/02/24. eng. PubMed Abstract | Publisher Full Text\n\nJothimani D, Kailasam E, Danielraj S, et al.: COVID-19: Poor outcomes in patients with zinc deficiency. International Journal of Infectious Diseases 2020 2020/11/01/; 100: 343–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNajafizade N, Hemati S, Gookizade A, et al.: Preventive effects of zinc sulfate on taste alterations in patients under irradiation for head and neck cancers: A randomized placebo-controlled trial. J Res Med Sci 2013 Feb; 18(2): 123–6. Epub 2013/08/06. eng. PubMed Abstract | Free Full Text\n\nRipamonti C, Zecca E, Brunelli C, et al.: A randomized, controlled clinical trial to evaluate the effects of zinc sulfate on cancer patients with taste alterations caused by head and neck irradiation. Cancer 1998 May 15; 82(10): 1938–45. Epub 1998/05/20. eng. PubMed Abstract | Publisher Full Text\n\nKar M, Khan NA, Panwar A, et al.: Zinc Chelation Specifically Inhibits Early Stages of Dengue Virus Replication by Activation of NF-κB and Induction of Antiviral Response in Epithelial Cells. Front Immunol 2019; 10: 2347. Epub 2019/10/22. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXiao Z, Ehrlich E, Luo K, et al.: Zinc chelation inhibits HIV Vif activity and liberates antiviral function of the cytidine deaminase APOBEC3G. Faseb J 2007 Jan; 21(1): 217–22. Epub 2006/12/01. eng. PubMed Abstract | Publisher Full Text\n\nSuliburska J, Duda G, Pupek-Musialik D: The influence of hypotensive drugs on the taste sensitivity in patients with primary hypertension. Acta Pol Pharm 2012 Jan-Feb; 69(1): 121–7. Epub 2012/05/12. eng. PubMed Abstract\n\nTsuruoka S, Wakaumi M, Araki N, et al.: Comparative study of taste disturbance by losartan and perindopril in healthy volunteers. J Clin Pharmacol 2005 Nov; 45(11): 1319–23. Epub 2005/10/22. eng. PubMed Abstract | Publisher Full Text\n\nTanasa IA, Manciuc C, Carauleanu A, et al.: Anosmia and ageusia associated with coronavirus infection (COVID-19) - what is known? Exp Ther Med 2020 2020/09/01; 20(3): 2344–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWitt M, Miller IJ Jr: Comparative lectin histochemistry on taste buds in foliate, circumvallate and fungiform papillae of the rabbit tongue. Histochemistry 1992 Oct; 98(3): 173–82. Epub 1992/10/11. eng. PubMed Abstract\n\nPushpass RG, Pellicciotta N, Kelly C, et al.: Reduced Salivary Mucin Binding and Glycosylation in Older Adults Influences Taste in an in vitro Cell Model. Nutrients 2019 Sep 24; 11(10). Epub 2019/09/27. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMilanetti E, Miotto M, Rienzo LD, et al.: In-Silico evidence for two receptors based strategy of SARS-CoV-2. bioRxiv 2020: 2020.03.24.006197. Publisher Full Text\n\nPark Y-J, Walls AC, Wang Z, et al.: Structures of MERS-CoV spike glycoprotein in complex with sialoside attachment receptors. Nat. Struct. Mol. Biol. 2019 2019/12/01; 26(12): 1151–7. Publisher Full Text\n\nXu H, Zhong L, Deng J, et al.: High expression of ACE2 receptor of 2019-nCoV on the epithelial cells of oral mucosa. Int. J. Oral Sci. 2020 2020/02/24; 12(1): 8. Publisher Full Text\n\nWang H, Zhou M, Brand J, et al.: Inflammation and taste disorders: mechanisms in taste buds. Ann N Y Acad Sci 2009; 1170: 596–603. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu L, Wei Q, Alvarez X, et al.: Epithelial cells lining salivary gland ducts are early target cells of severe acute respiratory syndrome coronavirus infection in the upper respiratory tracts of rhesus macaques. J Virol 2011 Apr; 85(8): 4025–30. Epub 2011/02/04. eng. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXu J, Li Y, Gan F, et al.: Salivary Glands: Potential Reservoirs for COVID-19 Asymptomatic Infection. J. Dent. Res. 2020 2020/07/01; 99(8): 989. PubMed Abstract | Publisher Full Text\n\nHarapan H: Anosmia and dysgeusia in SARS-CoV-2 infection: Incidence, effects on COVID-19 severity and mortality, and the possible pathobiology mechanisms - A systematic review and meta-analysis. figshare 2020 Journal contribution. Reference Source"
}
|
[
{
"id": "77883",
"date": "16 Feb 2021",
"name": "Cissy B. Kartasasmita",
"expertise": [
"Reviewer Expertise Pediatrics",
"Epidemiology",
"Vaccinology",
"and Respirology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nObjectives of study ‘\"The present study aimed to summarize the global evidence of anosmia and dysgeusia among COVID-19 patients, in order to assess their association with the severity and mortality of the disease, and provide a comprehensive review related to the possible pathogenesis of anosmia and dysgeusia in SARS-CoV-2 infection.”\nThe analysis on the assessment of the association with the severity and mortality of the disease is very short and needs more data to be reported.\nAssociation of anosmia and dysgeusia with COVID-19 severity and mortality \"Limited studies have assessed the association between anosmia and dysgeusia and the severity and mortality of COVID-19 cases. One study linked anosmia with a lower fatality rate and a lower ICU admission.\"\nIn the conclusion the authors did not include all the objectives of the study.\n\nNo statement on the association anosmia and dysgeusia with severity and mortality as stated in the objectives.\nConclusions\nMy recommendation would be to add the limitations of this study.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Yes\n\nAre the conclusions drawn adequately supported by the results presented in the review? Partly",
"responses": []
},
{
"id": "80214",
"date": "22 Mar 2021",
"name": "Seyi Samson Enitan",
"expertise": [
"Reviewer Expertise Medical Virology and Immunology of Infectious Diseases."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe present study examined the global prevalence of anosmia and dysgeusia in COVID-19 patients, their association with severity and mortality of COVID-19, as well as the possible pathobiological mechanisms of anosmia and dysgeusia in COVID-19. Authors reported a global prevalence of 38.2% and 36.6% for anosmia and dysgeusia, respectively. Identified potential mechanisms for anosmia include: Obstruction in the nasal airway, damage in olfactory sensory neurons, olfactory center damage in the brain, olfactory supporting cells dysfunction and inflammation-related olfactory epithelium dysfunction. On the other hand, the subsequent effect of cranial nerves dysfunction, Zinc deficiency, SARS-CoV-2-bound sialic acid, Direct attack on several oral sites are opined to be responsible for the dysgeusia.\nThe work is okay and the findings are worth-sharing with the scientific community. The introduction is considered satisfactory. Authors provided background that puts the manuscript into context and allows readers outside the field to understand the purpose and significance of the study. They also identified the existing gap in knowledge that needs to be filled. The methodology section was clearly presented to allow the reproduction of the study. Discussion and Conclusion were well written. However, the limitation of the study was not clearly stated. And except, the association of anosmia and dysgeusia with severity and mortality is properly discussed in the study, it should be deleted from the objectives.\nRating of the manuscript: Use (1 = Excellent) (2 = Very Good) (3 = Average) (4 = Fair) (5 = poor)\nOriginality 2\nContribution To The Field 1\nTechnical Quality 2\nClarity of Presentation 2\nDepth of Research 2\n\nRecommendation: Minor corrections are needed.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Yes\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes",
"responses": []
},
{
"id": "80213",
"date": "23 Mar 2021",
"name": "Mahir Gachabayov",
"expertise": [
"Reviewer Expertise clinical outcomes and evidence synthesis"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThank you for an interesting systematic review. This systematic review and meta-analysis aimed at evaluating the rate of anosmia and dysgeusia in patients with confirmed COVID-19 and their association with disease severity and mortality. The research question makes sense and the methodology chosen for statistical analysis is logical. The Introduction is comprehensible and easy to read, it clearly states the gap in the literature and the rationale behind this research question. The aim of the study was stated clearly. The Methods are well-formulated. Reporting of this review is compliant with the PRISMA guidelines. The search strategy is comprehensive, includes major databases as well as preprint servers. The details of the search strategy and the PRISMA flow diagrams were provided. Eligibility criteria and definitions were reported. Statistical analysis was adequate. The Results are well-written and clear. The findings of the statistical analysis were nicely summarized in tables and illustrated in forest plots. The Discussion is comprehensible and easy to read. The authors have provided interpretation of their findings, and clinical and scientific implications thereof. The authors have provided possible etiopathogenesis for anosmia and dysgeusia in COVID-19 patients. The Conclusion was justified by the statistical findings.\nIn order to improve the manuscript, I have a few suggestions:\nThere is some confusion between the terms incidence and prevalence. In fact, in the Methods under the Outcomes, the authors stated the primary outcomes to be incidence of anosmia and dysgeusia, whereas they described this metrics as prevalence in the rest of the manuscript. I believe neither of the terms fits the context. I would change these terms to rate, eg. anosmia rate (rate of anosmia) and dysgeusia rate (rate of dysgeusia).\n\nI would not use Newcastle-Ottawa scale for risk of bias assessment as it does not evaluate for heterogeneity due to the differences in the definitions of outcomes and interventions to measure the outcome. I would rather use ROBINS-I tool.\n\nI would address in the Discussion the reasons for substantial heterogeneity in the pooled rates of anosmia and dysgeusia. I would consider heterogeneity in geographic locations and community types, heterogeneity in the definitions of anosmia and dysgeusia, and heterogeneity in the definitions of disease severity across the included studies.\n\nI would add a brief paragraph to the Discussion (last paragraph) and acknowledge the strengths and limitations of this study.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Yes\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-40
|
https://f1000research.com/articles/10-39/v1
|
21 Jan 21
|
{
"type": "Research Article",
"title": "Comparison of different machine learning methods and dimensionality reduction for classification astrocytoma and glioblastoma tissues by mass spectra",
"authors": [
"Evgeny S. Zhvansky",
"Anatoly A. Sorokin",
"Denis S. Zavorotnyuk",
"Vsevolod A. Shurkhay",
"Vasiliy A. Eliferov",
"Denis S. Bormotov",
"Daniil G. Ivanov",
"Alexander A. Potapov",
"Anatoly A. Sorokin",
"Denis S. Zavorotnyuk",
"Vsevolod A. Shurkhay",
"Vasiliy A. Eliferov",
"Denis S. Bormotov",
"Daniil G. Ivanov",
"Alexander A. Potapov"
],
"abstract": "Background: Recently developed methods of ambient ionization allow rapid obtaining of large mass spectrometric datasets, which have a great application in biological and medical analysis. One of the areas that could employ such analysis is neurosurgery. The fast in situ identification of dissected tissues could assist the neurosurgery procedure. The additional information about tumor could help the tumor border monitoring. In this paper, tumor tissues of astrocytoma and glioblastoma are compared, as their identifications during surgery could influence the extent of resection and, hence, the median and overall survival. Methods: Mass spectrometric profiles of brain tumor tissues contain molecular information, which is rather hard to interpret in terms of identifications of individual molecules. The machine learning algorithms are employed for the fast automated mass spectra classification. Different algorithms of dimensionality reduction are considered to process the mass spectra before the classification task, as the initial dimensionality of mass spectra is too high compared with the number of mass spectra. Results: Different classifiers are compared for both just preprocessed data and after dimensionality reduction. The Non-Negative Matrix Factorization appears to be the most effective dimensionality reduction algorithm. The random forest algorithm demonstrated the most robust appearance on the tested data. Also, the comparison of the accuracy of the trained classifier on the mass spectra of tissues measured with different instruments and different resolution is provided in the paper. Conclusions: Machine learning classifiers overfit the raw mass spectrometric data. The dimensionality reduction allows the classification of both train and test data with 88% accuracy. Positive mode data provides better accuracy. A combination of principal component analysis and AdaBoost algorithms appears to be most robust to changing the instrument and conditions.",
"keywords": [
"mass spectra",
"astrocytoma and glioblastoma tumors",
"dimensionality reduction",
"classification",
"high- and low-resolution"
],
"content": "Introduction\n\nThe extent of tumor resection is important for patients with primary brain tumors in terms of life expectancy since tumor cells can provoke a disease recurrence1. Recently, we have seen a growing interest in the use of mass spectrometry for the identification of tumor tissues, typing, and detection of tumor boundaries2-4. Analysis of tumor samples using mass spectrometry is based on the observation that tumor cells differ significantly from normal ones in their metabolic processes and, as a consequence, have a different chemical composition5-8. Identification of the histological type and location of the brain tumor tissue during neurosurgical intervention allows for the correct dissection of the tumor and opens the way to a personalized strategy for further treatment of the patient with chemotherapy, taking into account the molecular features of the tumor. Comparative analysis of tumor types is fundamental, although elucidating tumor boundaries is of the highest priority for neurosurgeons9.\n\nFast mass spectrometric profiling allows rapid clinic and laboratory analyses but faces the problem of classification of high-dimensionality objects.\n\nFor the analysis of large mass spectrometry (MS) data, dimension reduction (DR) algorithms are commonly used as a previous step for statistical analysis and visualization. The most widely used DR methods are linear methods such as PLS-DA and PCA10-16. More advanced nonlinear methods have recently been developed, such as t-SNE and UMAP, to name a few17. DR methods allow visualization through the major components of the compressed data, e.g. the first three PCA components and three selected ions are the most commonly used characteristics in MS imaging18,19.\n\nDespite the visualization that could produce the results dividing the different types of tissues by their MS, these results are subjective and require further automated classification. Machine learning (ML) methods are commonly used for these purposes20-22.\n\nIn this paper, we compare the performance of six DR algorithms and ten ML algorithms in their ability to classify the MS profiles of astrocytoma and glioblastoma. Also, we compared the stability of the trained ML models on the data obtained with another instrument and under different conditions. The stability of the trained models to different instruments is very important for the wide spreading of the methods, as the clinical conditions could influence the mass spectra due to different mass analyzers. Different polarities, resolutions, and m/z ranges are also considered.\n\n\nMethods\n\nThe two instruments used in our study: for both high and low resolution (HR and LR) under laboratory conditions (Thermo LTQ Orbitrap XL) and only low resolution under clinical conditions (Thermo LTQ XL). Inline cartridge extraction23 followed by electrospray ionization was used for mass spectrometric profiling of samples. Spectra from Thermo LTQ Orbitrap XL were measured in both positive and negative ion modes with two resolutions: 24,000 at m/z=744 (high resolution using Orbitrap analyzer) and 2000 at m/z=744 (low resolution using LTQ analyzer). All spectra were measured in the m/z 500 — 1000 and m/z 100 — 2000 ranges.\n\nTissue samples were provided by the N.N. Burdenko NSPCN and analyzed under a protocol approved by N.N. Burdenko NSPCN Institutional Review Board (order 40 from 12.04.2016, revised with order 131 from 17.07.2018). A signed informed consent form, filled out in accordance with the requirements of the local ethical committee, specifically noting that all removed tissues can be used for further research, was obtained from all patients before surgery. The study was conducted in accordance with the Helsinki Declaration as revised in 2013. All procedures were carried out according to the relevant guidelines and regulations.\n\nThree fragments of tissue taken from a single patient were measured with each instrument to take into account and evaluate inner biological variability. In total, 76 astrocytoma fragments (26 patients) and 89 glioblastomas (31 patients) were measured with Thermo LTQ XL (Ltq) in LR, while 60 astrocytoma fragments (20 patients), 84 glioblastoma fragments (28 patients) were measured with Thermo LTQ Orbitrap XL (Orbitrap) in LR and HR. Samples from 37 patients were measured with both instruments, 20 patients were measured only with Ltq, 11 — only with Orbitrap. The precise schema is given in Figure 1.\n\nThere tissue samples (total 309) from 68 patients had different brain tumors: 21 anaplastic astrocytomas (WHO Grade III; 9 tumors with IDH-1 R132H mutation), 10 diffuse astrocytomas (WHO Grade II; 7 tumors with IDH-1 R132H mutation), 1 gemistocytic astrocytoma WHO Grade II (IDH-1 with IDH-1 R132H mutation) and 36 glioblastomas (WHO Grade IV; 11 tumors with IDH-1 R132H mutation).\n\nTissue samples were divided into several parts, for annotation and MS analysis. Tissue samples were annotated with routine hematoxylin and eosin staining and further immunohistochemical analysis of its fragment (expression of isocitrate dehydrogenase 1 (IDH-1)). Other fragments of tissue samples were measured with Thermo LTQ XL right after removal or were placed in the normal saline, frozen, and stored at -80°C until measurement with Thermo LTQ Orbitrap XL. Measurement for each sample was carried out continuously with alternating mode, range, and resolution if possible24.\n\nSpectra were processed with the algorithm similar to described previously25,26. Mass spectra were binned with binning width 0.01 m/z, and then spectra were convoluted with Gaussian (FWHM equals 0.4m/z for high-resolution and 0.2m/z for low-resolution). Spectra of each measurement were filtered by a moving median filter. The width and step of the median filter were chosen to 51. The baseline subtraction was carried out through Kneen and Annegarn’s algorithm27. Spectra similarity matrix (SSM) with cosine measure similarity was calculated as described previously26. Both dimensionality reduction and classification were made with Scikit-learn v0.23.128 machine learning library.\n\nAll calculations and visualizations were made using code written by the authors using MATLAB R2019b (GNU Octave 5.2.0 could be used for reproducing results) or Python 3.7. This code is freely available at http://doi.org/10.5281/zenodo.430770043.\n\nThe classification without dimensionality reduction is compared with the method of principal component analysis (PCA), non-negative matrix factorization (NNMF)29, isometric mapping (Isomap)30, Partial least squares discriminant analysis (PLS-DA)31, UMAP32, Diffusion map33. The number of the selected features after dimensionality reduction equals 5.\n\nNearest Neighbors34, linear support vector machine SVM35, Radial-basis function (RBF) kernel SVM35, Gaussian Process36, Decision Tree37, Random Forest38, Neural Net (Multi-layer Perceptron classifier with 1 hidden layer with 100 neurons and the log-loss function and “adam” optimizer and L2 regularization), AdaBoost39, Naive Bayes40, QDA41 were used as classifiers. Only one scan from each measurement after median filtration was taken part in the classification task.\n\n\nResults\n\nSSMs of astrocytoma and glioblastoma in different polarities are presented in Figure 2. There are two clusters of fresh and frozen samples. The first were measured right after surgery, the second after storing samples under −80°C. Figure 2 represents the large-range data (100-2000 m/z) measured in low-resolution.\n\n(A) Negative and (B) positive mode.\n\nFrozen glioblastoma MS profiles have a lot of outliers, which is obvious from the SSM Figure 2B. The outliers could be filtered out from the datasets for the classification task, but the classifiers will be created with only fresh samples. Thus, outliers will be tested on the trained classifiers and the results of visual inspection of SSM could be confirmed. The negative mode (Figure 2A) has more variance and spectra of the negative mode have less similarity. At the same time, there are no obvious differences between astrocytoma and glioblastoma. We, therefore, employ the machine learning methods for classifying the MS profiles.\n\nThe accuracy of classifiers for different dimensionality reduction algorithms is presented in Figure 3. The accuracy was calculated as the ratio of correct predictions to the total number of predictions on the given data and labels.\n\n(A-D) 100-2000 m/z; (E-H) 500-1000 m/z.\n\nFirst 70% of astrocytomas and 70% of glioblastomas were taken as the training data, the rest as the test data. Neither of the mass spectra of the samples corresponding to the single patient was in training and test data simultaneously to prevent the overfitting. Thus 46% of training data and 46% of test data were astrocytoma class, whereas 54% of the training and test data were glioblastoma class. Many of the combinations of classifiers and dimensionality reduction algorithms score about 46% or 54% of accuracy.\n\nThe PLS-DA is overfitted in all cases except probably negative mode and m/z 500-1000 (Figure 3E and F). The highest values of accuracy are achieved with NNMF for positive mode MS profiles. NNMF produced better results for a wide m/z range (100-2000) since classifiers produce similar accuracy in this case. The best result of NNMF and random forest combination for the positive mode, m/z 500-1000 is achieved for 5 components of NNMF. The most stable results were produced for the 100-2000 m/z range in positive mode. These conditions are considered below\n\nThe robustness of each combination of DR and ML algorithms is tested on the data obtained with another instrument (Orbitrap) both in low and high resolution (Figure 4). In this setup, the classifiers were taken from the previous step, so they were fitted with Ltq data. High-resolution data was roughed through Gaussian convolution to decrease resolving power to low-resolution data.\n\n(A) Low-resolution; (B) high-resolution converted to low-resolution.\n\nFigure 4 demonstrates the lack of classifier robustness to the different instruments despite the rather high similarity of MS profiles from Figure 2. The combinations of the algorithms that seem to be overfitted from Figure 3C and D (Diffusion map, PLS-DA) leads to almost random classification for both low-resolution and high-resolution converted to low-resolution data.\n\nConsidering data without outliers provides better results (Figure 5) for overfitted models, whereas it has an almost negligible influence on the other models’ accuracy. The exclusion of outliers with the SSM by indicating low-similarity blue strips mostly in glioblastoma samples reduces the number of samples from 144 to 116 for low-resolution data and from 144 to 113 for high-resolution. It increased the accuracy, for example, of NNMF+Linear SVM by five percent.\n\n(A) Low resolution; (B) high resolution.\n\nThe combination of the PCA and AdaBoost approach demonstrates here the most robustness to the instrument.\n\n\nDiscussion\n\nThe SSMs in Figure 2 demonstrate different variance of MS profiles in negative and positive mode. The negative mode has less similarity between MS profiles especially in the m/z range 100-2000. The positive mode has similar MS profiles excepting the outliers that are not contained in the negative mode24,42. Thus, the positive mode could be used for looking for general distinctions between classes, whereas the negative mode could contain fine distinctions in the class as well. The classifiers seem to have difficulties in application to such high variation objects as MS profiles have in negative mode.\n\nPreliminary feature selection (dimensionality reduction) is necessary for classifiers to not be overfitted, as the dimensionality of the object is equal to 7600 features in the given preprocessing (binning for m/z 100-2000 range) and the number of samples is about 150 (about 50 patients) in the dataset from the single instrument. The dimensionality reduction is expected to work rather effectively in the first several components only in case the variation is caused by interclass variability with classes characterized by a single Gaussian distribution. In the presented case, each class has a set of multi-dimensional Gaussian distributions. Also, astrocytoma tissues can have smooth changes of grades up to glioblastoma due to its origin. The whole variability of the data could be explained by six factors: tumor type, tumor localization, mutation status, interpatient variability, intratumor variability, and batch effect. Usually, interpatient variability is comparable with interclass variability. So, the presented data is unlikely to have most of its variability in the first components due to tumor type. Thus, both for the data with low (positive mode, Figure 2), and with high variance data (negative mode Figure 2) the first components of DR could be useless in terms of classification if the intraclass variability is higher than interclass.\n\nPCA (Figure 3) provides good results for positive mode: most classifiers achieve similar accuracy of about 80%. PCA is ineffective for negative mode, as the accuracy falls to ~46%/54%, which corresponds to the ratio of the classes in the dataset. At the same time accuracy in the training data is much higher, so PCA provides specific only for the training data, which leads to overfitting. PCA and further classifiers predict mostly one class for any test data in negative mode.\n\nNNMF components correspond to the highest accuracy values in most conditions. It seems to be not overfitted in positive mode. NNMF produces the best results for the m/z range 500-1000 in combination with random forest or decision tree models. NNMF and PCA repeat the accuracy of the test data with the KNN classifier for positive mode.\n\nUMAP seems to be perspective, whereas Isomap, Diffusion map, PLS-DA algorithms demonstrate worse results compared with others. Classifiers are the most overfitted if no DR algorithm is applied, as expected. Although the AdaBoost model has about 70% accuracy on the test data for each experimental condition for non-preprocessed data (Figure 3).\n\nPLS-DA is used for dimensionality reduction here, but it is common practice to use it as a classifier. PLS-DA appears to be overfitted for all cases except the 500-1000 m/z range in negative mode, where all other combinations are overfitted. Random forest demonstrates the best result for this case but still seems to be slightly overfitted, as the accuracy on the test data decreases by 20% compared with accuracy on the train data.\n\nNeural net and AdaBoost models demonstrated the best results for uncompressed data in negative mode.\n\nAlgorithms were applied to a similar dataset of frozen samples, measured with another instrument (Orbitrap). As the measurements were carried out in two resolutions, the accuracy matrices are calculated for both resolutions. Figure 4 shows a lower level of accuracy compared to Figure 3, e.g. for PCA+AdaBoost. This is explained by the outliers, which could be seen in Figure 2B for frozen glioblastomas as the blue crossing stripes correspond to MS profiles that don’t have similar MS profiles in the dataset. Excluding the outliers from the consideration reveals the accuracy improvements, but the accuracy is still not so high. The combination PCA+AdaBoost demonstrated the most robust result of classification over all the data in positive mode (Figure 3-5). Thus, the most independent combination of dimensionality reduction and classification algorithm to the instrument, resolution, and freeze-thaw process is supposed to be PCA+AdaBoost. This combination accuracy is not changing for positive mode from train to test data on any instrument and equals to about 70-80%.\n\n\nConclusions\n\n\n\n1. Positive mode mass spectra provide better accuracy for astrocytoma and glioblastoma classification by mass spectrometric profiles of samples without sample preparation.\n\n2. Astrocytoma and glioblastoma could be classified with 88% accuracy in low-resolution (NNMF+random forest) by positive-mode mass spectrometric profiles.\n\n3. The PCA and AdaBoost combination appeared to be most stable in positive mode while transferring the classifier from the Ltq’s to Orbitrap’s data. The accuracy of classification is about 65-70% for validation data.\n\n4. The dimensionality reduction algorithms combined with the classification models can process the outliers from the SSM as normal data.\n\n\nData availability\n\nZenodo: Data and code for comparison of different machine learning methods and dimensionality reduction for classification astrocytoma and glioblastoma tissues by mass spectra, https://doi.org/10.5281/zenodo.430770043.\n\nThis project contains the following underlying data:\n\nDatasets of mass spectrometric profiles for different instruments, ranges, polarities, and resolutions.\n\nSoftware files for figure replication.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Acknowledgments\n\nThe research used the equipment of Shared Research Facilities of N.N. Semenov Federal Research Center for Chemical Physics of the Russian Academy of Sciences.\n\n\nReferences\n\nErmolaev AY, Kravets LY, Smetanina SV, et al.: Cytologic control of the resection margins of hemispheric gliomas and metastases. Zh. Vopr. Neirokhir. Im. N N Burdenko 2020; 84: 33–42. PubMed Abstract | Publisher Full Text\n\nAgar NYR, Golby AJ, Ligon KL, et al.: Development of stereotactic mass spectrometry for brain tumor surgery. Neurosurgery 2011; 68: 280–89; discussion 290. PubMed Abstract | Publisher Full Text | Free Full Text\n\nClark AR, Calligaris D, Regan MS, et al.: Rapid discrimination of pediatric brain tumors by mass spectrometry imaging. J Neurooncol 2018; 140: 269–279. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEberlin LS, Norton I, Orringer D, et al.: Ambient mass spectrometry for the intraoperative molecular diagnosis of human brain tumors. Proc Natl Acad Sci USA 2013; 110: 1611–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSorokin A, Shurkhay V, Pekov S, et al.: Untangling the metabolic reprogramming in brain cancer: discovering key molecular players using mass spectrometry. Curr. Top. Med. Chem. 2019; 19: 1521–1534. PubMed Abstract | Publisher Full Text\n\nCarpinteiro A, Dumitru C, Schenck M, et al.: Ceramide-induced cell death in malignant cells. Cancer Lett. 2008; 264: 1–10. PubMed Abstract | Publisher Full Text\n\nWymann MP, Schneiter R: Lipid signalling in disease. Nat. Rev. Mol. Cell Biol. 2008; 9: 162–76. PubMed Abstract | Publisher Full Text\n\nHannun YA, Obeid LM: Principles of bioactive lipid signalling: lessons from sphingolipids. Nat. Rev. Mol. Cell Biol. 2008; 9: 139–150. PubMed Abstract | Publisher Full Text\n\nLau D, Hervey-Jumper SL, Han SJ, et al.: Intraoperative perception and estimates on extent of resection during awake glioma surgery: overcoming the learning curve. J. Neurosurg. 2018; 128: 1410–1418. PubMed Abstract | Publisher Full Text\n\nPovey JF, O’Malley CJ, Root T, et al.: Rapid high-throughput characterisation, classification and selection of recombinant mammalian cell line phenotypes using intact cell MALDI-ToF mass spectrometry fingerprinting and PLS-DA modelling. J. Biotechnol. 2014; 184: 84–93. PubMed Abstract | Publisher Full Text\n\nPereira HV, Amador VS, Sena MM, et al.: Paper spray mass spectrometry and PLS-DA improved by variable selection for the forensic discrimination of beers. Anal. Chim. Acta 2016; 940: 104–12. PubMed Abstract | Publisher Full Text\n\nCajka T, Smilowitz JT, Fiehn O: Validating Quantitative Untargeted Lipidomics Across Nine Liquid Chromatography-High-Resolution Mass Spectrometry Platforms. Anal. Chem. 2017; 89: 12360–12368. PubMed Abstract | Publisher Full Text\n\nAnderson TJ, Jones RW, Ai Y, et al.: High-resolution time-of-flight mass spectrometry fingerprinting of metabolites from cecum and distal colon contents of rats fed resistant starch. Anal. Bioanal. Chem. 2014; 406: 745–756. PubMed Abstract | Publisher Full Text\n\nZhou W, Xia L, Huang C, et al.: Rapid analysis and identification of meat species by laser-ablation electrospray mass spectrometry (LAESI-MS). Rapid Commun. Mass Spectrom. 2016; 30(Suppl 1): 116–121. PubMed Abstract | Publisher Full Text\n\nCortés M, Pareja E, Castell JV, et al.: Exploring mass spectrometry suitability to examine human liver graft metabonomic profiles. Transplant. Proc. 2010; 42: 2953–2958. PubMed Abstract | Publisher Full Text\n\nHänel L, Kwiatkowski M, Heikaus L, et al.: Mass spectrometry-based intraoperative tumor diagnostics. Future Science OA 2019; 5: FSO373. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoon KR, van Dijk D, Wang Z, et al.: Visualizing structure and transitions in high-dimensional biological data. Nat. Biotechnol. 2019; 37: 1482–1492. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRace AM, Bunch J: Optimisation of colour schemes to accurately display mass spectrometry imaging data based on human colour perception. Anal. Bioanal. Chem. 2015; 407: 2047–2054. PubMed Abstract | Publisher Full Text\n\nAbramowski P, Kraus O, Rohn S, et al.: Combined application of RGB marking and mass spectrometric imaging facilitates detection of tumor heterogeneity. Cancer Genomics Proteomics 2015; 12: 179–187. PubMed Abstract\n\nMascini NE, Teunissen J, Noorlag R, et al.: Tumor classification with MALDI-MSI data of tissue microarrays: A case study. Methods 2018; 151: 21–27. PubMed Abstract | Publisher Full Text\n\nChagovets VV, Starodubtseva NL, Tokareva AO, et al.: Validation of breast cancer margins by tissue spray mass spectrometry. Int. J. Mol. Sci. 2020; 21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEberlin LS, Norton I, Dill AL, et al.: Classifying human brain tumors by lipid imaging with mass spectrometry. Cancer Res. 2012; 72: 645–654. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPekov SI, Eliferov VA, Sorokin AA, et al.: Inline cartridge extraction for rapid brain tumor tissue identification by molecular profiling. Sci. Rep. 2019; 9: 18960. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhvansky ES, Eliferov VA, Sorokin AA, et al.: Assessment of variation of inline cartridge extraction mass spectra. J. Mass Spectrom. 2020: e4640. PubMed Abstract | Publisher Full Text\n\nZhvansky ES, Sorokin AA, Pekov SI, et al.: Unified representation of high- and low-resolution spectra to facilitate application of mass spectrometric techniques in clinical practice. Clin Mass Spectrom 2019; 12: 37–46. Publisher Full Text\n\nZhvansky ES, Pekov SI, Sorokin AA, et al.: Metrics for evaluating the stability and reproducibility of mass spectra. Sci. Rep. 2019; 9: 914. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKneen MA, Annegarn HJ: Algorithm for fitting XRF, SEM and PIXE X-ray spectra backgrounds. Nucl Instrum Methods Phys Res B 1996; 109–110: 209–213. Publisher Full Text\n\nPedregosa F, Varoquaux G, Gramfort A: Scikit-learn: Machine learning in Python. In: J Mach Learn Res 2011.\n\nCichocki A, Phan A-H: Fast local algorithms for large scale nonnegative matrix and tensor factorizations. IEICE T Fund Electr 2009; E92-A: 708–721. Publisher Full Text\n\nTenenbaum JB, de Silva V, Langford JC: A global geometric framework for nonlinear dimensionality reduction. Science 2000: 290, 2319–2323. PubMed Abstract | Publisher Full Text\n\nBarker M, Rayens W: Partial least squares for discrimination. J. Chemom. 2003; 17: 166–173. Publisher Full Text\n\nBecht E, McInnes L, Healy J, et al.: Dimensionality reduction for visualizing single-cell data using UMAP. Nat. Biotechnol. 2018; 37: 38–44. PubMed Abstract | Publisher Full Text\n\nBerry T, Harlim J: Variable bandwidth diffusion kernels. Appl. Comput. Harmon. Anal. 2016; 40: 68–96. Publisher Full Text\n\nAltman NS: An Introduction to Kernel and Nearest-Neighbor Nonparametric Regression. Am. Stat. 1992; 46: 175–185. Publisher Full Text\n\nChang CC, Lin CJ: LIBSVM: A library for support vector machines. ACM Trans. Intell. Syst. Technol. 2011; 2: 1–27. Publisher Full Text\n\nRasmussen CE, Williams CKI: Gaussian Processes for Machine Learning.Cambridge: MIT Press; 2005; Mass p. 266; ISBN 0-262-18253-X.\n\nBreiman L, Friedman JH, Olshen RA, et al.: Classification and regression trees. Biometrics 1984; 40: 874. Publisher Full Text\n\nBreiman L: Random Forests. Springer Science and Business Media LLC 2001. Publisher Full Text\n\nHastie T, Rosset S, Zhu J, et al.: Multi-class AdaBoost. Stat. Interface 2009; 2: 349–360. Publisher Full Text\n\nRennie JD, Shih L, Teevan J, et al.: Tackling the poor assumptions of naive bayes text classifiers. In: Proceedings of the 20th … 2003.\n\nHastie T, Tibshirani R, Friedman J: The Elements of Statistical Learning. In: Springer Series in Statistics 2nd ed.; New York, NY: Springer New York; 2009: pp. 106–119; ISBN 978-0-387-84857-0.\n\nEliferov VA, Zhvansky ES, Sorokin AA, et al.: The role of lipids in the classification of astrocytoma and glioblastoma using MS tumor profiling. Biomed. Khim. 2020; 66: 317–325. PubMed Abstract | Publisher Full Text\n\nZhvansky E, Sorokin A, Shurkhay V, et al.: Data and code for comparison of different machine learning methods and dimensionality reduction for classification astrocytoma and glioblastoma tissues by mass spectra [Data set]. Zenodo 2020. Publisher Full Text"
}
|
[
{
"id": "80908",
"date": "15 Mar 2021",
"name": "Sergei A. Moshkovskii",
"expertise": [
"Reviewer Expertise mass spectrometry",
"proteomics",
"proteogenomics",
"molecular medicine"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nA paper by Zhvansky et al aims to distinguish between glioblastoma and astrocytoma tumor tissues by analysis of mass-spectrometry tissue profiling. Mass-spectra recorded from tissues intra- and extra-surgery, using low and high resolution mass spectrometry. Authors tried to select an optimal mode for spectra recording and processing, to use, in perspective, a mass spectrometry during surgery. Similar efforts are recognized in the field. However, the paper may add to the prior art some new ways of data processing. In its present form, it is still difficult for perception by the readership and has to be crucially improved.\nAbstract. Please specify what certain mass spectrometry methods were used. Actually, they were two, a linear ion trap and Orbitrap, not just ‘different’, as it was stated. The abstract must better illustrate a specific content of the work. Please check the language, a ‘train’ in abstract has nothing to do with a railway.\n\nIntroduction. A text neglects needs of readers. Many abbreviations are given without deciphering. Check all these PLS-D, tSNE, UMAP et al, these are not established terms in the field of biomedicine. In the end, better define the novelty of your approach. Also specify the aim more clear. Is it to classify between glioblastoma and astrocytoma? Why, again, do we need this during surgery?\n\nA pipeline (clinics, analytics, data processing) is not quite clear. We have here two tumors with subclasses inside them, frozen vs. fresh tissues, two resolutions by two MS machines, and two modes of spectra recording. Please provide a figure (maybe instead of the Venn diagrams in Fig.1 or in addition) strictly designating what exactly was done. From the conclusion, it is only partly clear what type of pipeline gave the best result.\n\nWhat was done with the spectra produced from the same tumor? If they are used by classifiers independently, that is a mistake. Then, the model mixed intertumoral and intratumoral variability, which is not possible. Further, what was finally done with the outliers? An explanation in page 4, the last paragraph, is difficult to understand. Please clarify.\n\nThe idea of transition of the results between LTQ and Orbitrap seems to be not feasible. Even if ionization is the same, these ion traps act differently and may record different ions with different intensity, with inevitable loss in model performance. Please discuss with examples of similar use of classification models in mass spectrometry.\n\nDiscussion. Page 7, paragraph 2. A big piece of text with several unobvious statements was written without references, please provide them.\n\nFigures. All legends are short and unclear for the reader who looks through the paper, sometimes with abbreviations not disclosed. Please make the legends self-explanatory.\n\nLanguage. Many places are unclear due to uncertain English. Please ask the confident English speaker and writer to go through the text.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "99306",
"date": "30 Nov 2021",
"name": "Konstantin Chingin",
"expertise": [
"Reviewer Expertise mass spectrometry"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI fully agree with the first reviewer's comments.\n\nOn top of that, I suggest that the Introduction should be written in more detail. Also, the authors should provide more background on the determination of tumour boundaries by MS for other tissues as reported in earlier studies. Are there specific difficulties related to brain as compared to other types of tissues?\nLanguage should be improved with the help of a native English speaker.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-39
|
https://f1000research.com/articles/10-38/v1
|
21 Jan 21
|
{
"type": "Brief Report",
"title": "RhD blood type significantly influences susceptibility to contract COVID-19 among a study population in Iraq",
"authors": [
"Khalid R Majeed",
"Dhurgham Al-Fahad",
"Hayder Hussein Jalood",
"Haider A Hantosh",
"Mrtatha K Ali",
"Sumiktsal Sakthivel",
"Harry F Williams",
"Jonathan M Gibbins",
"Ketan Patel",
"M. Fazil Baksh",
"Sakthivel Vaiyapuri",
"Khalid R Majeed",
"Dhurgham Al-Fahad",
"Hayder Hussein Jalood",
"Haider A Hantosh",
"Mrtatha K Ali",
"Sumiktsal Sakthivel",
"Harry F Williams",
"Jonathan M Gibbins",
"Ketan Patel",
"M. Fazil Baksh"
],
"abstract": "The ABO blood type has been reported to be associated with several diseases such as hepatitis and malaria. Recently, some studies have reported that people with O blood type are protected against COVID-19, while people with A blood type are more susceptible to contract this disease. Here, we analysed data from 5668 COVID-19 patients along with the same number of control samples in a study population in Iraq. Our analysis confirms that people with O blood type are protected partially against COVID-19. Notably, we demonstrate that people with RhD- are more susceptible to contract COVID-19 than people with RhD+ blood type. The blood types are associated with some clinical symptoms such as headache and asthenia of COVID-19, but there is no association with other symptoms. There is no association between blood types and deaths among COVID-19 patients. This study suggests that in addition to ABO, RhD blood type influences the susceptibility to contract COVID-19. Overall, we conclude that susceptibility/protection against COVID-19 may not be determined based only on blood types among the global population as this might vary based on a number of other factors such as ethnicity, geographical locations, occupation and the level of exposure to infected people.",
"keywords": [
"COVID-19",
"ABO",
"RhD",
"Blood type",
"Risk factors",
"Blood group",
"Red blood cells",
"Iraq"
],
"content": "Introduction\n\nCOVID-19 is caused by the newly identified severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), which initiated a pandemic in early 20201,2. Following initial infections in Wuhan, Hubei Province of China, it has affected almost all countries in the world, resulting in a significant number of deaths (over 1.1 million until October 2020) and increased economic burden worldwide. The primary symptoms of COVID-19 include a fever, persistent cough, dyspnoea, thrombosis, myalgia, fatigue, and, in some cases, decreased white blood cell count and severe pneumonia, which requires mechanical ventilation and other intensive care support. Although the relationship between ABO blood types and various diseases is poorly understood, blood types have been reported to be associated with several diseases caused by organisms such as SARS-CoV-13, Helicobacter pylori4, hepatitis B virus5, Norwalk virus6, rotavirus and dengue virus, as well as malaria7. Similarly, the potential importance of ABO blood types in relevance to contracting COVID-19 has been sparsely reported. While some studies found no significant correlation between blood types and the impact of COVID-198–10, others have reported that people with O blood type are partially protected against COVID-19 infection and people with A or AB blood type are more susceptible to this disease11–16. This could be due to the presence of anti-A17 and/or anti-B antibodies in the circulation and their ability to bind to the spike proteins of the virus, thereby preventing them entering cells and proliferating. Notably, a genome wide association study has confirmed the potential involvement of the ABO blood group system in contracting COVID-1918. Although few studies have specifically established an association between RhD blood type and susceptibility to COVID-19, a recent study reported a high risk association between RhD+, COVID-19 infection and resulting deaths13. However, the robust correlation between the blood types specifically RhD and COVID-19 infection has not yet been fully established. Hence, we explored the potential association between blood types, particularly RhD and COVID-19 in a study population of Iraq.\n\n\nMethods\n\nThe aim of this study was to analyse the relationship between the blood types and COVID-19 associated complications in a small population of Iraq. Therefore, in this study, we collected data (i.e. blood group, primary symptoms, previous risk factors, age, gender and outcomes, severity level of COVID-19) from the records of 5,668 COVID-19 patients (confirmed as positive by real-time polymerase chain reaction and this was considered as the key criteria for eligibility to be included in this study) who were admitted to Al-Hussein Teaching Hospital, Thi-Qar and Alkarama Teaching Hospital, Wasit in Iraq between March and June 2020. Blood type data, were collected from the same number of control (non-COVID-19) individuals from the same hospitals (using their records database) and matched by age and gender to the cases.\n\nThe sample size was based on the number of COVID-19 patients admitted in these two hospitals during the study period. The accuracy of data was thoroughly checked by healthcare professionals working in these hospitals and the authors prior to analysis. The individuals who were involved in data collection did not directly analyse the data. The data analysis (between COVID-19 and non COVID-19 groups) was performed by different authors using anonymised data in order to avoid any potential bias. All the statistical analyses were performed using Pearson’s Chi-square test and logistic regression models in R statistical package (www.r-project.org).\n\nThis study design, data collection and consent were approved by the ethical committee (Ref no: 00563477) at Thi-Qar Health Directorate, the Ministry of Health, Republic of Iraq, and all the data were anonymised prior to analysis. Informed verbal consent was obtained from all patients to use their records, and where patients were minor (below 18 years old), consent was obtained from their parents or legal guardians. Verbal consent instead of written consent was obtained as the hospitals were very busy, as a result of the pandemic. Appropriate permissions from the hospitals were also obtained to use the anonymised data for this study.\n\n\nResults\n\nThe COVID-19 population studied here consisted of 3691 (65.2%) males and 1977 (34.8%) females. According to age group: ≤10 years, 267 (4.7%); 11-20 years, 511 (9%); 21–30 years, 1781 (31.4%); 31–40 years, 1221 (21.5%); 41–50 years, 870 (15.3%); 51–60 years, 569 (10%); 61–70 years, 273 (4.8%); 71–80 years, 138 (2.4%); 81–90 years, 32 (0.6%); 90+ years, 6 (0.1%) (with the highest age being 100). The median age of the population was 33 years, and the most infected age group in this population was 21–30 years old.\n\nAmong COVID-19 patients, 1572 (27.7%) had blood type A (A+ = 1493; A- = 79), 1880 (33.2%) had B (B+ = 1653; B- = 227), 645 (11.4%) had AB (AB+ = 511; AB- = 134) and 1571 (27.7%) had O (O+ = 1495; O- = 76). The control population (5668, matched for age and sex) had a similar ratio of blood types [1454 (25.7%) people had A (A+ = 1405; A- = 49); 1792 (31.6%) had B (B+ = 1692; B- = 100); 411 (7.3%) had AB (AB+ = 366; AB- = 45); 2011 (35.5%) had O (O+ = 1976; O- = 35)]. Among the control population, O+ was the most common blood type, and O- was the least common type with others being small population groups except A+ and B+.\n\nThere is very strong evidence (X2 = 209.51, df = 8, p <0.0001) for association between blood groups and COVID-19 infection among the study population. Exploring this association further demonstrates that the proportion of COVID-19 patients with O+ blood type is significantly lower than the control population, suggesting that this blood type is protective against this disease. In contrast, A-, AB-, AB+, B- and O- have shown increased susceptibility to COVID-19 compared to the control population (Figure 1A). Notably, COVID-19 patients with A+ and B+ blood types did not show any significant difference in the proportion infected when compared to the control group (Figure 1B). The risk of contracting COVID-19 was higher among RhD- patients compared to the RhD+ patients [OR = 2.38, 95% CI (2.03, 2.79), p= 0.0001]. There is no evidence for association between gender and blood type among COVID-19 patients (X2 = 4.97, df = 7, p = 0.664).\n\n(A) the frequency of different blood types in the study population. (B) the proportions of controls and COVID-19 cases within different blood types. (C) the Sankey plot shows the relationship between blood types and severity of COVID-19 among the study population.\n\nThe Sankey plot (Figure 1C) illustrates the association between different blood types and the severity of COVID-19 among patients (the severity level was classified according to the guidelines provided by the Ministry of Health in Iraq). In total, there were 77 patients classified as in ‘critical’ stage, and all of them died in hospital during treatment. A total of 90 patients were in the ‘severe’ category, of which only 4 recovered while the others died in hospitals. A further 4000 were in the ‘moderate’ category and 1501 patients were listed as ‘mild’. Based on our analysis, there was no significant association between blood types and the severity of disease among these patients.\n\nThere is evidence for the association of some clinical symptoms with certain blood types among COVID-19 patients when adjusted for age, gender, severity of disease and risk factors. For example, asthenia is significantly lower in people with blood type B+ compared to individuals with O+ [OR = 0.98, 95% CI (0.96, 0.99), p= 0.004]. Similarly, experiencing a headache was significantly lower in people with blood type AB+ compared to people with O+ [OR = 0.97, 95% CI (0.95, 0.99), p= 0.012]. It is noteworthy that the size of these effects is very small. There is no significant correlation between other clinical symptoms (such as fever, chills, cough, dyspnea, anosmia ageusia, loss of appetite, muscle ache, cyanosis, rhinorrhoea, sore throat, diarrhoea, nausea and vomiting) and blood types.\n\nIn total, 176 deaths were recorded in this study among COVID-19 patients. However, there is no evidence for association of blood types, including RhD, with death due to COVID-19 when adjustments were made for age, gender and risk factors (X2 = 0.037, df = 7, p = 0.999). The 172 patients who died were reported to have various preexisting health conditions as risk factors. For example, 63 patients had hypertension, 19 had type 2 diabetes and hypertension, 33 had only diabetes, 1 had type 2 diabetes with acute renal failure, 1 had type 2 diabetes with atherosclerosis and another had diabetes with chronic kidney disease. Notably, 18 patients had chronic obstructive pulmonary disease (COPD), 3 had COPD with diabetes, and one had COPD with hypertension. Another 3 patients had chronic kidney disease and 1 had chronic lung disease. In total, 17 patients had hypertension while 4 had acute renal failure and 4 had acute pulmonary conditions. Of the 5668 patients, only 37 were asymptomatic (hence fewer admission in hospitals) and others had one or more notable COVID-19 symptoms.\n\n\nDiscussion\n\nThe ABO blood types are known to be associated with several human diseases3–7. Similar to previous studies12,13, here we report that O+ blood type provides partial protection against COVID-19. While the association between RhD blood type and various diseases has not been fully established yet, RhD+ has been reported to be associated with high infection and death rate among COVID-19 patients in a population in the USA13. In contrast to this previous study13, interestingly, we report the significance of RhD blood type in influencing the susceptibility to contract COVID-19 among this study population; i.e. people with RhD- blood were more susceptible to contract COVID-19 than people with RhD+ blood type. The mechanisms behind this protection/susceptibility are unclear and further research is warranted to unravel the underlying questions. Moreover, we emphasise that the association between blood types (ABO and RhD) and COVID-19 is likely to be based on numerous factors including (but not limited to) ethnicity, geographical location, nature of occupation and the exposure to infected patients.\n\n\nData availability\n\nUniversity of Reading Research Data Archive: Blood types and their relationship with COVID-19 among a study population in Iraq, http://dx.doi.org/10.17864/1947.27819.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Acknowledgements\n\nWe would like to thank Abbas J. Hadi from the Thi-Qar Directorate of Public Health, Al-Hussein Teaching Hospital, Thi-Qar and Alkarama Teaching Hospital, Wasit, Iraq for providing all the data used in this study and their constant support.\n\n\nReferences\n\nWHO: Novel Coronovirus (2019-nCoV) Situation Report - 1. World Health Organisation. Reference Source\n\nPhelan AL, Katz R, Gostin LO: The Novel Coronavirus Originating in Wuhan, China: Challenges for Global Health Governance. JAMA. 2020; 323(8): 709–710. PubMed Abstract | Publisher Full Text\n\nCheng Y, Cheng G, Chui CH, et al.: ABO Blood Group and Susceptibility to Severe Acute Respiratory Syndrome. JAMA. 2005; 293(12): 1447–1451. PubMed Abstract | Publisher Full Text\n\nBoren T, Falk P, Roth KA, et al.: Attachment of Helicobacter pylori to human gastric epithelium mediated by blood group antigens. Science. 1993; 262(5141): 1892–1895. PubMed Abstract | Publisher Full Text\n\nWang DS, Chen DL, Ren C, et al.: ABO blood group, hepatitis B viral infection and risk of pancreatic cancer. Int J Cancer. 2012; 131(2): 461–8. PubMed Abstract | Publisher Full Text\n\nLindesmith L, Moe C, Marionneau S, et al.: Human susceptibility and resistance to Norwalk virus infection. Nat Med. 2003; 9(5): 548–53. PubMed Abstract | Publisher Full Text\n\nLoscertales MP, Owens S, O'Donnell S, et al.: ABO blood group phenotypes and Plasmodium falciparum malaria: unlocking a pivotal mechanism. Adv Parasitol. 2007; 65: 1–50. PubMed Abstract | Publisher Full Text\n\nGöker H, Karakulak EA, Demiroğlu H, et al.: The effects of blood group types on the risk of COVID-19 infection and its clinical outcome. Turk J Med Sci. 2020; 50(4): 679–83. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDzik S, Eliason K, Morris EB, et al.: COVID-19 and ABO blood groups. Transfusion. 2020; 60(8): 1883–1884. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMendy A, Keller JL, Apewokin S, et al.: Is Blood Type Associated with COVID-19 Severity? medRxiv. 2020; 2020.08.11.20172676. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWu Y, Feng Z, LI P, et al.: Relationship between ABO blood group distribution and clinical characteristics in patients with COVID-19. Clin Chim Acta. 2020; 509: 220–223. PubMed Abstract | Publisher Full Text\n\nZhao J, Yang Y, Huang H, et al.: Relationship between the ABO Blood Group and the COVID-19 Susceptibility. medRxiv. 2020; 2020.03.11.20031096. Publisher Full Text\n\nZietz M, Zucker J, Tatonetti NP: Testing the association between blood type and COVID-19 infection, intubation, and death. medRxiv. 2020; 2020.04.08.20058073. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLatz CA, DeCarlo C, Boitano L, et al.: Blood type and outcomes in patients with COVID-19. Ann Hematol. 2020; 99(9): 2113–2118. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHoiland RL, Fergusson NA, Mitra AR, et al.: The association of ABO blood group with indices of disease severity and multiorgan dysfunction in COVID-19. Blood Adv. 2020; 4(20): 4981–4989. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarnkob MB, Pottegård A, Støvring H, et al.: Reduced prevalence of SARS-CoV-2 infection in ABO blood group O. Blood Advances. 2020; 4(20): 4990–4993. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGuillon P, Clément M, Sébille V, et al.: Inhibition of the interaction between the SARS-CoV spike protein and its cellular receptor by anti-histo-blood group antibodies. Glycobiology. 2008; 18(12): 1085–1093. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEllinghaus D, Degenhardt F, Bujanda L, et al.: Genomewide Association Study of Severe Covid-19 with Respiratory Failure. N Engl J Med. 2020; 383(16): 1522–1534. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAl-Fahad D, Vaiyapuri S: Blood types and their relationship with COVID-19 among a study population in Iraq. University of Reading. Dataset. 2020. http://www.doi.org/10.17864/1947.278"
}
|
[
{
"id": "77833",
"date": "09 Feb 2021",
"name": "Sarah Jones",
"expertise": [
"Reviewer Expertise haematology",
"platelets",
"thrombosis",
"endothelial cells"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this brief report, Majeed and colleagues investigate the relationship between blood groups and COVID19 susceptibility and severity. The study analyses data from 5668 COVID19 patients admitted to two hospitals in Iraq, and an equal number of matched controls admitted to the same hospitals during the study period. The study compares the frequency of blood groups in the study population and the proportions of controls and COVID19 patients within each blood type. The results demonstrate that within the O+ blood type, the proportion of COVID19 patients is significantly lower compared to controls, indicating reduced susceptibility to the virus. In contrast, in blood types A-, B-, AB-, O- and AB+ there is a higher proportion of COVID19 cases compared to controls, suggesting increased susceptibility. No relationship was found between blood group and disease severity. The methods are clearly reported, and the data appears sound. The report would benefit from a demographics table or information specifically relating to ethnicity. It has been established that ethnicity impacts susceptibility to SARs-CoV-2 and the distribution of blood groups vary between ethnic groups. Does the data represent a difference in ethnicity rather than blood group? And might this explain conflicting results with larger published studies concerning RhD status and COVID19 susceptibility?\nA few minor points to improve clarity:\nIt is clearly stated that O+ was the most common blood type in the control population, it would be useful to have a similar statement for the COVID19 group, stating the most common blood type.\n\nAlso, to improve clarity it would be useful to present the data comparing the percentage of patients that are RhD- in each group. It appears to be more than double in the COVID19 group (Control 4%; COVID19 9.1%).\n\nThe final statement, emphasising the factors likely to be responsible for the associations observed between the blood types and COVID19 requires further clarification and supporting citations. What is the evidence that the proportion of each blood type varies in different occupations or geographical locations? Were the study groups not all from the same geographical location (local to the hospitals)? Presenting the ethnicity information for the study group would also help to address this point.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "259728",
"date": "04 Apr 2024",
"name": "Abeer Shaker El-Moursy Ali",
"expertise": [
"Reviewer Expertise tumor marker",
"IBD",
"COVID-19",
"Nephrology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe abstract and introduction of the manuscript \"RhD blood type significantly influences susceptibility to contract COVID-19 among a study population in Iraq\" appear to be well-written and clearly presented. Here's a breakdown of its strengths and weaknesses: Strengths:\nClearly states the research question: investigating the association between blood types (ABO and RhD) and COVID-19 susceptibility. Mentions the controversy around ABO blood types and COVID-19. Describes the study population, data collection methods, and statistical analysis. Highlights the finding of RhD- blood type being more susceptible to COVID-19. Acknowledges limitations like needing further research to understand the mechanisms.\nWeaknesses:\nDoesn't mention potential biases in the study design (discussed later). Doesn't elaborate on the novelty of the RhD finding compared to existing studies.\nHere are some specific questions to consider for each aspect of the manuscript: Clarity and Presentation:\nIs the work clearly and accurately presented and does it cite the current literature?\nYes, the abstract and introduction are clear and cite relevant literature on ABO blood types and COVID-19. Study Design and Methods:\nIs the study design appropriate and is the work technically sound? Are sufficient details of methods and analysis provided to allow replication by others?\nThe manuscript describes a case-control study design, which can be appropriate for investigating associations. However, limitations like selection bias (choosing participants not representative of the population) and confounding variables (factors affecting both blood type and COVID-19 risk) need to be addressed in the methods section. Statistical Analysis:\nIf applicable, is the statistical analysis and its interpretation appropriate?\nThe abstract mentions using Chi-square tests and logistic regression, which could be appropriate depending on the data. However, a more detailed description of the statistical analysis is needed to assess its appropriateness. Data Availability:\nAre all the source data underlying the results available to ensure full reproducibility?\nThe manuscript doesn't mention data availability. Ideally, the data should be available for scrutiny or re-analysis. Conclusions:\nAre the conclusions drawn adequately supported by the results?\nThe conclusions about the association of ABO and RhD blood types with COVID-19 susceptibility seem to be based on the findings, but limitations of the study design (potential biases) need to be considered when interpreting the results. Overall, the abstract and introduction provide a good foundation for the research, but a more critical evaluation of the methodology and results is necessary\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Partly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-38
|
https://f1000research.com/articles/9-1199/v1
|
05 Oct 20
|
{
"type": "Method Article",
"title": "Multi-parametric characterization of drug effects on cells",
"authors": [
"Yael Paran",
"Yuvalal Liron",
"Sarit Batsir",
"Nicola Mabjeesh",
"Benjamin Geiger",
"Zvi Kam",
"Yael Paran",
"Yuvalal Liron",
"Sarit Batsir",
"Nicola Mabjeesh",
"Benjamin Geiger"
],
"abstract": "We present here a novel multi-parametric approach for the characterization of multiple cellular features, using images acquired by high-throughput and high-definition light microscopy. We specifically used this approach for deep and unbiased analysis of the effects of a drug library on five cultured cell lines. The presented method enables the acquisition and analysis of millions of images, of treated and control cells, followed by an automated identification of drugs inducing strong responses, evaluating the median effect concentrations and those cellular properties that are most highly affected by the drug. The tools described here provide standardized quantification of multiple attributes for systems level dissection of complex functions in normal and diseased cells, using multiple perturbations. Such analysis of cells, derived from pathological samples, may help in the diagnosis and follow-up of treatment in patients.",
"keywords": [
"Microscope imaging",
"Multiparameter scoring",
"Drug effect screening"
],
"content": "Introduction\n\nAdvanced precision medicine enables the use of genetic and proteomic information for the characterization of disease states, based on the correlation with detailed medical records and specific pathological manifestations1–11. Indeed, multi-component “omics” profiling can report a large number of components of the genome, transcriptome, proteome, interactome and metabolome (to name a few), and detect even a small fraction of them, indicating significant changes from normal12–16, yet this fraction may not fully overlap with the currently used repertoire of medical manifestations in complex diseases such as heterogeneous cancers17–26. We address here the possibility of using multi-parametric characterizations of cellular features for identifying novel signatures of functional cell states, and offer quantitative diagnostic, as well as mechanistic measures, of disease progression or suppression following therapy.\n\nIt is widely recognized that meaningful understanding of disease states needs to be achieved in the context of the whole body physiology and tissue functional morphology, yet cell-level functional abnormalities lie at the basis of many pathologies, and may thus be identified by recordable cellular attributes. Developments in quantitative light microscopy, either in two-dimensional or in three-dimensional model settings, has been combined with live cell microscopy and recent adaptation of microfluidics technologies to screening and personalized treatment optimization with primary cells, spheroids, organoids and tissue biopsies27–33. The difficulty in quantitative characterization of multiple cellular features of biological specimens with diverse morphological and molecular properties creates an urgent need for methodologies that can be standardized, yield strong statistical scores and can be automated for effective application in systems-level biomedicine.\n\nThe value of multi-parametric analysis is well recognized in flow cytometry34, and is mandatory when the strategies aiming at functional screening depart from cell reporters for specific drug-target interactions. The application of multi-parametric analysis to cell screening has become a common place. Perlman et al.35,36 have designed a titration-invariant similarity score (TISS) based on multi-parametric doze response matrix for each drug, allowing comparison of drug effects to reflect similarity of mechanisms of action. Classification of patterns displayed by tagged proteins in cells37–43 is a powerful way to sort sub-cellular localizations, but is bound to cellular responses reflected by the labeled epitopes (e.g. nucleus vs. cytoplasm localization). Tanaka et al.44 based their analysis on multi-parameter evaluations, using Principal Components Analysis (PCA) to reduce the number of independent parameters so that the multi-dimensional state vectors of cells treated with drugs could be displayed in 3D plots to differentiate or correlate effects and infer similar mechanisms. Melnick et al.45 obtained their multi-parametric data vectors from 35 tyrosine-kinase-activated cellular assays responding to 1400 kinase inhibitors, and used clustering in Euclidian space to sort perturbations similar in action to known drugs. Screens have also been developed to dissect cellular mechanisms42,46 and identify proteins involved in cellular processes using interference RNA perturbations43,47. Rather than prior definition of the type of effect (e.g. cell death) or a marker to a specific function, the analysis developed here quantifies drug effects on cell phenotypic and molecular attributes in high dimensionality space.\n\nIn this study we have determined the effects of the NCI COMBO drug collection [48, Plate number 3948] on five cell lines; the well-established cervical carcinoma-derived HeLa cells and four cell lines derived from bladder cancers49, and used light microscopy-based screening to record multiple cellular responses to the tested drugs. Parameters such as total protein levels, cytoplasmic-nuclear distributions and cell death were analyzed at low magnifications (X10/0.25 objectives)50, while high-definition imaging, providing detailed intracellular data about fine protein localization (e.g. cytoskeletal fibers, sub-cellular organelle morphology, etc.) used higher power objectives (60x/.95), providing multi-dimensional information needed to link drug responses with molecular mechanisms and cellular functions.\n\nIn order to be able to read properties of cells and measure as many aspects of system-level response to the treatments, we developed an analysis platform for multi-parametric characterization of cellular features and unbiased scores, based on Mahalanobis distances, for identifying the potency of drugs producing a wide variety of effects. The multi-parametric score is a single value representing the difference from control cells, that contain, however, the complete information on the individual contributions of each of the measured parameters, enabling us to identify those that are mostly affected by each perturbation. Furthermore, the score allows quantification of time-dependencies, examination of cell-line differential responses, and prediction of expected synergy of drug combinations.\n\n\nMethods\n\nHeLa, and four cell lines derived from bladder cancers (UMUC3, TCCSUP, HT1376 and RT41) were obtained from the ATCC. Cell lines were maintained in DMEM supplemented with 10% FBS (Sigma-Aldrich, Rehovot, Israel) and penicillin/streptomycin antibiotics. Cells were grown in polystyrene petri dishes.\n\nDrugs were obtained from the National Cancer Institute (NCI). The COMBO plate number 394848 includes 77 compounds, 23 of which are FDA-approved anti-cancer drugs. Mechanisms of action include anti-metabolite activity, tubulin binding, DNA damage, Hsp90 binding, as well as inhibition of topoisomerases, kinases, the proteasome, angiogenesis, ion channels, palmitoylation, phosphodiesterase, cyclooxygenase, and aromatase.\n\nCells were suspended and transferred to 384 well plates (thin plastic bottom, Cat #781091 Greiner-Bio One, D-72636 Frickenhausen) using BioMek FX liquid handling robot (Beckman Coulter, Fullerton, CA 92834) at a density of 1000 cells/well and cultured for one day. Drugs were then added at the ten three-fold dilutions, concentration range: (10-5x10-4)*GI50, (GI50 is the concentration of the least sensitive line in the NCI60, see Extended data: Table S151). Following the specified incubation time with the drugs, cells were fixed, labeled with DAPI for nuclei (Life, Molecular Probes D1306), FITC-Phalloidin for F-Actin (Life, Molecular Probes F432) and indirectly immunolabeled for tubulin (Primary Antitubulin antibodies (SIGMA T6199)) and Cy3-labled secondary antibodies (Alexa Fluor546 –Life, Molecular Probes A11030), and washed by the robot. For drug effect measurement, ten 3-fold dilutions in duplicates were arranged in plate rows, starting from 50 times the median effect concentration specified for the COMBO collection. The transfer order and concentrations from the 96-well COMBO drugs plate into six 384-well plates for each of the 5 cell lines are listed in Extended data: Table S151.\n\nA total of 16 binary combinations were selected from 20 drugs that showed cellular effects. Matrices with two-fold dilutions in duplicate with drug concentrations up and down from the median effect concentrations, as listed in Table 1, were prepared for two cell lines (TCCSUP & UMUC3). Four such matrices were arranged in one 384-well plate, and rearranged for presentation, including duplicate averaging, as “virtual plates”.\n\nMedian-effect values were fitted to the scores in Figure 1 using the equation in the Methods. The table lists Dm the column number of half effect for the drug rows marked by arrows in Figure 1. The higher the Dm number is, the more effective the drug is at higher dilutions. The corresponding concentration for a specific drug is 10*GI50 *3^(1-Dm/2) (3-fold dilutions in duplicates from initial 50 fold dilution of the COMBO supplied concentration, set as 500 times the drug GI50).\n\nPlate scanning and multi-color image acquisition was performed by WiScan Argus (Idea-biomedical, Rehovot 76705, Israel; https://idea-bio.com) a fast, high-resolution screening microscope system52–54. Images were stored locally during the screen, and transferred to 60 TeraBytes storage [NetApp, Sunnyvale, CA 94089] accessible to the analysis workstations via fast local Ethernet for visualization of raw images, interactive optimization of the analysis strategy and tuning of user-controlled parameters (see below). The automated analysis pipeline was then run, accumulating analyzed results to a data base, allowing results display and interpretation.\n\nIn order to process the images, we used a computer cluster (Sun Microsystems), consisting of AMD dual-core computer nodes running Linux RedHat. The software used was WiSoft Minerva (Idea-biomedical).\n\nThe analysis scripts pipeline (Extended data55; scripts can be used with any software) combines throughput with modularity. Analysis of different specimens and diverse assays requires a wide range of alternative algorithms and interactive capability to select between them and optimize user-defined parameters before processing the whole data. To achieve such flexibility, we broke the analysis into modular steps categorized by their functionality (Pre-processing, Segmentation and Quantification see Extended data: Figure S151). Sequences of the analysis modules can be integrated like in a jigsaw puzzle and looped in cycle on all images acquired in an experiment. While such modular structure typically compromises performance, we implemented fast communication of images and data between modules via Shared Memory, achieving processing time as fast as would have been the processing time of an integrated optimized program. Every image is annotated during the acquisition (time, fluorescent color or transmitted light, position inside well, well in plate, plate in whole experiment), and the analyzed data carries these annotations, keeping track of the experimental organization (metadata). Calculation and display of statistics of the analyzed results is therefore directly linked with the experimental design structure. The “Plate GUI” is used to display plate-wide scores, and to interactively show (by a mouse click on a well) the original images and the corresponding numerical and statistical analyzed data. The access to all levels of the data is necessary for rational mining of the TeraBytes of digital information at all levels: original images, montages of image tiles, segmented objects, quantified parameters and statistical profiles. The scoring algorithm used here is based on Mahalanobis distances in multi-parametric space (see Extended data: Figure S251). The attributes calculated for each cell are listed in Extended data: Table S251.\n\nCombination matrix scores were fitted using Loewe-additivity and Bliss-independence models56–58 using the following steps:\n\nI. Fitting Dm and s in Chou’s medial effect equation to the single-drug response curve (first rows and columns in the combination matrices):\n\nAf+B=A/[1+(Dm/D)s]+B\n\nwhere:\n\nf the fractional effect, here from the Mahalanobis score,\n\nA+B the measured score amplitude at infinite drug concentration,\n\nB the score baseline at zero drug concentration.\n\nDm the median effect concentration of the drug\n\ns the Hill parameter\n\nThe parameters A,B were first evaluated from the scores response curve minimum and maximum, Dm and s were obtained from the linearized equation for the log of normalized fractional effect data: log(1/f-1)=s*log(Dm/D)\n\nThe four parameters A,B,Dm,s were then used as initial estimates for a non-linear Marquet-Levenberg fit.\n\nII. Using the parameters s1,2 for each of the two drugs to “span” the combination effect matrix, fcomb, as a function of the two concentrations, D1,2, and solving fcomb from Loewe additivity equation:\n\nD1/[fcomb/(1-fcomb)1/s1] + D2/[fcomb/(1-fcomb)1/s2] = 1\n\nIII. Using the medial effect parameters Dm1,2,s1,2 for each of the two drugs to “span” the combination effect matrix, fcomb, assuming Bliss independence:\n\nfcomb=f1*f2=1/[1+(Dm1/D1)s1] /[1+(Dm2/D2)s2]\n\nIV. Fitting the four parameters Dm1,2,s1,2 using all the values in the combination matrix simultaneously, based on Loewe or Bliss models.\n\n\nResults\n\nFive established cell lines; four derived from urinary-bladder cancers, and one from cervical carcinoma, were plated in 384-well plates at a density of 1000 cells per well. Following incubation for 24h, the cells were treated with serial dilutions of the COMBO plate drugs48 (ten, three-fold dilutions, concentration range: (10-5×10-4)*GI50 , where GI50 is the concentration of the least sensitive line in the NCI60, see Extended data: Table S151) and further incubated for 18h (see Methods). The cells were then fixed and labeled (see Methods). Images were acquired in these three fluorescent channels (see top panels, Figure 1), and analyzed as described in the Methods and in Extended data: Figure S151. The analysis yielded, for each well, a list of segmented cells, each characterized by a “vector” of quantified features (‘attributes’).\n\nFive cell lines were cultured each in six 384 plates treaded with concentration-dilution series of the COMBO drug plate (see Table 1 for drug order, each plate row is a dilution series in duplicates). Cells were imaged in three fluorescent colors (Red: Acin, Green:Tubulin & Blue:Nucleus, See image examples in top pannels). Individual cells were segmented, and for each cell an array of attributes was quantified. Attributes include cells and nuclei morphological and fluorescence intensities attributes and microtubules and actin fiber attributes (see Methods and Extended data: Table S2). Color coded Mahalanobis Scores (see Extended data: Figure S2) are presented here for each well in six plates for the five cell lines. A number of drugs with strong cellular effects are displayed (red rows), and indicate some cell-line dependence. Robot error effects were minimized in the analysis by rejecting outliers lying outside the robust PCA ellipsoids.\n\nWhole cell segmentation was commonly achieved by diffuse cytoplasmic staining (e.g. “cell mask”, Invitrogen), with enhanced nuclear concentration. Cytoplasmic fluorescence (due to the monomer fraction of tubulin and actin and even cell autofluorescence) plus the definition of the nucleus by DAPI (often excluding cytoskeletal proteins) provided highly reliable segmentation of individual cells, even in densely-packed cell islands. This approach enabled us to ‘free’ color channels for labeling cells for additional cellular markers of interest. In addition, the segmentation process allowed us to also calculate the total covered cell area which can report on cell spreading/proliferation/death during the time of treatment of each of the tested cell lines. Notably, although all lines are of epithelial origin, the RT4 cells (and to some extent HT1376) grow in islands even at low plating density, while TCCSUP and UMUC3 cells grow, primarily as isolated cells, and may be more susceptible to shrinking or contraction. For this reason, the analysis is based primarily on “per-cell” attributes.\n\nThe profiling of drug responses based on a single attribute may be misleading, as demonstrated in Figure 2. Dose-response profiles of a single drug (nocodazole) is displayed by eight different attributes (cell area, MT intensity/ area/ texture, actin intensity/ area, nucleus intensity/ area) for two cell lines (TCCSUP, UMUC3). The results are displayed as a “virtual plate”. Various attributes display different responses: tubulin-related attributes in the first four rows indicate the disruption (Blue) of the microtubule network; attributes related to the actin cytoskeleton show little effect after 18 hours; nucleus-related attributes show an increase (Red) in DNA content per cell, possibly due to cell-cycle arrest.\n\nTCCSUP and UMUC3 cell lines were treated by nocodazole. Image analysis recorded nuclei, cells microtubule and actin-related scores. The response curves clearly depend on the attribute used to quantify the effects, and the strength of the effect depends on the cell line.\n\nIn order to define an unbiased score that will accumulate drug-induced changes from multi-parametric characterization of cell properties, we have to consider three factors: (1) different attributes have different dimensions and scales; (2) the cell-by-cell variability of each attribute can be quite different; (3) it is not easy to select independent attributes. For example, cell content of a specific detected protein (either by expression of fluorescent derivative or by specific immune-labeling) is highly correlated with cell size. While average protein concentration accounts for this dependency, there are less obvious correlations, reflecting regulated cellular properties, which may become relaxed or tightened in response to drug treatments.\n\nPCA is a well-established method for the characterization of multi-dimensional variability and correlations. We have applied PCA to the quantified attributes in the control wells to best fit a multi-dimensional hyper-ellipsoid, where elongated axes indicate correlated attributes. Using this hyper-ellipsoid, we defined a score based on Mahalanobis distances between each treated cell to the control cells (see Methods and Extended data: Figure S251). This score balances the scales and variability of the measured attributes and accounts for correlations between them. In addition, the hyper-ellipsoid for each of the treated wells provides a multi-parametric scale for identifying outliers, created by technical artifacts such as bubbles, dead cells, cell clumps etc. Figure 1 shows the multi-parametric Mahalanobis scores for all the COMBO plate drugs applied to the five cell lines tested. The figure compiles 360 Gigabytes of image data (12GB/plate) imaged from 30 multi-well plates. The scores coded in spectral colors are displayed in the plate format that is also used as a “Graphic User Interface” (Plate GUI) showing, upon a mouse-click on a well, montages of the acquired images, outlined cell segments, as well as the full resolution image color components, and a wealth of statistical presentations and cell-by-cell numerical values, facilitating data mining within the large volume of information (see Extended data: Figure S151).\n\nThe scores display differential responses to some of the drugs for some of the cell lines. Table 1 lists the AC50 values obtained for these drugs, as described in the Methods. It should be noted that defined as a “distance”, Mahalanobis scores are always positive. The multi-parametric score is a faithful reporter of changes in any of the measured attributes for all drugs that showed effects. Moreover, once an effect is detected by scoring large Mahalanobis distance to the control ellipsoid, the largest contributions of each of the attributes to the score identify the most significant attribute changes (see examples in Table 2). The plate scores for individual high contributors were qualitatively similar (though not identical) to the multi-parametric Mahalanobis scores, yet the latter present in one picture what can only be exhaustively reviewed by many single attribute scores. Altogether, this offers a fast and systematic method with internal standardization for navigating in multi-dimensional attribute space towards focusing on the cellular phenotype responding to perturbations.\n\nFigure 3 depicts another important feature of cellular responses to drug treatments, namely, time dependence. The effect of drugs with specific molecular target is typically documented using reporter cells displaying target activation. However, the medical value of many drugs often stems from indirect effects. Multi-parametric cell-based measurements can probe both direct and indirect effects. Here the fast disruption of microtubules compared to the slower cell cycle arrest (depicted by DAPI-labeled DNA content) can be seen. Fingerprints of drug effects, based on an array of features or time-dependent of a single feature, suffer from different scales and variabilities of the measured features. The Mahalanobis-scaled fingerprints can include many features at several times, and present more balanced measures for analyses such as multiparametric similarity of drug effects.\n\nListed are examples for three drugs, Curcubitacin, Helenalin and Nocodazole, that strongly affected TCCSUP cells as seen in Figure 1 Plate 1 row 9, plate 3 row 11 and plate 6 row 12 respectively.\n\n(A) Curcubitacin-treated TCCSUP indicate rounding up compared to the control cells (Shrinking cells long axis, compare to control cells, the top of Figure 1), loss of actin stress fibers (strongly stained condensed actin), and longer, strongly labeled tubulin fibers.\n\n(B) Helenalin-treated cells show stronger total per-cell staining of actin & tubulin fibers, and also stronger DAPI staining, compatible with cell cycle arrest.\n\n(C) Nocodazole-treated cells indicate loss of tubulin fibers with no disturbance of actin filaments.\n\nAll the above descriptions based on image visualization are compatible with the three mostly-contributing attributes to the Mahalanobis distance from controls, listed below. Other attributes have normalized Mahalanobis contributions of less than 0.2.\n\nThree plates were treated with 6 drugs for 3 times: 1, 5 & 18h (shown only Nocodazole for 2 cell lines for several attributes). The “virtual plate” displays in rows the different attributes evaluating cellular effects at different time points.\n\nTime-response fingerprinting based on a single attribute depends on the attribute chosen, and is not reliably reporting time-dependence of effects. For example, fast (minutes) disruption of microtubules, indirect effects on the actin filaments (an hour), changes in cell spreading, slower cell-cycle arrest (many hours) and cell death (days) are only reported by several time points.\n\nIn conclusion, multi-parametric analysis of cellular responses to drugs, visualized by high-definition light microscopy screening, allows departing from target-guided drug screening and build essays more sensitive to cell-level functional effects. In addition, since systematic screen of all drug combinations is not practical, selection based on multi-parametric analysis may increase the chance of identifying interactions between different cellular mechanisms favorably affected by drug combinations. In Figure 4 we show matrices of the responses of one combination (Radicinin and Brasilin) out of 16 binary combinations tested at five 2-fold dilutions around the “AC50” concentration of the single drug treatments. Differences between the two treated cell lines are seen for nuclear-related features. However, Loewe or Bliss model fitting to the Mahalanobis scores does not indicate synergism.\n\nThe combinations were selected from 20 \"responsive\" drugs (Table 1). 16 two-drug combinations matrices were applied to two cell lines (TCCSUP & UMUC3). 5x5 matrices (10x5 wells including duplicates) + single-drug rows and columns + controls were prepared. 4 dual combinations were set in each plate, 8 plates total (4 for each cell line).\n\nExample of 7 single features response for the combination of Radicinin (rows) and Brasilin (columns) in concentrations of 4-0.25µM are displayed for the two cell lines (TOP). The data require preparation of larger matrices with more dilutions and repeats. Nevertheless we attempted to fit the Mahalanobis scores (BOTTOM) to Loewe or Bliss models (see Methods) but found no significant synergism.\n\n\nDiscussion\n\nImages recording cell, nuclear and cytoskeleton structures were analyzed for five cell lines treated with the COMBO drug library and revealed differential responses. Cell morphology and tagged protein intensity and distribution attributes depicted cellular responses as reported by various features. The transition with time from direct and specific effects on the drug target to indirect and distributed responses manifest the advantages of high-resolution imaging for characterization of cell-level responses by multi-parametric scores. Identification of the features mostly affected by each treatment help guide to relevant cellular functions mediating these responses.\n\nBoth basic biological research and biomedical challenges need high-throughput technologies to address the complexity of cellular mechanisms. The information-content in high-throughput experiments, initially confined to biochemical assays reporting a specific target protein, is extending to detailed quantitative sub-cellular characterization, classically the basis for understanding cellular functions. This work developed analysis for high-resolution (recording cytoskeletal fibers) multi-color cellular images from large-scale screens, and display of the results in a digestible form with cell-informatics mining capability. Multi-parametric quantification of cellular effects offers a method to identify cellular responses to drugs without selecting in advance a specific drug targets as a reporter. Effective AC50 obtained from multi-parametric analysis integrate many cell-level responses and is therefore less essay-biased then measurements based on single-attribute analysis. Scores based on Mahalanobis distance offer standardized measures of responses in comparison to controls.\n\nPerturbations to living systems, even when the treatment is directed towards a highly specific target with well-defined short-term effects, cause at longer times distributed cellular responses. Characterization of the time-dependence of the direct and indirect responses to drugs using multi-parametric analysis contain causal information to help dissect mechanisms, and may provide rational basis for optimization of temporal schedules for drug administration, already recognized valuable in chemotherapy51.\n\nMulti-component drug combinations are commonly used in chemotherapy, AIDS and also in antibiotic treatments to fight development of resistant bacteria clones. The effectiveness of such cocktails is often a result of interactions at the physiological level, and was optimized through slow clinical trials. Redundancies and multi-functionalities in molecular networks mediating cellular mechanisms60–62 and the multi-genetic nature of diseases such as cancer, with abnormal protein interactions and modified cellular mechanisms, suggest a potential for the discovery and optimization of multiple-perturbation approaches using cell preparations. Many classical drug targets are shared by the disease phenotypes and the normal ones, causing undesired side effects at the potent dozes. Multiple-drug treatments offer degrees of freedom to optimize synergistic interference with a specific mechanism, preferably more effective in the diseased version and in a given cell lineage, with reduced side effects due to lower doses of each individual drugs in the mixture and minimize compensation by alternative pathways. The benefits of compounds with lower affinity and multiple targets is therefore re-evaluated due to their potential in combinatorial therapies63–67.\n\nThe advancement in our understanding of the molecular mechanisms of cellular functions serves as a basis for rational design of drugs. Combinatorial drug effect profiling is in fact a tool to probe normal network structures60–62. However, first principles quantitative modeling of complex cellular networks depends on many and not readily measurable parameters. Cell-based high-content drug response profiling is therefore a practical method to study multiple perturbations, and to optimize mixtures with advantageous responses. Cell-based assays also open possibilities for patient-specific profiling of abnormal pathways and personal optimization of treatments.\n\n\nData availability\n\nThe raw data includes 10 Terabytes of images and are therefore too large to share. The images are kept on storage disks as OME-TIFF files with .xml headers metadata for each image. Individuals wishing to access this data can apply to the corresponding author in order to obtain the files. In this situation, these files will be uploaded to and available from the Weizmann Institute of Science website. Example images for one plate are provided in the Extended data67.\n\nFigshare: Multi-parametric characterization of drug effects on cells, https://doi.org/10.6084/m9.figshare.12981518.v155.\n\nThis project contains the following extended data:\n\n- Table S1. The order of transfer from the COMBO 96-well plates and the COMBO plate handout.\n\n- Table S2. The list of attributes calculated for each cell.\n\n- Figure S1. Analysis Flow Diagram.\n\n- Figure S2. Mahalanobis distance score.\n\nFigshare: Multi-parametric characterization... Original Data, https://doi.org/10.6084/m9.figshare.12981875.v155.\n\n- Analysis scripts\n\n- Example images for one plate\n\n- Analyzed data for the example images (Excel file)\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).\n\n\nNotes\n\n1RT4 cells were received to the lab in November 2006 from the ATCC. RT4 cells are much smaller than HeLa and have different morphology. We could clearly see that there was no HeLa contamination in our RT4 cells.",
"appendix": "Acknowledgements\n\nWe thank the NCI for supplying the Combo drugs plate. Pierre Choukroun built our computer grid, and helped us continuously to develop and maintain our computation environment.\n\n\nReferences\n\nZeggini E, Gloyn AL, Barton AC, et al.: Translational genomics and precision medicine: Moving from the lab to the clinic. Science. 2019; 365(6460): 1409–1413. Publisher Full Text\n\nGoto T: Patient-Derived Tumor Xenograft Models: Toward the Establishment of Precision Cancer Medicine. J Pers Med. 2020; 10(3): E64. PubMed Abstract | Publisher Full Text\n\nKu SY, Gleave ME, Beltran H: Towards precision oncology in advanced prostate cancer. Nat Rev Urol. 2019; 16(11): 645–654. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoutsouliak K, Veeraraghavan J, Sethunath V, et al.: Towards personalized treatment for early stage HER2-positive breast cancer. Nat Rev Clin Oncol. 2020; ; 17(4): 233–250. PubMed Abstract | Publisher Full Text\n\nHo D, Quake SR, McCabe ERB, et al.: Enabling Technologies for Personalized and Precision Medicine. Trends Biotechnol. 2020; 38(5): 497–518. PubMed Abstract | Publisher Full Text\n\nYadav A, Vidal M, Luck K: Precision medicine - networks to the rescue. Curr Opin Biotechnol. 2020; 63: 177–189. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTsimberidou AM, Fountzilas E, Nikanjam M, et al.: Review of precision cancer medicine: Evolution of the treatment paradigm. Cancer Treat Rev. 2020; 86: 102019. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGambardella V, Tarazona N, Cejalvo JM, et al.: Personalized Medicine: Recent Progress in Cancer Therapy. Cancers (Basel). 2020; 12(4): 1009. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJariyal H, Weinberg F, Achreja A, et al.: Synthetic lethality: a step forward for personalized medicine in cancer. Drug Discov Today. 2020; 25(2): 305–320. PubMed Abstract | Publisher Full Text\n\nHey SP, Gerlach CV, Dunlap G, et al.: The evidence landscape in precision medicine. Sci Transl Med. 2020; 12(540): eaaw7745. PubMed Abstract | Publisher Full Text\n\nMartorell-Marugán J, Tabik S, Benhammou Y, et al.: Deep Learning in Omics Data Analysis and Precision Medicine. In: Husi H, editor. Computational Biology [Internet]. Brisbane (AU): Codon Publications; 2019; Chapter 3. PubMed Abstract | Publisher Full Text\n\nDiaz-Ortiz ME, Chen-Plotkin AS: Omics in Neurodegenerative Disease: Hope or Hype? Trends Genet. 2020; 36(3): 152–159. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoufraqech M, Nilubol N: Multi-omics Signatures and Translational Potential to Improve Thyroid Cancer Patient Outcome. Cancers (Basel). 2019; 11(12): 1988. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDugourd A, Saez-Rodriguez J: Footprint-based functional analysis of multiomic data. Curr Opin Syst Biol. 2019; 15: 82–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYe Y, Zhang Z, Liu Y, et al.: Multi-omic signatures identify pan-cancer classes of tumors beyond tissue of origin. A Multi-Omics Perspective of Quantitative Trait Loci in Precision Medicine. Trends Genet. 2020; 36(5): 318–336.\n\nChen B, Garmire L, Calvisi DF, et al.: Harnessing big 'omics' data and AI for drug discovery in hepatocellular carcinoma. Nat Rev Gastroenterol Hepatol. 2020; 17(4): 238–251. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHausser J, Alon U: Tumour heterogeneity and the evolutionary trade-offs of cancer. Nat Rev Cancer. 2020; 20(4): 247–257. PubMed Abstract | Publisher Full Text\n\nBonin S, Stanta G: Pre-analytics and tumor heterogeneity. N Biotechnol. 2020; 55: 30–35. PubMed Abstract | Publisher Full Text\n\nMarusyk A, Janiszewska M, Polyak K: Intratumor Heterogeneity: The Rosetta Stone of Therapy Resistance. Cancer Cell. 2020; 37(4): 471–484. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVessoni AT, Filippi-Chiela EC, Lenz G, et al.: Tumor propagating cells: drivers of tumor plasticity, heterogeneity, and recurrence. Oncogene. 2020; 39(10): 2055–2068. PubMed Abstract | Publisher Full Text\n\nSenosain MF, Massion PP: Intratumor Heterogeneity in Early Lung Adenocarcinoma. Front Oncol. 2020; 10: 349. PubMed Abstract | Publisher Full Text | Free Full Text\n\nScherer F: Capturing Tumor Heterogeneity and Clonal Evolution by Circulating Tumor DNA Profiling. Recent Results Cancer Res. 2020; 215: 213–230. PubMed Abstract | Publisher Full Text\n\nJaniszewska M: The microcosmos of intratumor heterogeneity: the space-time of cancer evolution. Oncogene. 2020; 39(10): 2031–2039. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKutasovic JR, McCart Reed AE, Sokolova A, et al.: Morphologic and Genomic Heterogeneity in the Evolution and Progression of Breast Cancer. Cancers (Basel). 2020; 12(4): 848. PubMed Abstract | Publisher Full Text | Free Full Text\n\nParker TM, Henriques V, Beltran A, et al.: Cell competition and tumor heterogeneity. Semin Cancer Biol. 2020; 63: 1–10. PubMed Abstract | Publisher Full Text\n\nDagogo-Jack I, Shaw AT: Tumour heterogeneity and resistance to cancer therapies. Nat Rev Clin Oncol. 2018; 15(2): 81–94. PubMed Abstract | Publisher Full Text\n\nMa X, Liu J, Zhu W, et al.: 3D bioprinting of functional tissue models for personalized drug screening and in vitro disease modeling. Adv Drug Deliv Rev. 2018; 132: 235–251. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPuls TJ, Tan X, Husain M, et al.: Development of a Novel 3D Tumor-tissue Invasion Model for High-throughput, High-content Phenotypic Drug Screening. SCientifiC RePorTS. 2018; 8: 13039. Publisher Full Text\n\nHou S, Tiriac H, Sridharan BP, et al.: Advanced Development of Primary Pancreatic Organoid Tumor Models for High-Throughput Phenotypic Drug Screening. SLAS Discov. 2018; 23(6): 574–584. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCzerniecki SM, Cruz NM, Harder JL, et al.: High-Throughput Screening Enhances Kidney Organoid Differentiation from Human Pluripotent Stem Cells and Enables Automated Multidimensional Phenotyping. Cell Stem Cell. 2018; 22(6): 929–940.e4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDamiati S, Kompella UB, Damiati SA, et al.: Microfluidic Devices for Drug Delivery Systems and Drug Screening. Genes (Basel). 2018; 9(2): 103. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEsch EW, Bahinski A, Huh D: Organs-on-chips at the frontiers of drug discovery. Nat Rev Drug Discov. 2015; 14(4): 248–260. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSnijder B, Vladimer GI, Krall N, et al.: Image-based ex-vivo drug screening for patients with aggressive haematological malignancies: interim results from a single-arm, open-label, pilot study. Lancet Haematol. 2017; 4(12): e595–e606. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPerez OD, Mitchell D, Campos R, et al.: Multiparameter analysis of intracellular phosphoepitopes in immunophenotyped cell populations by flow cytometry. Curr Protoc Cytom. 2005; Chapter 6: Unit6.20. PubMed Abstract | Publisher Full Text\n\nPerlman ZE, Slack MD, Feng Y, et al.: Multidimensional drug profiling by automated microscopy. Science. 2004; 306(5699): 1194–1198. PubMed Abstract | Publisher Full Text\n\nPerlman ZE, Mitchison TJ, Mayer TU: High-content screening and profiling of drug activity in an automated centrosome-duplication assay. Chembiochem. 2005; 6(1): 145–151. PubMed Abstract | Publisher Full Text\n\nConrad C, Erfle H, Warnat P, et al.: Automatic identification of subcellular phenotypes on human cell arrays. Genome Res. 2004; 14(6): 1130–1136. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuang K, Murphey RF: From quantitative microscopy to automated image understanding. J Biomed Opt. 2004; 9(5): 893–912. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKiger AA, Baum B, Jones S, et al.: A functional genomic analysis of cell morphology using RNA interference. J Biol. 2003; 2(4): 27. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFraser AG, Kamath RS, Zipperlen P, et al.: Functional genomic analysis of C. elegans chromosome I by systematic RNA interference. Nature. 2000; 408(6810): 325–330. PubMed Abstract | Publisher Full Text\n\nHuh WK, Falvo JV, Gerke LC, et al.: Global analysis of protein localization in budding yeast. Nature. 2003; 425(6959): 686–691. PubMed Abstract | Publisher Full Text\n\nNeumann B, Walter T, Hériché JK, et al.: Phenotypic profiling of the human genome by time-lapse microscopy reveals cell division genes. Nature. 2011; 464(7289): 721–727. PubMed Abstract | Publisher Full Text | Free Full Text\n\nConrad C, Wünsche A, Tan TH, et al.: Micropilot: automation of fluorescence microscopy-based imaging for systems biology. Nat Methods. 2011; 8(3): 246–249. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTanaka M, Bateman R, Rauh D, et al.: An unbiased cell morphology-based screen for new, biologically active small molecules. PLoS Biol. 2005; 3(5): e128. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMelnick JS, Janes J, Kim S, et al.: An efficient rapid system for profiling the cellular activities of molecular libraries. Proc Natl Acad Sci U S A. 2006; 103(9): 3153–3158. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPegoraro G, Misteli T: High-Throughput Imaging for the Discovery of Cellular Mechanisms of Disease. Trends Genet. 2017; 33(9): 604–615. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaddison PJ, Silva JM, Conklin DS, et al.: A resource for large-scale RNA-interference-based screens in mammals. Nature. 2004; 428(6981): 427–31. PubMed Abstract | Publisher Full Text\n\nhttps://dtp.cancer.gov/dtpstandard/servlet/PlateMap?searchlist=3948&outputformat=html&searchtype=plate&Submit=Submit\n\nReuter VE: The pathology of bladder cancer. Urology. 2006; 67(3 Suppl): 11–7; discussion 17-8. PubMed Abstract | Publisher Full Text\n\nFarkash-Amar S, Eden E, Cohen A, et al.: Dynamic proteomics of human protein level and localization across the cell cycle. PLoS One. 2012; 7(11): e48722. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKam Z, Geiger B, Paran Y, et al.: Multi-parametric characterization of drug effects on cells. figshare. Figure. 2020. http://www.doi.org/10.6084/m9.figshare.12981518.v1\n\nLiron Y, Paran Y, Zatorsky NG, et al.: Laser autofocusing system for high-resolution cell biological imaging. J Microsc. 2006; 221(Pt 2): 145–151. PubMed Abstract | Publisher Full Text\n\nParan Y, Lavelin I, Naffar-Abu-Amara S, et al.: Development and application of automatic high-resolution light microscopy for cell-based screens. Methods Enzymol. 2006; 414: 228–247. PubMed Abstract | Publisher Full Text\n\nParan Y, Ilan M, Kashman Y, et al.: High-throughput screening of cellular features using high-resolution light-microscopy; application for profiling drug effects on cell adhesion. J Struct Biol. 2007; 158(2): 233–243. PubMed Abstract | Publisher Full Text\n\nKam Z: Multi-parametric characterization... Original Data. figshare. Dataset. 2020. http://www.doi.org/10.6084/m9.figshare.12981875.v1\n\nChou TC, Talalay P: A simple generalized equation for the analysis of multiple inhibitions of Michaelis-Menten kinetic systems. J Biol Chem. 1977; 252(18): 6438–6442. PubMed Abstract\n\nChou TC: Derivation and properties of Michaelis-Menten type and Hill type equations for reference ligands. J Theor Biol. 1976; 59(2): 253–276. PubMed Abstract | Publisher Full Text\n\nChou TC: On the determination of availability of ligand binding sites in steady-state systems. J Theor Biol. 1977; 65(2): 345–356. PubMed Abstract | Publisher Full Text\n\nWu J, Waxman DJ: Immunogenic chemotherapy: Dose and schedule dependence and combination with immunotherapy. Cancer Lett. 2018; 419: 210–221. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKeith CT, Borisy AA, Stockwell BR: Multicomponent therapeutics for networked systems. Nat Rev Drug Discov. 2005; 4(1): 71–8. PubMed Abstract | Publisher Full Text\n\nYeh P, Tschumi AI, Kishony R: Functional classification of drugs by properties of their pairwise interactions. Nat Genet. 2006; 38(4): 489–494. PubMed Abstract | Publisher Full Text\n\nYeh P, Kishony R: Networks from drug-drug surfaces. Mol Syst Biol. 2007; 3: 85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFitzgerald JB, Schoeberl B, Nielsen UB, et al.: Systems biology and combination therapy in the quest for clinical efficacy. Nat Chem Biol. 2006; 2(9): 458–466. PubMed Abstract | Publisher Full Text\n\nLee MS, Johansen L, Zhang Y, et al.: The novel combination of chlorpromazine and pentamidine exerts synergistic antiproliferative effects through dual mitotic action. Cancer Res. 2007; 67(23): 11359–67. PubMed Abstract | Publisher Full Text\n\nLehár J, Krueger A, Zimmermann G, et al.: High-order combination effects and biological robustness. Mol Syst Biol. 2008; 4: 215. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVerissimo CS, Overmeer RM, Ponsioen B, et al.: Targeting mutant RAS in patient-derived colorectal cancer organoids by combinatorial drug screening. eLife. 2016; 5: e18489. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPemovska T, Bigenzahn JW, Superti-Furga G: Recent advances in combinatorial drug screening and synergy scoring. Curr Opin Pharmacol. 2018; 42: 102–110. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "72645",
"date": "26 Oct 2020",
"name": "John Lock",
"expertise": [
"Reviewer Expertise Systems Microscopy of cancer cell biology",
"including signalling",
"cytoskeletal regulation",
"adhesion and migration biology. Phenotypic drug screening. Development of image analysis",
"statistical analysis",
"and data visualisation tools for quantitative single cell analysis based on imaging data."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article delineates a novel strategy and accompanying analytical platform to assess and compare the complex, multiparametric effects of drug treatments on cell populations (at the single cell level). The platform, described in detail, couples cell phenotype (fluorescence) imaging and an integrated quantitative image analysis and statistical data analysis/visualisation capability that enable rapid exploration of cellular states/responses across multiple scales of the data hierarchy.\nFocused on data capturing the level of single cell phenotypes measured through imaging with subcellular resolution, this article captures an analytical strategy that is relevant to drug discovery, but also to fundamental analysis of cellular processes, their underlying mechanisms, and potentially, diagnostic assessment/identification of defective processes in disease states.\n\nUsing the NCI COMBO drug library as a basis for analysis, the authors detail the experimental, imaging, and image analysis process, but reserve particular focus for the methods of drug effect analysis, primarily relying on the Mahalanobis distance. This measure captures a compact representation of drug effect strengths relative to control variation, whilst also embedding information about the (relative weights) of specific cellular features that capture/comprise the effects of any given perturbation (drug). Mahalanobis distance thus emerges as a powerful tool to assess, compare, sort, and visualise drug effects, especially when compared (as the authors show) with single feature responses, which are far less systematic in their indications.\n\nThe authors also highlight the strengths of their approach for identifying complex temporal dynamics of drug response, i.e. how changes in different phenotypic features emerge at different times after treatment. Reflecting the way that even highly specific initial perturbation impacts are propagated outward over time within cellular information processing and responses networks, this is an important additional dimension that is precisely addressed using the authors' analytical strategy.\n\nImportantly, due to the very large scale of the datasets described, the authors indicate (reasonably so) that it is not feasible to share the entire dataset online. However, they indicate that, on request, components, or the entirety of the dataset can be accessed based on a request to the corresponding author. I interpret this as constituting data availability in practice.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "75113",
"date": "14 Dec 2020",
"name": "Robert Murphy",
"expertise": [
"Reviewer Expertise computational cell biology",
"biological image analysis and modeling"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript describes high throughput microscopy measurement and analysis of the effects of drugs using multiple fluorescent probes. While the study has been carefully done and the results for particular drugs may be of interest to researchers, the manuscript does not represent a significant advance either in experimental approach or in subsequent analysis. The use of multiple simultaneous fluorescent probes is widespread in biomedical research, including in drug development. For example, the work on cell painting described by Bray et al. is has been widely cited and used.1 The use of similar multivariate statistical methods for analysis of the signals from those probes is also routine.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly",
"responses": [
{
"c_id": "6215",
"date": "23 Dec 2020",
"name": "Zvi Kam",
"role": "Author Response",
"response": "We believe that the ability to detect drug effects on cells and relate them, in an unbiased fashion to quantitative attributes extracted from the images is a key feature that is missing when screens are directed towards identification of drugs or genes related to predetermined functional or morphological effects. The unbiased approach provided by Mahalanobis score and the identification of \"most contributing attributes\" to score differences from controls may reveal unexpected results in laborious screen preparations, independently of the initial aims of the work."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1199
|
https://f1000research.com/articles/10-37/v1
|
19 Jan 21
|
{
"type": "Research Article",
"title": "Analysis of ion currents in mass spectrometric profiles using glioblastoma tissue",
"authors": [
"Anatoly A. Sorokin",
"Evgeny S. Zhvansky",
"Denis S. Zavorotnyuk",
"Vsevolod A. Shurkhay",
"Denis S. Bormotov",
"Alexander A. Potapov",
"Anatoly A. Sorokin",
"Denis S. Zavorotnyuk",
"Vsevolod A. Shurkhay",
"Denis S. Bormotov",
"Alexander A. Potapov"
],
"abstract": "Background: The development of direct ambient ionization methods makes way for fast mass-spectrometry profiling of biological samples, which has great potential in medicine. Those methods, unlike traditional mass spectrometric analysis with chromatographic separation, are not able to take into account inter-ion interaction, ion suppression, and matrix effect due to the absence of chromatographic separation of the mixture components. So dynamics of ion current during direct ambient ionization mass-spectra is governed by the component micro-extraction and electrospray ionization influenced by the geometry of the sample, its position, and internal heterogeneity. Despite the progress in mass-spectrometry of biological samples, not much is known about the influence of sample type and structure on its molecular profile peculiarities. Methods: In this work, we propose to use analysis of the correlation between individual ion currents for a better understanding of ion current variability sources and grouping ions of high biological importance. Several fragments of glioblastoma tissue from a single patient are used for these purposes. Results: Ion currents have different dynamics considering different ions in different fragments. The correlation of two selected ion currents could be positive or negative for single fragment measurement. Correlations have persistent or alternating signs in different fragments for two selected ions. The spread of correlations of each pair of ion currents is calculated for evaluation of the signs’ stability. Conclusions: We were able to group ions according to the primary reason for their variabilities such as micro-extraction, mass-spectrometry measurement, or specimens' properties. Such grouping would allow the development of more reliable and reproducible methods of mass-spectrometry data analysis and improve the accuracy of results of its application in medicine.",
"keywords": [
"mass spectra",
"ion current",
"microextraction processes",
"ambient ionization"
],
"content": "Introduction\n\nNew techniques for rapid tissue profiling in mass spectrometry have made it possible to develop methods for express analysis of biological substances1–7. Many approaches to integrate mass spectrometry into the clinical routine have been published8–12. Interpretation and analysis of mass spectra of complex mixtures of biological molecules are still a rather difficult task due to the enormous variety of molecule types contained in biological tissues and their concentrations13,14. The LC-MS/MS analysis, which is considered a gold standard between methods for the accurate identification of complex mixture components, is time-consuming and cannot be used as a method for rapid analysis15. Therefore, for rapid analysis, one has to use direct mass spectrometry without preliminary chromatographic separation of mixtures. Neurosurgery is one of the areas that requires rapid analysis of biological tissues dissected during procedure16 to determine the boundaries of the tumor and control the volume of the excised tissue. It was shown that this problem could be solved using the mass spectrometric approach17.\n\nPrimary brain tumors account for 1.4% of all cancers and 2.4% of all cancer-related mortality in the United States. Approximately 20,500 newly diagnosed cases and 12,500 deaths are associated with primary malignant brain tumors each year18. According to recent reports, glial tumors are considered the most common primary brain tumors. They represent 31% of all brain and central nervous system (CNS) tumors diagnosed in the United States, and 81% of all malignant brain and CNS tumors19,20. Glial tumors can be divided into different subtypes based on morphological and genetic data following the WHO classification21.\n\nIt should be noted that all glial tumors can be very heterogeneous at the morphological level not only in between patients but even in within-patient due to the relatively random mixture of more benign and more aggressive parts22. This variability can become a serious problem for the rapid, accurate, and objective diagnosis of both the tumor border and its level of malignancy. The task of detecting tumor cells is most important in the border area since the undissected tumor cell may contribute to recurrence in the future23.\n\nDuring the transfer of the results of biological experiments into clinical practice a number of questions needs to be addressed23, including a study of the stability of the mass spectrometric characteristics of the sample and the reproducibility of the results under variable external (mode, range, temperature, resolution, time) or internal (sample characteristics, biological variability) conditions. For this analysis, the stability of the whole spectrum structure and the analysis of the dynamics of individual ion currents are of interest.\n\nThe study of the dynamics of individual ion currents is widely used in the analysis of mass spectrometric images24–27. In this work, using the analysis of the dynamics of ion currents, we investigated the influence of the internal heterogeneity of the sample, as well as the processes of ion suppression and extraction on the interpretation of mass spectrometric profiles of glioblastoma samples.\n\n\nMethods\n\nThe study used a Thermo LTQ Orbitrap XL high-resolution mass spectrometer, which allows measurements in both high- and low-resolution spectra. For mass spectrometric profiling of the samples, extraction using an integrated cartridge (ICE)28 was used, followed by electrospray ionization. Mass spectra were measured according to a previously developed protocol29. In this work, we used only low-resolution spectra in the range of 100–2000 m/z in both polarities.\n\nA tissue sample dissected at the border of the tumor was provided by the N.N. Burdenko neurosurgery center and analyzed following the protocol approved by the ethical committee of the N.N. Burdenko neurosurgery center (order 40 from 12.04.2016, revised with order 131 from 17.07.2018). Signed informed consent was obtained from the patients before surgery, in which it was specifically noted that all removed tissues could be used for further research. The study was conducted in accordance with the 2013 edition of the Declaration of Helsinki. All procedures were carried out according to the relevant guidelines and regulations approved by the N.N. Burdenko neurosurgery center. Written informed consent for publication of the patients’ details was obtained from the patients.\n\nThe tissue sample was split into two parts. The first was characterized by hematoxylin and eosin staining and further immunohistochemical analysis to determine the diagnosis and grade of malignancy. As a result of the analyses carried out, the diagnosis \"glioblastoma, IDH-1 wild type, without MGMT mutation\" was obtained. The second part was frozen and kept at -80°C until mass spectrometry measurements were taken. Glioblastoma tissue was divided into six fragments, and a mass spectrometric profile was measured for each of them. Measurements were carried out continuously by alternating modes of polarity and range and repeated twice for 10 minutes. Thus, the interval between repetitions of the measurement in each condition was approximately five minutes.\n\nMass spectra were processed in a similar way to previously described23–30. So, mass spectra were converted to vectors with 1m/z binning width and were normalized to total ion current. Previously we calculated similarity spectra matrices (SSM)29,30, where each matrix element was calculated as a cosine measure between two mass spectra (vectors), and these mass spectra (scans) were taken from a whole measurement. Thus, SSM demonstrated the similarity between scans of single measurement and/or different measurements. In this paper, we calculate similarity measure matrices, where similarity measure is Pearson’s R correlation between ion currents of different m/z, and ion currents are continuously sampled from scan to scan and from measurement to measurement.\n\nAlso, we calculated the spread of correlations for different tissue samples and each m/z. The matrices of correlation spread were created in the following way: the correlation matrix of ion currents was calculated for each fragment of tissues; each value of matrices (R) was rounded to -1, 0, or +1 by the binning with bin boundaries -0.2 and 0.2 respectively; the difference between maximum and minimum R in the set of six fragments formed the values of matrices of correlation spread.\n\nAll calculations and visualizations were made using code written by the authors using MATLAB (available in the Underlying data31).\n\n\nResults\n\nCorrelation matrices of ion currents are shown in Figure 1 and Figure 2 for negative and positive ion modes, respectively. The top and the right panels of the figure show the average spectrum for all measurements, the left panel represents the similarity matrices. So, the right and the top panels are the aggregated mass spectrum of the spectra of all 6 measurements normalized to the total ion current. The m/z range on the axis of the spectra corresponds to the range on the axes of the similarity matrix, i.e. the values of the axes of the similarity matrix and the spectra coincide.\n\nSeveral regions of the high ion correlation represented by compact red squares could be seen in the negative ion current correlation matrix (Figure 1). Within these regions, ion currents have a correlation close to unity, which indicates the similarity in dynamics regardless of the relative intensity of the ion current in samples. Correlation values close to the negative unity indicate a negative linear dependence of the intensities of a pair of ions in all samples.\n\nInterestingly, there are two groups of ions in the approximate ranges of 700–900m/z and 1500–1700m/z. Ions in these groups are correlated both within and among themselves. The blue stripes at the boundaries of these red squares indicate a negative correlation of these groups with ions in the ranges of 100–700m/z and 900–1500m/z.\n\nThe correlation matrix for positive ions (Figure 2) contains significantly fewer areas of high correlation on the main diagonal of the similarity matrix. This means that ions close to each other in mass in the positive ion mode behave more independently than negative ions. Nevertheless, several correlated squares on the main diagonal with a correlation between 0.2 and 0.4 can be recognized in Figure 2. The highest correlated ions group in all samples are located around 1500–1600m/z. Contrary to the small number of high positive correlation values, the high negative correlation values are more common in the matrix of similarity of ion currents in positive polarity spectra.\n\nCorrelation matrices in Figure 1 and Figure 2 were built using all 6 fragments of the biological sample. In this case, the orientation of the fragment in the cartridge during measurement, as well as the internal heterogeneity of the sample, can affect the value of the correlation. Figure 3 and Figure 4 demonstrate the dynamics of the currents of several pairs of negative and positive ions, respectively. Figure 3 and Figure 4 represent the mutual distribution of ion currents. Figure 3 and Figures 4 A,C,E,G show a time course for pairs of ion currents grouped by fragments, where dashed lines separate repetitions of measurements for each fragment, whereas different glioblastoma fragments are separated by vertical lines. Ion currents are shifted to zero mean and normalized to the standard deviation (z-normalization). Figure 3 and Figures 4 B,D,F,H are scatter-plots for the corresponding pairs of ion currents. The coordinate of each point is determined by the value of the ion intensities normalized to the total ion current in a single mass spectrum.\n\nData correlations are above the graphs.\n\nData correlations are above the graphs.\n\nBoth pairs of ions with a high positive correlation (“correlating”, Figures 3A, 3C, Figures 4A, 4C) and with a noticeable negative correlation (“anti-correlating”, Figures 3E, 3G, Figures 4E, and 4G) were selected. For clarity, the measurements for individual fragments are separated by a solid line, and repetitions within one fragment are separated by a dotted line. Figure 3A and Figure 4A demonstrate the dynamics of pairs of ion currents whose correlations are positive in all fragments. Figure 3B and Figure 4B prove this statement since the cloud of points is strongly elongated along the y=k*x line. Ion pairs that are shown in Figure 3C and Figure 4C also have a positive total correlation. However, the correlation of ion currents in the fifth and sixth fragments in Figure 3C and the second and fifth fragments in Figure 4C is negative.\n\nUnlike the previous two pairs, the ion currents of the pair shown in Figure 3E and Figure 4E have a negative total correlation, which can be seen from the descending shape of the point clouds in Figure 3F and Figure 4F. In this case, the correlation of currents in the first and fourth fragments in Figure 3E and the third fragment in Figure 4E is positive.\n\nThe spread matrices shown in Figures 5B and 5D were constructed for visual assessment of the presence of ions with alternating correlations in various fragments in addition to the correlation matrices (Figures 5A and 5C the same as Figure 1 and Figure 2). Most of the anticorrelating regions for negative ions (blue stripes in Figure 5A) are alternating signs (corresponding red stripes in Figure 5B). Some of the low-mass ions have spread close to 1, despite the relatively high overall correlation. This means that these ions showed a correlation close to zero for some of the fragments. The similarity matrices of ion currents for individual fragments, which were used to calculate the spread of correlations, are presented in Figures S3 and S4.\n\n\nDiscussion\n\nThe ions in the matrices (Figure 1 and Figure 2) are ordered by the m/z ratio (along the axis of matrices). The understanding of conformity of ion currents with the obtained mass spectrometric profiles is convenient in Figure 1 and Figure 2. However, ion currents can be reordered, for example, using hierarchical clustering methods so that the most similar ions are found in the closest rows (and columns) (Figures S1 and S2). This method of analysis allows for exploring different groups of ions. For example, groups of isotopes and adducts can be combined into one feature, hence, the dimension of the mass spectrum feature space can be reduced. Such reduction is of crucial importance for the application of machine learning techniques since the performance of many algorithms is degraded with the growth of the number of features and the presence of highly correlated features. It also should be taken into account that both adducts and isotopes represent the same pool of molecules, so redistribution of the concentration into a number of features obscured the overall dynamics of the ion current and makes interpretation of the spectra more difficult. The most common distance between the closest ions is equal to one for positive ions and one or two for negative ions (Figures S1C and S2C). The presence of chlorine adducts in negative mode causes this effect, as chlorine adducts give a significant isotopic peak at a distance of 2m/z from the monoisotopic peak. Such peaks with similar ion currents dynamics and 1 or 2 m/z difference form about 10% of all peaks in both polarities.\n\nSome of the pairs of ion currents have persistent signs of correlation, i.e. the correlation has the same sign for all measured fragments (Figure 3A, Figures 4A,E). Adducts and isotopic peaks belong to such pairs. At the same time, ions that are neither isotopes nor adducts, but have a high positive persistent correlation of ion currents (Figure 3A, Figure 4A), can be a result of coordinated metabolic processes, for example, intermediates of one metabolic pathway. The identification and detailed analysis of such groups of ions in different fragments of the sample can provide additional information for the characterization of the sample itself. Thus, groups of ions with a high positive correlation of ion currents, whose sign does not change from fragment to fragment, are functionally coupled: either as adducts/isotopes or as biochemically related molecules.\n\nStrictly negative persistent values of the correlation of ion currents (Figure 4E) indicate the ion suppression, the matrix effect. Such groups of ions could be associated with the measuring characteristics of the mass spectrometer.\n\nIon pairs with alternating signs of correlation values are associated with sample heterogeneity and extraction specifics of corresponding substances. So, two substances in one fragment were extracted in the same way from similar locations, whereas the regions of high concentration of these substances could be located differently in the other fragment, for example, on the surface and deeper inside the fragment. In this case, in the second fragment, the first substance might have been exhausted by the time the second substance could just start to be actively extracted. Examples of such pairs can be seen in Figures 3E, 3G in the second fragment, when the ion intensities change take place at the repetition border. It should be noticed that due to the peculiarities of the measurement protocol, the time between two repeated measurements of one sample is equal to five minutes, which is enough for a significant change in the ion concentration during the extraction process. The study of the internal heterogeneity of the sample would be very interesting since the sample was located at the border of the tumor and thus contained both tumor tissues and reactively modified brain tissues of the transitional region.\n\nIon currents of a pair of molecules can be either correlated or anticorrelated within one fragment, whereas the trend can be stable or alternating in different fragments. These factors allow the statement of the reason for the dynamics of ion currents: ion suppression, the sample position, or the functional relationship of molecules.\n\nThe alternating-sign correlation is rare in the positive mode compared to the negative ion mode, if we compare the correlation of the corresponding ion currents in different fragments (Figure 5D). Thus, the effects associated with the specifics of extraction, heterogeneity of the sample, and the location of the sample in the ion source are quite rare for the positive mode. At the same time, the alternating correlation of ion currents between different fragments dominates in the negative mode, which indicates the manifestation of such effects. Consequently, the biopsy tissues contain large quantities of molecules that are detected in the positive mode, and these molecules are rather evenly distributed within the sample. The effects associated with the geometry of the sample and its internal heterogeneity are visible in the negative ion mode. This confirms earlier observations that positive ions better describe inter-patient variability, while negative ions reflect intra-sample variability29.\n\nThus, all pairs of ions between that have a significant correlation value can be divided into three groups:\n\n1. The persistent positive or negative relationship caused by the processes of measuring the mass spectrum (isotopes and adducts or ion suppression respectively);\n\n2. Positive biochemical relationship caused by the relationship of molecules concentrations in one tissue type, for example, due to a common precursor;\n\n3. Sign-alternating relationship caused by the position of different tissue types with different molecules concentrations in the sample and their orientation to the eluent flow.\n\nWe will refer to the three types of relationships described above as measure-, biochemical-, extraction-related coupling. All types of these couplings are present in mass spectra, but they can be distinguished. For example, if the difference m/z is equal to 1 or the mass of the adduct and the overall correlation is positive and has a constant sign, then this is a positive measure-related coupling; positive persistent correlation with an unobvious difference in mass — is a biochemical-related coupling candidate; negative persistent correlation with an unobvious difference in mass — negative measure-related coupling; alternating correlation in different fragments — extraction-related coupling.\n\nThus, we have three types of coupling, two signs for each coupling type. Four of these six cases can be identified with high accuracy by analysis of ion currents using mass spectrometric profiling, and the other two require additional identification of molecules and their mapping to metabolic networks. For example, the difference between a positive measure-related coupling due to the presence of a chlorine adduct and a biochemical-related coupling due to the presence of two lipid species, which fatty acid residues differ by one double bond, can be reliably revealed only by the identification of molecules through the LC-MS/MS analysis.\n\nThe study of biochemical-related coupling is of great interest. The ratio of saturated and unsaturated lipids could be changed due to metabolic reprogramming during the development of the tumor17. So, the differences in the population of fatty acids between healthy and tumor tissues become noticeable. This can lead to significant ion currents correlations due to common precursors — an altered set of fatty acids. For example, if stearate is present in excess, lipids containing C18 will be at sufficiently high concentrations and, hence, the correlation between the peaks will reflect an excess of C18 within a particular tissue type, despite the actual class of lipids (phosphoserines, cardiolipins, ethanolamines, etc.). If we assume that the microextraction processes for similar molecules have similar dynamics, then their ion currents will have a high correlation. Thus, biochemical specifics appear as a positive relationship of ion currents. Moreover, these ion currents can be affected by the extraction and geometric variability, which leads to a low absolute value of the overall correlation, although the correlation of ion currents in different fragments stays positive.\n\nThe mass spectrometric profiles of glial tumors in the negative mode consist mainly of ions, whose dynamics reflect the specifics of microextraction and geometry of a particular sample. In the positive mode, mass spectra consist mainly of evenly distributed ions that are not significantly affected by the geometry of the sample, and processes of microextraction are constant in various fragments of tumor tissue. It is worth noting that ions in the positive and negative modes correspond to different molecules and even classes of molecules.\n\nMethods created in the laboratory should be proven to be stable for the transition from laboratory studies to routine clinical mass spectrometric analysis. Routine clinical practice requires a reduction in the analysis time, therefore, recently, ambient ionization MS methods have been widely developed. Ambient ionization, unlike traditional LC-MS/MS analysis, allow measurements in the shortest possible time (minutes for the analysis vs hours for chromatography). At the same time, markers of a particular tissue found using LC-MS/MS can disappear from the ambient MS data due to the ion suppression, which can be found through analysis of Figures 5A and 5B or Figures 5C and 5D. The indicator of ion suppression is the case, where both values of ion currents correlation and correlation spread matrices are closed to zero (twice blue pixels). Thus, an understanding of the processes of ion suppression makes it possible to predict the behavior of markers found by LC-MS/MS analysis under conditions of ambient ionization setup. Aggregation of ions that behavior correlated due to measurement procedure (isotopes, adducts, etc.) is another possibility to simplify the analysis of mass spectrometric profiles. Such groups of ions are well distinguished using hierarchical clustering (Figures S1 and S2): Figures S1C and S2C demonstrate high peaks at 1m/z and 2m/z, which indicates the isotopic structure and the presence of chlorine adducts in the negative ion mode or saturated bonds in lipid tails.\n\n\nConclusion\n\nThe currents of ions forming the mass spectrometric profiles obtained from glial tumors are grouped according to their similarity with each other due to several factors that include the specifics of measuring, extraction, and biochemical processes.\n\nThe dynamics of ions in the mass spectrometric profiles of glioblastoma samples obtained in the positive mode proves that the bulk of ions corresponds to molecules uniformly distributed in the sample volume in a relatively stable concentration. At the same time, the profiles obtained in the negative mode contain ions with high heterogeneity of the spatial distribution, which is confirmed by a large number of pairs with an alternating sign of correlation. This makes the negative range the preferred choice when looking for markers of glial tumors. On the other hand, this fact allows us to conclude that marker lipids are differently distributed within the sample.\n\n\nData availability\n\nZenodo: Data and code for analysis of ion currents in mass spectrometric profiles using glioblastoma tissue, http://doi.org/10.5281/zenodo.430907731.\n\nThis project contains the following underlying data:\n\nDataset of mass spectrometric profiles of glioblastoma tissue fragments.\n\nSoftware file for figures replication.\n\nZenodo: Figure S1. Sorted similarity matrix of ion currents for negative mode, https://doi.org/10.5281/zenodo.430895432.\n\nZenodo: Figure S2. Sorted similarity matrix of ion currents for positive mode, https://doi.org/10.5281/zenodo.430895633.\n\nZenodo: Figure S3. Similarity matrix of ion currents in separated fragments of tissues for negative mode, https://doi.org/10.5281/zenodo.430895834.\n\nZenodo: Figure S4. Similarity matrix of ion currents in separated fragments of tissues for positive mode, https://doi.org/10.5281/zenodo.430896235.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Acknowledgments\n\nThe research used the equipment of Shared Research Facilities of N.N. Semenov Federal Research Center for Chemical Physics of the Russian Academy of Sciences.\n\n\nReferences\n\nFerreira CR, Yannell KE, Jarmusc AK, et al.: Ambient Ionization Mass Spectrometry for Point-of-Care Diagnostics and Other Clinical Measurements. Clin Chem. 2016; 62(1): 99–110. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXu X, Li W, Li T, et al.: Direct infusion-three-dimensional mass spectrometry enables rapid chemome comparison among herbal medicines. Anal Chem. 2020; 92(11): 7646–7656. PubMed Abstract | Publisher Full Text\n\nGonzález-Domínguez R, Sayago A, Fernández-Recamales Á: Direct infusion mass spectrometry for metabolomic phenotyping of diseases. Bioanalysis. 2017; 9(1): 131–148 . PubMed Abstract | Publisher Full Text\n\nRathahao-Paris E, Alves S, Boussaid N, et al.: Evaluation and validation of an analytical approach for high-throughput metabolomic fingerprinting using direct introduction-high-resolution mass spectrometry: Applicability to classification of urine of scrapie-infected ewes. Eur J Mass Spectrom (Chichester). 2019; 25(2): 251–258. PubMed Abstract | Publisher Full Text\n\nFreund DM, Sammons KA, Makunga NP, et al.: Leaf Spray Mass Spectrometry: A Rapid Ambient Ionization Technique to Directly Assess Metabolites from Plant Tissues. J Vis Exp. 2018; (136): 57949. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTakáts Z, Wiseman JM, Gologan B, et al.: Mass spectrometry sampling under ambient conditions with desorption electrospray ionization. Science. 2004; 306(5695): 471–473. PubMed Abstract | Publisher Full Text\n\nChiang S, Zhang W, Ouyang Z: Paper spray ionization mass spectrometry: recent advances and clinical applications. Expert Rev Proteomics. 2018; 15(10): 781–789. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPirro V, Llor RS, Jarmusch AK, et al.: Analysis of human gliomas by swab touch spray-mass spectrometry: applications to intraoperative assessment of surgical margins and presence of oncometabolites. Analyst. 2017; 142(21): 4058–4066. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPirro V, Alfaro CM, Jarmusch AK, et al.: Intraoperative assessment of tumor margins during glioma resection by desorption electrospray ionization-mass spectrometry. Proc Natl Acad Sci U S A. 2017; 114(26): 6700–6705. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu J, Cooks RG, Ouyang Z: Biological tissue diagnostics using needle biopsy and spray ionization mass spectrometry. Anal Chem. 2011; 83(24): 9221–9225. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSt John ER, Balog J, McKenzie JS, et al.: Rapid evaporative ionisation mass spectrometry of electrosurgical vapours for the identification of breast pathology: towards an intelligent knife for breast cancer surgery. Breast Cancer Res. 2017; 19(1): 59. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPhelps DL, Balog J, Gildea LF, et al.: The surgical intelligent knife distinguishes normal, borderline and malignant gynaecological tissues using rapid evaporative ionisation mass spectrometry (REIMS). Br J Cancer. 2018; 118(10): 1349–1358. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJarmusch AK, Pirro V, Baird Z, et al.: Lipid and metabolite profiles of human brain tumors by desorption electrospray ionization-MS. Proc Natl Acad Sci U S A. 2016; 113(6): 1486–1491. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBalog J, Sasi-Szabó L, Kinross J, et al.: Intraoperative tissue identification using rapid evaporative ionization mass spectrometry. Sci Transl Med. 2013; 5(194): 194ra93. PubMed Abstract | Publisher Full Text\n\nMa X, Ouyang Z: Ambient ionization and miniature mass spectrometry system for chemical and biological analysis. Trends Analyt Chem. 2016; 85(A): 10–19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEberlin LS, Norton I, Orringer D, et al.: Ambient mass spectrometry for the intraoperative molecular diagnosis of human brain tumors. Proc Natl Acad Sci U S A. 2013; 110(5): 1611–1616. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSorokin A, Shurkhay V, Pekov S, et al.: Untangling the metabolic reprogramming in brain cancer: discovering key molecular players using mass spectrometry. Curr Top Med Chem. 2019; 19(17): 1521–1534. PubMed Abstract | Publisher Full Text\n\nGladson CL, Prayson RA, Liu WM: The pathobiology of glioma tumors. Annu Rev Pathol. 2010; 5: 33–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOstrom QT, Gittleman H, Stetson L, et al.: Epidemiology of gliomas. Cancer Treat Res. 2015; 163: 1–14. PubMed Abstract | Publisher Full Text\n\nOstrom QT, Gittleman H, Farah P, et al.: CBTRUS statistical report: Primary brain and central nervous system tumors diagnosed in the United States in 2006-2010. Neuro Oncol. 2013; 15(Suppl 2): ii1–ii56. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDiamandis P, Aldape K: World health organization 2016 classification of central nervous system tumors. Neurol Clin. 2018; 36(3): 439–447. PubMed Abstract | Publisher Full Text\n\nPedeutour-Braccini Z, Burel-Vandenbos F, Gozé C, et al.: Microfoci of malignant progression in diffuse low-grade gliomas: towards the creation of an intermediate grade in glioma classification? Virchows Arch. 2015; 466(4): 433–444. PubMed Abstract | Publisher Full Text\n\nZhvansky ES, Sorokin AA, Indeykina MI, et al.: Unified representation of high- and low-resolution spectra to facilitate application of mass spectrometric techniques in clinical practice. Clinical Mass Spectrometry. 2019; 12: 37–46. Publisher Full Text\n\nKhatib-Shahidi S, Andersson M, Herman JL, et al.: Direct molecular analysis of whole-body animal tissue sections by imaging MALDI mass spectrometry. Anal Chem. 2006; 78(18): 6448–6456. PubMed Abstract | Publisher Full Text\n\nKaddi CD, Parry RM, Wang MD: Multivariate hypergeometric similarity measure. IEEE/ACM Trans Comput Biol Bioinform. 2013; 10(6): 1505–1516. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWüllems K, Kölling J, Bednarz H, et al.: Detection and visualization of communities in mass spectrometry imaging data. BMC Bioinformatics. 2019; 20(1): 303. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlexandrov T, Kobarg JH: Efficient spatial segmentation of large imaging mass spectrometry datasets with spatially aware clustering. Bioinformatics. 2011; 27(13): i230–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPekov SI, Eliferov VA, Sorokin AA, et al.: Inline cartridge extraction for rapid brain tumor tissue identification by molecular profiling. Sci Rep. 2019; 9(1): 18960. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhvansky ES, Eliferov VA, Sorokin AA, et al.: Assessment of variation of inline cartridge extraction mass spectra. J Mass Spectrom. 2020; e4640. PubMed Abstract | Publisher Full Text\n\nZhvansky ES, Pekov SI, Sorokin AA, et al.: Metrics for evaluating the stability and reproducibility of mass spectra. Sci Rep. 2019; 9(1): 914. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhvansky E, Sorokin A, Zavorotnyuk D, et al.: Data and code for analysis of ion currents in mass spectrometric profiles using glioblastoma tissue [Data set]. Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4309077\n\nZhvansky E: Figure S1. Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4308954\n\nZhvansky E: Figure S2. Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4308956\n\nZhvansky E: Figure S3. Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4308958\n\nZhvansky E: Figure S4. Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4308962"
}
|
[
{
"id": "88228",
"date": "30 Jul 2021",
"name": "Vladimir Frankevich",
"expertise": [
"Reviewer Expertise mass spectrometry",
"clinical mass spectrometry",
"metabolomics",
"instrumentation"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study is aimed to determine the causes of the ion current dynamic. This question is very actual for ambient ionization mass spectrometry since there are numerous attempts to make predictions as classification or regression tasks by mass spectrometric profiles obtained with different ambient ion sources. This paper demonstrates an approach to differ the internal heterogeneity of the sample, extraction processes, and measuring process by the mutual dynamics of the ion currents. The research is based on the correlation calculation between ion currents. It was clearly demonstrated that some of the ions are “correlating” with each other, while others are “anti-correlating” or not sufficiently interconnected. The results are justified enough. The data is full enough to reproduce the results. The core of the method is simple but appears to be effective in the task of grouping ions by the nature of their dynamics. The paper has a high level of novelty and could be applied in the analysis of ambient mass spectrometry data.\nHowever, some minor revisions would improve the readability and quality of the paper:\nThe paper would benefit from English correction.\n\nThe resolution of the spectra should be noted.\n\nThe sentence describing the spread of correlations should be rewritten in a better manner.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "144390",
"date": "18 Jul 2022",
"name": "Joris Meurs",
"expertise": [
"Reviewer Expertise Analytical chemistry",
"mass spectrometry"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present to use of ion currents to assess MS profiles of glioblastoma tissue. I have some suggestions for further improvement of the article:\nPlease add more detail to the figure caption to make the figures easier to understand.\n\nI was wondering why the authors used the correlation of the average spectrum to assess and not the variability in ion intensity between the repeats.\n\nBrief details should be added on the mass spectrometry analysis (extraction solvent, mass resolution).\n\nIt was not clear what measure in the 'spread matrices'.\n\nThe authors could elaborate more on the use of the method for clinical practice.\n\nHow do the results reflect on the immunostaining?\n\nHow were the ions chosen for correlation (Figure 4)? Were these randomly chosen, are these fragments/clusters, or biologically-related?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-37
|
https://f1000research.com/articles/10-34/v1
|
19 Jan 21
|
{
"type": "Systematic Review",
"title": "Hemorrhagic and ischemic stroke in patients with coronavirus disease 2019: incidence, risk factors, and pathogenesis - a systematic review and meta-analysis",
"authors": [
"Syahrul Syahrul",
"Helnida Anggun Maliga",
"Muhammad Ilmawan",
"Marhami Fahriani",
"Sukamto S. Mamada",
"Jonny Karunia Fajar",
"Andri Frediansyah",
"Faza Nabila Syahrul",
"Imran Imran",
"Salim Haris",
"Aldy Safruddin Rambe",
"Talha Bin Emran",
"Ali A. Rabaan",
"Ruchi Tiwari",
"Kuldeep Dhama",
"Firzan Nainu",
"Endang Mutiawati",
"Harapan Harapan",
"Helnida Anggun Maliga",
"Muhammad Ilmawan",
"Marhami Fahriani",
"Sukamto S. Mamada",
"Jonny Karunia Fajar",
"Andri Frediansyah",
"Faza Nabila Syahrul",
"Imran Imran",
"Salim Haris",
"Aldy Safruddin Rambe",
"Talha Bin Emran",
"Ali A. Rabaan",
"Ruchi Tiwari",
"Kuldeep Dhama",
"Firzan Nainu",
"Endang Mutiawati",
"Harapan Harapan"
],
"abstract": "Background: In this study, we aimed to determine the global prevalence, chronological order of symptom appearance, and mortality rates with regard to hemorrhagic and ischemic stroke in patients with coronavirus disease 2019 (COVID-19) and to discuss possible pathogeneses of hemorrhagic and ischemic stroke in individuals with the disease. Methods: We searched the PubMed, Scopus, and Web of Science databases for relevant articles published up to November 8, 2020. Data regarding study characteristics, hemorrhagic stroke, ischemic stroke, and COVID-19 were retrieved in accordance with the PRISMA guidelines. The Newcastle-Ottawa scale was used to assess the quality of the eligible studies. The pooled prevalence and mortality rate of hemorrhagic and ischemic stroke were calculated. Results: The pooled estimate of prevalence of hemorrhagic stroke was 0.46% (95% CI 0.40%–0.53%; I2=89.81%) among 67,155 COVID-19 patients and that of ischemic stroke was 1.11% (95% CI 1.03%–1.22%; I2=94.07%) among 58,104 COVID-19 patients. Ischemic stroke was more predominant (incidence: 71.58%) than hemorrhagic stroke (incidence: 28.42%) in COVID-19 patients who experienced a stroke. In COVID-19 patients who experienced a stroke, hospital admission with respiratory symptoms was more commonly reported than that with neurological symptoms (20.83% for hemorrhagic stroke and 5.51% for ischemic stroke versus 6.94% for hemorrhagic stroke and 5.33% for ischemic stroke, respectively). The pooled mortality rate of COVID-19 patients who experienced a hemorrhagic and ischemic stroke was 44.72% (95% CI 36.73%–52.98%) and 36.23% (95% CI 30.63%–42.24%), respectively. Conclusions: Although the occurrence of hemorrhagic and ischemic stroke is low, the mortality rates of both stroke types in patients with COVID-19 are concerning, and therefore, despite several potential pathogeneses that have been proposed, studies aimed at definitively elucidating the mechanisms of hemorrhagic and ischemic stroke in individuals with COVID-19 are warranted. PROSPERO registration: CRD42020224470 (04/12/20)",
"keywords": [
"COVID-19",
"haemorrhagic stroke",
"ischemic stroke",
"meta-analysis",
"pathogenesis",
"SARS-CoV-2",
"systematic review"
],
"content": "Introduction\n\nCoronavirus disease 2019 (COVID-19), which is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection, has become a global human pandemic that is believed to have begun in late December 2019. The disease quickly spread to 217 countries, infected more than 82 million individuals, and has caused more than 1.8 million deaths as of December 30, 20201. Moreover, the second wave of this pandemic remains ongoing in various countries2. Numerous treatment strategies and drugs have been proposed3–5; however, a definitive therapy or treatment for COVID-19 has not yet been announced by the World Health Organization6. SARS-CoV-2 is a novel coronavirus, which is reported to have originated initially from an animal source7. The mortality rate of SARS-CoV-2 infection is the lowest among the infections caused by other members of the coronavirus family that have previously infected humans, including severe acute respiratory syndrome coronavirus (SARS-CoV) and Middle East respiratory syndrome coronavirus (MERS-CoV)8,9. However, SARS-CoV-2 has a higher reproduction rate (R0) and thus a higher transmission rate than SARS-CoV and MERS-CoV10.\n\nThe majority of individuals infected with SARS-CoV-2 are generally asymptomatic, although the most common symptoms of COVID-19 include dry cough, fever, dyspnea, chest pain, headache, and muscle ache11,12. The issue of hypercoagulability-related thrombotic vascular events in those infected with SARS-CoV-2 is emerging13,14. Evidence suggests that COVID-19 patients may experience increased rates of thromboembolism, as high as 15%–26%15. In addition, another concern is the increased risk of a hemorrhagic stroke among COVID-19 patients16–18. The World Stroke Organization has reported that COVID-19 increases the risk for an ischemic stroke by approximately 5% (95% confidence interval (CI) 2.8%–8.7%)19. Other possible explanations for the occurrence of ischemic stroke in COVID-19 patients include reduced angiotensin (ANG) (1-7) synthesis20, cardioembolism21,22, hyperviscosity23,24, and an induced hypercoagulative state25,26. Discussions surrounding the possible mechanism(s) underlying hemorrhagic stroke in COVID-19 patients have included the expression of angiotensin-converting enzyme 2 (ACE2), immunity, inflammation, endothelial dysfunction at the blood-brain-barrier (BBB), aging, stress, and anxiety27.\n\nThe aims of our present study were to determine the global incidence of ischemic and hemorrhagic stroke in patients with COVID-19; determine the mortality rate of ischemic and hemorrhagic stroke in individuals with COVID-19; assess the frequency of symptoms that lead to hospital admission among COVID-19 patients who have experienced an ischemic or a hemorrhagic stroke; assess the risk factors for ischemic and hemorrhagic stroke in COVID-19; assess the association between ischemic and hemorrhagic stroke and the severity of COVID-19; and assess the association between ischemic and hemorrhagic stroke and mortality in COVID-19. In addition, we also sought to propose possible pathogeneses of ischemic and hemorrhagic stroke in individuals with SARS-CoV-2 infection.\n\n\nMethods\n\nThis systematic review and meta-analysis identified the stroke proportion among COVID-19 confirmed cases. We followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) recommendation to search electronic databases (see completed checklist28)29. The protocol of this study was registered in PROSPERO (CRD42020224470) on 4th December 2020.\n\nThe inclusion criteria were articles written in English which identified stroke as a comorbidity among randomly sampled COVID-19 cases. All case reports, case series, editorials, reviews, commentaries, and studies in targeted specific groups, such as in children were excluded.\n\nThe systematic searches were conducted in three databases (PubMed, Scopus, and Web of Science) to identify the potential articles as of November 8th, 2020. The search criteria were as follows. Scopus (TITLE(\"SARS-CoV-2\" OR \"COVID-19\" OR \"Wuhan coronavirus\" OR \"Wuhan virus\" OR \"novel coronavirus\" OR \"nCoV\" OR \"severe acute respiratory syndrome coronavirus 2\" OR \"coronavirus disease 2019 virus\" OR \"2019-nCoV\" OR \"2019 novel coronavirus\" OR \"severe acute respiratory syndrome coronavirus 2\" OR \"coronavirus\" OR \"coronaviruses\" OR \"SARS 2\" OR \"2019-nCoV acute respiratory disease\" OR \"novel coronavirus pneumonia\" OR \"COVID\") AND ALL(\"Stroke \" OR \"cerebrovascular disorders\" OR \"brain ischemia\" OR \"brain haemorrhage\" OR \"cerebrovascular accident\" OR \"intracerebral haemorrhage\" OR \"subarachnoid haemorrhage\" OR \"transient ischemic attack\" OR \"brain attack\" OR \"cerebral embolism\" OR \"cerebral thrombosis\" OR \"cerebral haemorrhage\" OR \"cerebrovascular insult\" OR \"intraparenchymal haemorrhage\" OR \"intraventricular haemorrhage\" OR \"cerebral hypoperfusion\" OR \"brain infarct\" OR \"cerebral infarct\"). Web of Science (TITLE(\"SARS-CoV-2\" OR \"COVID-19\" OR \"Wuhan coronavirus\" OR \"Wuhan virus\" OR \"novel coronavirus\" OR \"nCoV\" OR \"severe acute respiratory syndrome coronavirus 2\" OR \"coronavirus disease 2019 virus\" OR \"2019-nCoV\" OR \"2019 novel coronavirus\" OR \"severe acute respiratory syndrome coronavirus 2\" OR \"coronavirus\" OR \"coronaviruses\" OR \"SARS 2\" OR \"2019-nCoV acute respiratory disease\" OR \"novel coronavirus pneumonia\" OR \"COVID\") AND ALL=(\"stroke\" OR \"cerebrovascular Disorders\" OR \"brain ischemia\" OR \"brain haemorrhage\" OR \"cerebrovascular accident\" OR \"intracerebral haemorrhage\" OR \"subarachnoid haemorrhage\" OR \"transient ischemic attack\" OR \"brain attack\" OR \"cerebral embolism\" OR \"cerebral thrombosis\" OR \"cerebral haemorrhage\" OR \"cerebrovascular insult\" OR \"intraparenchymal haemorrhage\" OR \"intraventricular haemorrhage\" OR \"cerebral hypoperfusion\" OR \"brain infarct\" OR \"cerebral infarct\"). PubMed (TITLE(\"SARS-CoV-2\" OR \"COVID-19\" OR \"Wuhan coronavirus\" OR \"Wuhan virus\" OR \"novel coronavirus\" OR \"nCoV\" OR \"severe acute respiratory syndrome coronavirus 2\" OR \"coronavirus disease 2019 virus\" OR \"2019-nCoV\" OR \"2019 novel coronavirus\" OR \"severe acute respiratory syndrome coronavirus 2\" OR \"coronavirus\" OR \"coronaviruses\" OR \"SARS 2\" OR \"2019-nCoV acute respiratory disease\" OR \"novel coronavirus pneumonia\" OR \"COVID\") AND (\"stroke \" OR \"cerebrovascular Disorders\" OR \"brain ischemia\" OR \"brain haemorrhage\" OR \"cerebrovascular accident\" OR \"intracerebral haemorrhage\" OR \"subarachnoid haemorrhage\" OR \"transient ischemic attack\" OR \"brain attack\" OR \"cerebral embolism\" OR \"cerebral thrombosis\" OR \"cerebral haemorrhage\" OR \"cerebrovascular insult\" OR \"intraparenchymal haemorrhage\" OR \"intraventricular haemorrhage\" OR \"cerebral hypoperfusion\" OR \"brain infarct\" OR \"cerebral infarct\").\n\nData from the articles and the supplementary materials were extracted. Reference lists from the eligible articles were retrieved for further relevant studies.\n\nEndNote X9 (Thompson Reuters, Philadelphia, PA, USA) was used to import all titles and abstracts of the identified articles, and duplicated records were removed. Potentially eligible articles were identified through screening of the titles and abstracts. The full texts of the resulting studies were then thoroughly reviewed by two authors (HAM and MI) and the eligibility of each study was decided. Any disagreements between the investigators were solved by consulting with another investigator (MF).\n\nCollected information included study characteristics (author, study site, study design), number of patients with ischemic stroke or haemorrhagic stroke in COVID-19 cases and their mortality cases, the chronological order of patient admission to hospital based on symptoms (COVID-19 first, ischemic or haemorrhagic stroke first) and the COVID-19 characteristics (number of patients, severity, and mortality).\n\nThis study received no external funding.\n\nThe primary outcomes were: (a) global incidence of ischemic and haemorrhagic stroke in COVID-19 patients; (b) mortality rate of ischemic and haemorrhagic stroke in COVID-19; (c) the frequency of symptoms related to hospital admission among COVID-19 patients with ischemic or haemorrhagic stroke; (d) risk factors of ischemic and haemorrhagic stroke in COVID-19; (e) association of ischemic and haemorrhagic stroke with COVID-19 severity; and (f) association of ischemic and haemorrhagic stroke with mortality of COVID-19. The possible pathogenesis mechanisms of ischemic and haemorrhagic stroke in SARS-CoV-2 infection were also explained in this review.\n\nThe global prevalence of ischemic stroke was calculated as the number of COVID-19 patients who experienced ischemia divided by the total number of COVID-19 patients with or without ischemic or hemorrhagic stroke, expressed as frequency (%) and 95% CI. The frequency of symptoms that led to hospital admission was calculated as the total number of COVID-19 patients presenting with either respiratory or neurological symptoms first, divided by the total number of ischemic stroke cases among COVID-19 patients, and expressed as percentage and 95% CI. The mortality rate was calculated as the number of deaths of COVID-19 patients who experienced ischemic stroke divided by the total number of COVID-19 patients who experienced ischemic stroke. The same calculations were performed for hemorrhagic stroke.\n\nThe Newcastle-Ottawa scale (NOS)30 was used to critically assess the quality of the studies included in the meta-analysis. The Q test was used to evaluate the heterogeneity and potential publication bias of the data gathered from the studies.\n\nThe association between ischemic and hemorrhagic stroke and the occurrence of COVID-19 was calculated and expressed as the cumulative odds ratio and 95% CI using the Z test; differences with p < 0.05 were considered to be statistically significant. Heterogeneity among studies was assessed using the Q test, and heterogeneous data were analyzed using a random effects model. Publication bias was assessed using Egger’s test and funnel plots (p < 0.05 was considered to indicate potential for publication bias). The data were analyzed using Review Manager version 5.3 (The Cochrane Collaboration)31.\n\n\nResults\n\nA total of 1915 articles were retrieved via a literature search, with 1416 citations remaining after the duplicates were removed. An additional 685 articles were excluded after screening the titles and abstracts, leaving 731 studies (Figure 1), the full texts of which were reviewed for eligibility, with an additional 713 excluded. Exclusions included reviews, irrelevant studies, case series, case reports, and studies with insufficient data. This process resulted in 18 studies being included in the final analysis.\n\nAll 18 studies were included in the meta-analysis to calculate the global prevalence of hemorrhagic stroke in COVID-19 patients and the frequency of symptoms leading to hospital admission32–49. Data from nine studies were included to calculate the mortality rate of hemorrhagic stroke in COVID-19 patients32,33,35–37,40,45,48,49, while the other nine studies did not report relevant data. The 18 studies and the prevalence of hemorrhagic stroke reported in each of them are summarized in Table 1.\n\nNOS = Newcastle-Ottawa scale score, COVID-19 = coronavirus disease 19\n\nOnly 16 studies were included in the meta-analysis to calculate the global prevalence of ischemic stroke in COVID-19 patients and the frequency of symptoms leading to hospital admission34–49. Data from six studies were included to calculate the mortality rate of ischemic stroke in COVID-19 patients35–37,40,48,49, as no relevant data was reported in the remaining 12 studies. The prevalences of ischemic stroke reported in each of these studies are summarized in Table 2.\n\nNOS = Newcastle-Ottawa scale score, COVID-19 = coronavirus disease 19\n\nHemorrhagic stroke was reported in 18 studies including 67,155 COVID-19 patients, with a pooled estimate of prevalence of 0.46% (95% CI 0.40%–0.53%), I2=89.81% (Figure 2). Ischemic stroke was identified in 544 of 58,104 COVID-19 patients in 16 studies, which corresponded to a pooled prevalence estimate of 1.11% (95% CI 1.03%–1.22%), I2=94.07% (Figure 3). The incidence of hemorrhagic and ischemic stroke in COVID-19 patients was 28.42% (216/760) and 71.58% (544/760), respectively.\n\nThe pooled estimate of haemorrhagic stroke prevalence is 0.46% with 95% CI 0.40%–0.53%, p<0.0001; p-value for Egger and heterogeneity is 0.882 and <0.0001, respectively with I2 89.81%.\n\nThe pooled estimate of ischemic stroke prevalence is 1.11% with 95% CI 1.03%–1.22%, p<0.0001; p-value for Egger and heterogeneity is 0.730 and <0.0001, respectively with I2 94.07%.\n\nAmong COVID-19 patients who experienced hemorrhagic stroke, 45 of 216 patients (20.83%; 95% CI 15.41%–26.24%) complained of respiratory symptoms before neurological symptoms, while neurological symptoms preceded respiratory symptoms in 15 of 216 patients (6.94%; 95% CI 3.55%–10.33%). No clear onset of either neurological or respiratory symptoms in COVID-19 patients was reported in 11 studies (156/216, 72.22%) (Table 1).\n\nAmong COVID-19 patients who experienced ischemic stroke, 30 of 544 patients (5.51%; 95% CI 3.59%–7.43%) experienced respiratory symptoms before neurological symptoms, while 29 of 544 patients (5.33%; 95% CI 3.44%–7.21%) complained of neurological symptoms before respiratory symptoms. No clear onset of either neurological or respiratory symptoms in COVID-19 patients was reported in 11 studies (485/544, 89.15%) (Table 2).\n\nThe mortality rate of COVID-19 patients who experienced hemorrhagic stroke was 44.72% (95% CI 36.73%–52.98%), and the rate was slightly lower in those who experienced ischemic stroke (36.23%; 95% CI 30.63%–42.24%) (Figure 4).\n\n(A) The pooled mortality rate of haemorrhagic stroke is 44.72% (95% CI 36.73%–52.98%, p=0.2099); heterogeneity p=0.7343, I2=0%, and Egger’s p=<0.0001. (B) The pooled mortality rate of ischemic stroke is 36.23% (95% CI 30.63%–42.24%, p<0.0001); heterogeneity p=0.438, I2 0%, and Egger’s p<0.0001.\n\nThere was a lack of data regarding the association between hemorrhagic and ischemic stroke and the severity of the disease and mortality rate of patients with COVID-19. One study reported that 63% of COVID-19 patients who experienced ischemic stroke required admission to the intensive care unit (ICU)50.\n\n\nDiscussion\n\nThe cumulative incidence of hemorrhagic stroke in COVID-19 patients in the present study was 0.3% (216 of 67,155). This result is lower than a previously reported incidence (0.7%; 95% CI 0.50%–0.9%)51. The incidence of hemorrhagic stroke among all stroke types in COVID-19 patients fluctuated, with 12.2% in May 202052, 9.7% in June 202053, 17.2% in July 202054, 11.6% in September 202055, and 28.42% in the present systematic review.\n\nAmong 216 COVID-19 patients who experienced hemorrhagic stroke, 20.83% were admitted to a hospital owing to respiratory symptoms and developed a brain hemorrhage during hospitalization, while 6.94% (15/216) were admitted owing to neurological symptoms. The disease course is important for predicting the severity of the systemic disease as COVID-19 in the former group was more severe with abnormal vital sign(s), elevated inflammatory and coagulopathy markers, altered mental status, and the patients likely required mechanical ventilation and ICU care43. Diffuse microhemorrhages have been observed previously in COVID-19 patients, via brain imaging, and such microhemorrhages are scattered mostly in the juxtacortical white matter, corpus callosum, and brain stem56–61. The fatality rate of COVID-19 patients who experienced hemorrhagic stroke was 44.72% in our study, which is slightly lower than the rate reported previously (48.6%)51.\n\nThe pooled prevalence of ischemic stroke among COVID-19 patients in our systematic review was 0.94% (544 of 58,104). This result is much lower than that described in another report in which acute ischemic stroke was observed in 1.6% of the patients who presented with or were hospitalized owing to COVID-1962. Another study also reported increased incidence of ischemic stroke among COVID-19 patients compared to that in the non-COVID-19 group (81.4% vs. 74.6%)63. The incidence of ischemic stroke in our investigation was significantly lower compared with that in another study (71.58% vs. 87.4%, respectively)55.\n\nCOVID-19 patients who were admitted with respiratory symptoms and later experienced acute brain infarct comprised 5.51% (30 of 544) of the sample. Meanwhile, 5.33% were admitted owing to neurological symptoms related to ischemic stroke and had a positive result on real-time polymerase chain reaction (RT-PCR) testing for SARS-CoV-2 after screening; there was no clear explanation for hospital admission in the remainder (89.15%). This result is lower than that reported in a previous study, in which ischemic stroke was the reason for hospital admission in 26% of the COVID-19 patients62.\n\nThe cause of ischemic stroke is multifactorial in those with SARS-CoV-2 infection, and it may be due to systemic embolization and diffuse microvascular thrombosis (attributed to a significant increase in prothrombotic factors)53. In our systematic review, the mortality rate of COVID-19 patients who experienced ischemic stroke was 36.23%. This rate is higher than that in two previous studies, which reported inpatient mortality rates of 32% and 22.8%50. among COVID-19 patients who experienced ischemic stroke.\n\nHemorrhagic stroke is caused by the rupture of cerebral vessels, leading to the extravasation of blood components into the surrounding brain tissue. However, the molecular mechanism by which SARS-CoV-2 infection causes hemorrhagic stroke remains unclear. The ACE2 receptor occupied by the virus appears to be the primary culprit, which then induces subsequent damage to host cells64. Dysfunction of the ACE2 receptor is linked to the elevation of Ang II levels. Ang II is produced from Ang I, and this reaction is catalyzed by the action of ACE. Ang II-related effects are generated after being bound to the AT1 receptor. To counteract the dangerous effects caused by the excessive level of ACE/Ang II/AT1R axis, the ACE2/Ang (1-7)/Mas axis is activated65.\n\nSARS-CoV-2 infection increases Ang II levels. SARS-CoV-2 uses the ACE2 receptor as a portal to enter host cells66. Along with a protease, i.e., TMPRSS2, this receptor assists the virus in infecting cells66. Viral occupation of the ACE2 receptor affects the normal physiological function of the receptor, which is to degrade Ang II, resulting in the accumulation of Ang II in the blood. Elevated Ang II levels are associated with damage linked to the occurrence of hemorrhagic stroke67.\n\nA previous study proposed four modes of action used by Ang II to exert its effects (i.e., direct impact on the vascular system) causing vasoconstriction, promotion of platelet aggregation, increased free radical production, and a reduction in insulin sensitivity68. These actions of Ang II are associated with the occurrence of hemorrhagic stroke. Vasoconstriction is a vital physiological alteration that occurs in hypertension, which is recognized as one of the major risk factors for hemorrhagic stroke67. Ang II is also related to the activation of thrombogenic factors. This may explain the elevation of D-dimer levels, which are monitored in COVID-19 patients, particularly in those with severe infection69,70. Consequently, the activation of a procoagulant state may induce a hemorrhagic stroke71. Ang II is also known to inhibit the PI3K/AKT signaling pathway, which regulates the secretion of insulin, leading to lowered insulin sensitivity68,72, which is another risk factor for hemorrhagic stroke. A study using a rat model confirmed that diabetes could degrade tight junction (TJ) proteins mediated by the action of matrix metalloproteinases (MMPs)73. The correlation between junctional disruption, MMPs, and hemorrhagic stroke is described in the next section.\n\nSARS-CoV-2 infection causes a cytokine storm that induces degradation of the extracellular matrix. When SARS-CoV-2 infects the body, the immune system produces massive amounts of pro-inflammatory cytokines in response. An excessive amount of pro-inflammatory cytokines, including tumor necrosis factor-alpha (TNF-α), interleukin (IL)-1β, and IL-6, has been reported in most COVID-19 patients74. This phenomenon—known as the “cytokine storm”—results in the failure of multiple organs and contributes to COVID-19-related death. Ang II is strongly linked to the activation of nicotinamide adenine dinucleotide phosphate (NADPH) oxidase, which is responsible for the development of oxidative stress, a condition that has been known to be correlated with the excessive production of pro-inflammatory cytokines67,75. An in vitro study using a human BBB model revealed that the spike protein of SARS-CoV-2 could elevate the levels of IL-1β and IL-676. Several mechanisms have been proposed to explain the role of these cytokines in weakening vessel walls and the subsequent increase in the risk of hemorrhagic stroke, including its effect in the degradation of the extracellular matrix (ECM), which is the primary structure responsible for maintaining the integrity of vascular endothelial cells.\n\nDegradation of the ECM caused by MMPs increases BBB permeability, promotes extravasation of blood components, and contributes to hemorrhagic brain injury77. Many studies have reported that TNF-α can induce the production of MMPs, which are proteolytic enzymes that degrade the ECM. For example, a study reported that TNF-α administered intravenously to mice produced an elevation in the MMP-9 levels, followed by a significant increase in BBB permeability78. Another study demonstrated that MMP-3 expression was upregulated in porcine choroid plexus epithelial cells, which was followed by a reduction in transepithelial electrical resistance, indicating decreased cellular tightness79. Another pro-inflammatory cytokine, IL-1β, is also involved in the induction of MMPs, resulting in the destruction of the ECM. The expression and activity of MMP-2 in cardiac microvascular endothelial cells were induced by IL-1β80. After experimenting with chondrocytes, a study has also revealed that IL-1β exposure leads to MMP-1 upregulation81. This action is suggested to involve various signaling pathways (i.e., ERK1/2, JNKs) and protein kinase C (PKC)80–82. Studies have also reported the upregulation of MMP expression and activity in various models after exposure to IL-6. This cytokine increases MMP-9 activity during aortic aneurysms and ruptures in mice83. MMP-2 and MMP-9 levels have also been found to be increased in COVID-19 patients, which is consistent with the elevation of IL-6 expression in patients with lymphoma84. A STAT3 signaling pathway has been proposed as the pathway used by IL-6 to upregulate MMPs83,85.\n\nMoreover, the impairment of ECM caused by those pro-inflammatory cytokines may be significantly associated with the action of reactive oxygen species (ROS), such as superoxide and singlet oxygen, and reactive nitrogen species, such as nitrogen oxide and peroxynitrite86,87. A study found that TNF-α and IL-6 administration in human brain microvascular endothelial cells (HBMVEC) induced increased levels of ROS75. Oxidative stress-related BBB disruption leading to the incidence of stroke is strongly related to the activation of MMPs88–91.\n\nInterestingly, MMPs could alter the regulation of junctional proteins. A study using a rat model demonstrated that TJ damage in cerebral vessels was mediated by MMP-2 and MMP-9, and that this action could be inhibited by the MMP inhibitor BB-110192. A study found that the degradation of occludin, a transmembrane protein of the TJ, in BBB model bEnd3 monolayer was mediated by MMP-293, while another study confirmed that MMP-9 mediated the destruction of TJ protein in a BBB model hCMEC/D394.\n\nSARS-CoV-2 infection induces a cytokine storm that causes disturbance in junctional protein formation. Elevation of cytokine levels caused by SARS-CoV-2 infection could weaken vessel walls and ultimately increase the incidence of hemorrhagic stroke by impairing cellular junctional proteins, which is also the primary structure responsible for maintaining vascular endothelial cell integrity. The integrity of vascular endothelial cells is, in large part, determined by the presence of junctional proteins. In general, three major junctions are located in the BBB: TJs, adherens junctions (AJs), and gap junctions95. Any disturbances occurring in any of these structures will ultimately lead to vascular endothelial dysfunction. A previous study proposed that serum levels of TJ proteins may be used to predict the incidence of a hemorrhagic event(s) following ischemic stroke96.\n\nDisruption of junctional proteins could be caused by pro-inflammatory cytokines. Using bEnd.3 endothelial cells as the BBB model, a previous study demonstrated that TNF-α and IL-6 exposure produced a significant increase in cellular permeability, which could be attributed to the decreased expression of ZO-1 and claudins97. These findings are supported by a study investigating primary cerebral microvessels isolated from sheep, which revealed that 100 ng/mL of IL-6 reduced the expression of occludin98. The expression of cadherin, occludin, and claudin-5 led to a dose-dependent decrease in a HBMVEC model after treatment with TNF-α and IL-675. Using human umbilical vein endothelial cells (HUVECs), another study reported that the expression of occludin and E-cadherin was downregulated following exposure to interferon (IFN)-γ99. TNF-α treatment to HUVECs caused a change in localization of claudin-5 and JAM-A, while this cytokine also reduced the expression of occludin100.\n\nDamage to junctional proteins could also be associated with the disruption of polarity complex proteins. Polarity proteins work by regulating many aspects of cellular differentiation and proliferation, including junctional protein formation and localization101,102. Although the understanding of polarity complexes is mainly supported by a wide range of experiments involving epithelial cells, the complexes also play pivotal roles in endothelial cells103,104. Thus, impairments to polarity complexes could subsequently result in damage to transmembrane and cytoplasmic junctional proteins. For example, PATJ knockdown Caco2 cells affect the localization of occludin and ZO-3 in TJ formation105. Another study reported that VE-cadherin, the major transmembrane protein of AJ, connected to Pals1 during the formation of vascular lumen indicating the specific role of this polarity protein in regulating junctional formation in endothelial cells106. Interestingly, SARS-CoV-2 has been suggested to interact with the Pals1 protein in host cells through its envelope (E) protein107. A study demonstrated that another betacoronavirus—SARS—also uses this mode of interaction with host cells108.\n\nThe action of cytokines in downregulating junctional proteins could be mediated by the activation of NADPH oxidase75. This enzyme is one of the main sources of ROS in the vascular system, along with mitochondrial enzymes and xanthin oxidase109. It should be noted that the activation of NADPH oxidase is also linked to endothelial dysfunction leading to COVID-19-related thrombotic events110. It has been proposed that SARS-CoV-2 induces a thrombosis event by stimulating various tissue factors that are dependent on the activation of NADPH oxidase following its attack on endothelial cells111.\n\nCollectively, hemorrhagic stroke in COVID-19 patients may be associated with the elevation of Ang II levels, which is an event subsequent to SARS-CoV-2 occupation of the ACE2 receptor. The cytokine storm is also responsible for the degradation of some important components of cerebral vessels, such as MMPs and TJ, triggering cerebral vascular rupture.\n\nSARS-CoV-2 infection could cause ischemic stroke through the induction of a hypercoagulative state, endothelial injury, cytokine storm, and/or cardiogenic embolism21. Dysfunction of endothelial cells (induced by SARS-CoV-2 infection) may increase thrombin formation and fibrinolysis112. Coagulopathy due to a thrombosis event has been observed in COVID-19 patients, with elevated D-dimer and fibrinogen, although with no significant prolonged prothrombin time and activated partial thromboplastin time25,26,113. Increased fibrinogen levels also contribute to hyperviscosity, which is consistently found in COVID-19 patients, in whom viscosity varies between 1.9 to 4.2 centipoise (normal range, 1.4–1.8 centipoise)23. Hyperviscosity is not only caused by increases in fibrinogen level, but also by the cytokine storm, which plays an important role in increasing viscosity levels in those with COVID-19 by inducing the excessive release of IL-6 and TNF-α114,115.\n\nSystemic inflammation, which activates the complement pathway, induces the excessive release of inflammatory cytokines, causing venous thromboembolism by platelets and also inducing a hypercoagulative state116–118. This hypercoagulability could lead to macro- and microthrombus formation, which ultimately leads to cerebrovascular incidents119,120.\n\nThe ACE2 receptor also plays an important role in the neurological manifestations of SARS-CoV-2 infection. ACE2 converts Ang II into ANG (1-7), which plays an essential role as a neuroprotector. Administration of ANG (1-7) in animal models resulted in a decrease in neurological deficits and infarct size in rats with ischemic stroke121,122. Therefore, in COVID-19, the SARS-CoV-2 spike protein binds to the ACE2 receptor, resulting in decreased ANG (1-7) synthesis123. Cardioembolic (19.21%) and atherothrombotic (7.39%) events have also been reported to contribute to the etiology of ischemic stroke in COVID-19 patients124.\n\n\nConclusion\n\nAlthough the prevalence of hemorrhagic and ischemic stroke is low in COVID-19 patients, this systematic review may increase awareness among clinicians regarding the potentially high mortality rate of individuals with this infection who experience a stroke, especially those with severe infection.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\nFigshare: PRISMA checklist for ‘Hemorrhagic and ischemic stroke in patients with coronavirus disease 2019: Incidence, risk factors, and pathogenesis - A systematic review and meta-analysis’, https://doi.org/10.6084/m9.figshare.1351350928.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Acknowledgement\n\nAuthors would like to thank HT Editorial Service in assisting during manuscript writing processes.\n\n\nEthics statement\n\nNot required.\n\n\nReferences\n\nWorldometers: COVID-19 coronavirus pandemic. [cited 2020 November 14]. Reference Source\n\nFan G, Yang Z, Lin Q, et al.: Decreased Case Fatality Rate of COVID-19 in the Second Wave: A study in 53 countries or regions. Transbound Emerg Dis. 2020. PubMed Abstract | Publisher Full Text\n\nFrediansyah A, Nainu F, Dhama K, et al.: Remdesivir and its antiviral activity against COVID-19: A systematic review. Clin Epidemiol Glob Health. 2021; 9: 123–127. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFrediansyah A, Tiwari R, Sharun K, et al.: Antivirals for COVID-19: a critical review. Clin Epidemiol Glob Health. 2021; 9: 90–98. Publisher Full Text\n\nMudatsir M, Yufika A, Nainu F, et al.: Antiviral Activity of Ivermectin Against SARS-CoV-2: An Old-Fashioned Dog with a New Trick—A Literature Review. Sci Pharm. 2020; 88(3): 36. Publisher Full Text\n\nWHO: Solidarity clinical trial for COVID-19 treatments.2020. Reference Source\n\nHarapan H, Itoh N, Yufika A, et al.: Coronavirus disease 2019 (COVID-19): A literature review. J Infect Public Health. 2020; 13(5): 667–673. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDhama K, Patel SK, Pathak M, et al.: An update on SARS-CoV-2/COVID-19 with particular reference to its clinical pathology, pathogenesis, immunopathology and mitigation strategies. Travel Med Infect Dis. 2020; 37: 101755. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDhama K, Khan S, Tiwari R, et al.: Coronavirus Disease 2019-COVID-19. Clin Microbiol Rev. 2020; 33(4): e00028–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPetersen E, Koopmans M, Go U, et al.: Comparing SARS-CoV-2 with SARS-CoV and influenza pandemics. Lancet Infect Dis. 2020; 20(9): e238–e244. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang Y, Wang Y, Chen Y, et al.: Unique epidemiological and clinical features of the emerging 2019 novel coronavirus pneumonia (COVID-19) implicate special control measures. J Med Virol. 2020; 92(6): 568–76. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRodriguez-Morales AJ, Cardona-Ospina JA, Gutiérrez-Ocampo E, et al.: Clinical, laboratory and imaging features of COVID-19: A systematic review and meta-analysis. Travel Med Infect Dis. 2020; 34: 101623. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRothstein A, Oldridge O, Schwennesen H, et al.: Acute cerebrovascular events in hospitalized COVID-19 patients. Stroke. 2020; 51(9): e219–e22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDorgalaleh A, Dabbagh A, Tabibian S, et al.: Patients with Congenital Bleeding Disorders Appear to be Less Severely Affected by SARS-CoV-2: Is Inherited Hypocoagulability Overcoming Acquired Hypercoagulability of Coronavirus Disease 2019 (COVID-19)? Thieme Medical Publishers. Semin Thromb Hemost. 2020; 46(7): 853–855. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMalas MB, Naazie IN, Elsayed N, et al.: Thromboembolism risk of COVID-19 is high and associated with a higher risk of mortality: A systematic review and meta-analysis. EClinicalMedicine. 2020; 29: 100639. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPoyiadji N, Shahin G, Noujaim D, et al.: COVID-19-associated Acute Hemorrhagic Necrotizing Encephalopathy: Imaging Features. Radiology. 2020; 296(2): E119–E120. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMuhammad S, Petridisa A, Cornelius JF, et al.: Letter to editor: Severe brain haemorrhage and concomitant COVID-19 Infection: A neurovascular complication of COVID-19. Brain Behav Immun. 2020; 87: 150–151. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAl Saiegh F, Ghosh R, Leibold A, et al.: Status of SARS-CoV-2 in cerebrospinal fluid of patients with COVID-19 and stroke. J Neurol Neurosurg Psychiatry. 2020; 91(8): 846–848. PubMed Abstract | Publisher Full Text\n\nQureshi AI, Abd-Allah F, Al-Senani F, et al.: Management of acute ischemic stroke in patients with COVID-19 infection: Report of an international panel. Int J Stroke. 2020; 15(5): 540–554. PubMed Abstract | Publisher Full Text\n\nSardu C, Gambardella J, Morelli M, et al.: Is COVID-19 an Endothelial Disease? Clinical and Basic Evidence. Preprints. 2020; 2020040204. Publisher Full Text\n\nMarkus HS, Brainin M: COVID-19 and stroke—A global World Stroke Organization perspective. Int J Stroke. 2020; 15(4): 361–4. PubMed Abstract | Publisher Full Text\n\nLodigiani C, Iapichino G, Carenzo L, et al.: Venous and arterial thromboembolic complications in COVID-19 patients admitted to an academic hospital in Milan, Italy. Thromb Res. 2020; 191: 9–14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaier CL, Truong AD, Auld SC, et al.: COVID-19-associated hyperviscosity: a link between inflammation and thrombophilia? Lancet. 2020; 395(10239): 1758–1759. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen G, Wu D, Guo W, et al.: Clinical and immunological features of severe and moderate coronavirus disease 2019. J Clin Invest. 2020; 130(5): 2620–2629. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTang N, Bai H, Chen X, et al.: Anticoagulant treatment is associated with decreased mortality in severe coronavirus disease 2019 patients with coagulopathy. J Thromb Haemost. 2020; 18(5): 1094–1099. PubMed Abstract | Publisher Full Text\n\nHelms J, Tacquard C, Severac F, et al.: High risk of thrombosis in patients with severe SARS-CoV-2 infection: a multicenter prospective cohort study. Intensive Care Med. 2020; 46(6): 1089–98. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJohansson K, Jansson JH, Johansson L, et al.: D-dimer is associated with first-ever intracerebral hemorrhage: A nested case-control study. Stroke. 2018; 49(9): 2034–2039. PubMed Abstract | Publisher Full Text\n\nHarapan H: Hemorrhagic and ischemic stroke in patients with coronavirus disease 2019: Incidence, risk factors, and pathogenesis - A systematic review and meta-analysis. figshare. Journal contribution. 2021. http://www.doi.org/10.6084/m9.figshare.13513509.v1\n\nMoher D, Liberati A, Tetzlaff J, et al.: Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med. 2009; 6(7): e1000097. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStang A: Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol. 2010; 25(9): 603–5. PubMed Abstract | Publisher Full Text\n\nCochrane T: Review Manager (RevMan) 5.3. Copenhagen: The Nordic Cochrane Centre. 2008; 373. Reference Source\n\nAltschul DJ, Unda SR, de La Garza Ramos R, et al.: Hemorrhagic presentations of COVID-19: Risk factors for mortality. Clin Neurol Neurosurg. 2020; 198: 106112. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDogra S, Jain R, Cao M, et al.: Hemorrhagic stroke and anticoagulation in COVID-19. J Stroke Cerebrovasc Dis. 2020; 29(8): 104984. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGarcía-Moncó JC, Cabrera-Muras A, Collía-Fernández A, et al.: Neurological reasons for consultation and hospitalization during the COVID-19 pandemic. Neurol Sci. 2020; 41(11): 3031–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHernández-Fernández F, Sandoval Valencia H, Barbella-Aponte RA, et al.: Cerebrovascular disease in patients with COVID-19: neuroimaging, histological and clinical description. Brain. 2020; 143(10): 3089–3103. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJain R, Young M, Dogra S, et al.: COVID-19 related neuroimaging findings: A signal of thromboembolic complications and a strong prognostic marker of poor patient outcome. J Neurol Sci. 2020; 414: 116923. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJohn S, Hussain SI, Piechowski-Jozwiak B, et al.: Clinical characteristics and admission patterns of stroke patients during the COVID 19 pandemic: A single center retrospective, observational study from the Abu Dhabi, United Arab Emirates. Clin Neurol Neurosurg. 2020; 199: 106227. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKaradaş Ö, Öztürk B, Sonkaya AR: A prospective clinical study of detailed neurological manifestations in patients with COVID-19. Neurol Sci. 2020; 41(8): 1991–1995. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKatz JM, Libman RB, Wang JJ, et al.: Cerebrovascular Complications of COVID-19. Stroke. 2020; 51(9): e227–e231. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi Y, Li M, Wang M, et al.: Acute cerebrovascular disease following COVID-19: a single center, retrospective, observational study. Stroke Vasc Neurol. 2020; 5(3): 279–284. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiotta EM, Batra A, Clark JR, et al.: Frequent neurologic manifestations and encephalopathy-associated morbidity in Covid-19 patients. Ann Clin Transl Neurol. 2020; 7(11): 2221–2230. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLuigetti M, Iorio R, Bentivoglio AR, et al.: Assessment of neurological manifestations in hospitalized patients with COVID-19. Eur J Neurol. 2020; 27(11): 2322–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPinna P, Grewal P, Hall JP, et al.: Neurological manifestations and COVID-19: Experiences from a tertiary care center at the Frontline. J Neurol Sci. 2020; 415: 116969. PubMed Abstract | Publisher Full Text\n\nRifino N, Censori B, Agazzi E, et al.: Neurologic manifestations in 1760 COVID-19 patients admitted to Papa Giovanni XXIII Hospital, Bergamo, Italy. J Neurol. 2020; 1–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSawlani V, Scotton S, Nader K, et al.: COVID-19-related intracranial imaging findings: a large single-centre experience. Clin Radiol. 2021; 76(2): 108–116. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShahjouei S, Naderi S, Li J, et al.: Risk of stroke in hospitalized SARS-CoV-2 infected patients: A multinational study. EBioMedicine. 2020; 59: 102939. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShekhar R, Sheikh AB, Suriya SS, et al.: Neurological Complications Among Native Americans with COVID-19: Our Experience at a Tertiary Care Academic Hospital in the U.S. J Stroke Cerebrovasc Dis. 2020; 29(12): 105260. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSiegler JE, Cardona P, Arenillas JF, et al.: Cerebrovascular events and outcomes in hospitalized patients with COVID-19: The SVIN COVID-19 Multinational Registry. Int J Stroke. 2020; 1747493020959216. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang LJ, Sun WW, Wang YJ, et al.: Clinical Course and Mortality of Stroke Patients With Coronavirus Disease 2019 in Wuhan, China. Stroke. 2020; 51(9): 2674–82. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCagnazzo F, Arquizan C, Derraz I, et al.: Neurological manifestations of patients infected with the SARS-CoV-2: a systematic review of the literature. J Neurol. 2020; 1–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCheruiyot I, Sehmi P, Ominde B, et al.: Intracranial hemorrhage in coronavirus disease 2019 (COVID-19) patients. Neurol Sci. 2020; 1–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBhatia R, Pedapati R, Komakula S, et al.: Stroke in Coronavirus Disease 2019: A Systematic Review. J Stroke. 2020; 22(3): 324–35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee KW, Yusof Khan AHK, Ching SM, et al.: Stroke and Novel Coronavirus Infection in Humans: A Systematic Review and Meta-Analysis. Front Neurol. 2020; 11(1196): 579070. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSiow I, Lee KS, Zhang JJY, et al.: Stroke as a Neurological Complication of COVID-19: A Systematic Review and Meta-Analysis of Incidence, Outcomes and Predictors. J Stroke Cerebrovasc Dis. 2020; 30(3): 105549. PubMed Abstract | Publisher Full Text\n\nNannoni S, de Groot R, Bell S, et al.: Stroke in COVID-19: A systematic review and meta-analysis. Int J Stroke. 2020; 1747493020972922. PubMed Abstract | Publisher Full Text\n\nVattoth S, Abdelhady M, Alsoub H, et al.: Critical illness-associated cerebral microbleeds in COVID-19. Neuroradiol J. 2020; 33(5): 374–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGupta NA, Lien C, Iv M: Critical illness-associated cerebral microbleeds in severe COVID-19 infection. Clin Imaging. 2020; 68: 239–41. PubMed Abstract | Publisher Full Text\n\nFitsiori A, Pugin D, Thieffry C, et al.: COVID-19 is Associated with an Unusual Pattern of Brain Microbleeds in Critically Ill Patients. J Neuroimaging. 2020; 30(5): 593–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCannac O, Martinez-Almoyna L, Hraiech S: Critical illness-associated cerebral microbleeds in COVID-19 acute respiratory distress syndrome. Neurology. 2020; 95(11): 498–499. PubMed Abstract | Publisher Full Text\n\nDixon L, McNamara C, Gaur P, et al.: Cerebral microhaemorrhage in COVID-19: a critical illness related phenomenon? Stroke Vasc Neurol. 2020; 5(4): 315–322. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFitsiori A, Pugin D, Thieffry C, et al.: COVID-19 is Associated with an Unusual Pattern of Brain Microbleeds in Critically Ill Patients. J Neuroimaging. 2020; 30(5): 593–597. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMerkler AE, Parikh NS, Mir S, et al.: Risk of Ischemic Stroke in Patients With Coronavirus Disease 2019 (COVID-19) vs Patients With Influenza. JAMA Neurol. 2020; 77(11): 1366–72. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBenussi A, Pilotto A, Premi E, et al.: Clinical characteristics and outcomes of inpatients with neurologic disease and COVID-19 in Brescia, Lombardy, Italy. Neurology. 2020; 95(7): e910–e920. PubMed Abstract | Publisher Full Text\n\nSpence JD, de Freitas GR, Pettigrew LC, et al.: Mechanisms of Stroke in COVID-19. Cerebrovasc Dis. 2020; 49(4): 451–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRegenhardt RW, Bennion DM, Sumners C: Cerebroprotective action of angiotensin peptides in stroke. Clin Sci (Lond). 2014; 126(3): 195–205. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHoffmann M, Kleine-Weber H, Schroeder S, et al.: SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor. Cell. 2020; 181(2): 271–80.e8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nQureshi AI, Abd-Allah F, Al-Senani F, et al.: Management of acute ischemic stroke in patients with COVID-19 infection: Report of an international panel. Int J Stroke. 2020; 15(5): 540–54. PubMed Abstract | Publisher Full Text\n\nDandona P, Dhindsa S, Ghanim H, et al.: Angiotensin II and inflammation: the effect of angiotensin-converting enzyme inhibition and angiotensin II receptor blockade. J Hum Hypertens. 2007; 21(1): 20–7. PubMed Abstract | Publisher Full Text\n\nPaliogiannis P, Mangoni AA, Dettori P, et al.: D-Dimer Concentrations and COVID-19 Severity: A Systematic Review and Meta-Analysis. Front Public Health. 2020; 8: 432. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYu HH, Qin C, Chen M, et al.: D-dimer level is associated with the severity of COVID-19. Thromb Res. 2020; 195: 219–225. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAn SJ, Kim TJ, Yoon BW: Epidemiology, Risk Factors, and Clinical Features of Intracerebral Hemorrhage: An Update. J Stroke. 2017; 19(1): 3–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuang X, Liu G, Guo J, et al.: The PI3K/AKT pathway in obesity and type 2 diabetes. Int J Biol Sci. 2018; 14(11): 1483–96. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHawkins BT, Lundeen TF, Norwood KM, et al.: Increased blood-brain barrier permeability and altered tight junctions in experimental diabetes in the rat:contribution of hyperglycaemia and matrix metalloproteinases. Diabetologia. 2007; 50(1): 202–11. PubMed Abstract | Publisher Full Text\n\nSong P, Li W, Xie J, et al.: Cytokine storm induced by SARS-CoV-2. Clin Chim Acta. 2020; 509: 280–287. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRochfort KD, Collins LE, Murphy RP, et al.: Downregulation of Blood-Brain Barrier Phenotype by Proinflammatory Cytokines Involves NADPH Oxidase-Dependent ROS Generation: Consequences for Interendothelial Adherens and Tight Junctions. PLoS One. 2014; 9(7): e101815. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBuzhdygan TP, DeOre BJ, Baldwin-Leclair A, et al.: The SARS-CoV-2 spike protein alters barrier function in 2D static and 3D microfluidic in-vitro models of the human blood–brain barrier. Neurobiol Dis. 2020; 146: 105131. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLapchak PA, Wu Q: Vascular Dysfunction in Brain Hemorrhage: Translational Pathways to Developing New Treatments from Old Targets. J Neurol Neurophysiol. 2011; 2011: S1–e001. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTsuge M, Yasui K, Ichiyawa T, et al.: Increase of tumor necrosis factor-alpha in the blood induces early activation of matrix metalloproteinase-9 in the brain. Microbiol Immunol. 2010; 54(7): 417–24. PubMed Abstract | Publisher Full Text\n\nZeni P, Doepker E, Schulze-Topphoff U, et al.: MMPs contribute to TNF-alpha-induced alteration of the blood-cerebrospinal fluid barrier in vitro. Am J Physiol Cell Physiol. 2007; 293(3): C855–64. PubMed Abstract | Publisher Full Text\n\nMountain DJ, Singh M, Menon B, et al.: Interleukin-1beta increases expression and activity of matrix metalloproteinase-2 in cardiac microvascular endothelial cells: role of PKCalpha/beta1 and MAPKs. Am J Physiol Cell Physiol. 2007; 292(2): C867–75. PubMed Abstract | Publisher Full Text\n\nRaymond L, Eck S, Mollmark J, et al.: Interleukin-1 beta induction of matrix metalloproteinase-1 transcription in chondrocytes requires ERK-dependent activation of CCAAT enhancer-binding protein-beta. J Cell Physiol. 2006; 207(3): 683–8. PubMed Abstract | Publisher Full Text\n\nBauer J, Huy C, Brenmoehl J, et al.: Matrix metalloproteinase-1 expression induced by IL-1β requires acid sphingomyelinase. FEBS Lett. 2009; 583(5): 915–20. PubMed Abstract | Publisher Full Text\n\nJu X, Ijaz T, Sun H, et al.: IL-6 regulates extracellular matrix remodeling associated with aortic dilation in a fibrillin-1 hypomorphic mgR/mgR mouse model of severe Marfan syndrome. J Am Heart Assoc. 2014; 3(1): e000476. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKossakowska AE, Edwards DR, Prusinkiewicz C, et al.: Interleukin-6 regulation of matrix metalloproteinase (MMP-2 and MMP-9) and tissue inhibitor of metalloproteinase (TIMP-1) expression in malignant non-Hodgkin's lymphomas. Blood. 1999; 94(6): 2080–9. PubMed Abstract\n\nHuang X, Dai S, Dai J, et al.: Luteolin decreases invasiveness, deactivates STAT3 signaling, and reverses interleukin-6 induced epithelial-mesenchymal transition and matrix metalloproteinase secretion of pancreatic cancer cells. Onco Targets Ther. 2015; 8: 2989–3001. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSiwik DA, Colucci WS: Regulation of Matrix Metalloproteinases by Cytokines and Reactive Oxygen/Nitrogen Species in the Myocardium. Heart Fail Rev. 2004; 9(1): 43–51. PubMed Abstract | Publisher Full Text\n\nYang C, Hawkins KE, Doré S, et al.: Neuroinflammatory mechanisms of blood-brain barrier damage in ischemic stroke. Am J Physiol Cell Physiol. 2019; 316(2): C135–c53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFreeman LR, Keller JN: Oxidative stress and cerebral endothelial cells: regulation of the blood-brain-barrier and antioxidant based interventions. Biochim Biophys Acta. 2012; 1822(5): 822–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGu Y, Dee CM, Shen J: Interaction of free radicals, matrix metalloproteinases and caveolin-1 impacts blood-brain barrier permeability. Front Biosci (Schol Ed). 2011; 3: 1216–31. PubMed Abstract | Publisher Full Text\n\nJian Liu K, Rosenberg GA: Matrix metalloproteinases and free radicals in cerebral ischemia. Free Radic Biol Med. 2005; 39(1): 71–80. PubMed Abstract | Publisher Full Text\n\nLakhan SE, Kirchgessner A, Tepper D, et al.: Matrix metalloproteinases and blood-brain barrier disruption in acute ischemic stroke. Front Neurol. 2013; 4: 32. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang Y, Estrada EY, Thompson JF, et al.: Matrix metalloproteinase-mediated disruption of tight junction proteins in cerebral vessels is reversed by synthetic matrix metalloproteinase inhibitor in focal ischemia in rat. J Cereb Blood Flow Metab. 2007; 27(4): 697–709. PubMed Abstract | Publisher Full Text\n\nLiu J, Jin X, Liu KJ, et al.: Matrix metalloproteinase-2-mediated occludin degradation and caveolin-1-mediated claudin-5 redistribution contribute to blood-brain barrier damage in early ischemic stroke stage. J Neurosci. 2012; 32(9): 3044–57. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrilha S, Ong CWM, Weksler B, et al.: Matrix metalloproteinase-9 activity and a downregulated Hedgehog pathway impair blood-brain barrier function in an in vitro model of CNS tuberculosis. Sci Rep. 2017; 7(1): 16031. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKeep RF, Andjelkovic AV, Xiang J, et al.: Brain endothelial cell junctions after cerebral hemorrhage: Changes, mechanisms and therapeutic targets. J Cereb Blood Flow Metab. 2018; 38(8): 1255–1275. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKazmierski R, Michalak S, Wencel-Warot A, et al.: Serum tight-junction proteins predict hemorrhagic transformation in ischemic stroke patients. Neurology. 2012; 79(16): 1677–85. PubMed Abstract | Publisher Full Text\n\nVoirin AC, Perek N, Roche F: Inflammatory stress induced by a combination of cytokines (IL-6, IL-17, TNF-α) leads to a loss of integrity on bEnd.3 endothelial cells in vitro BBB model. Brain Research. 2020; 1730: 146647. PubMed Abstract | Publisher Full Text\n\nCohen SS, Min M, Cummings EE, et al.: Effects of Interleukin-6 on the Expression of Tight Junction Proteins in Isolated Cerebral Microvessels from Yearling and Adult Sheep. Neuroimmunomodulation. 2013; 20(5): 264–73. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFeng J, Ito M, Ichikawa K, et al.: Inhibitory phosphorylation site for Rho-associated kinase on smooth muscle myosin phosphatase. J Biol Chem. 1999; 274(52): 37385–90. PubMed Abstract | Publisher Full Text\n\nOzaki H, Ishii K, Horiuchi H, et al.: Cutting edge: combined treatment of TNF-alpha and IFN-gamma causes redistribution of junctional adhesion molecule in human endothelial cells. J Immunol. 1999; 163(2): 553–7. PubMed Abstract\n\nAssémat E, Bazellières E, Pallesi-Pocachard E, et al.: Polarity complex proteins. Biochim Biophys Acta. 2008; 1778(3): 614–30. PubMed Abstract | Publisher Full Text\n\nBauer HC, Krizbai IA, Bauer H, et al.: “You Shall Not Pass”—tight junctions of the blood brain barrier. Front Neurosci. 2014; 8: 392. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKruse C, Kurz ARM, Pálfi K, et al.: Polarity Protein Scrib Facilitates Endothelial Inflammatory Signaling. Arterioscler Thromb Vasc Biol. 2015; 35(9): 1954–62. PubMed Abstract | Publisher Full Text\n\nLizama CO, Zovein AC: Polarizing pathways: balancing endothelial polarity, permeability, and lumen formation. Exp Cell Res. 2013; 319(9): 1247–54. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMichel D, Arsanto JP, Massey-Harroche D, et al.: PATJ connects and stabilizes apical and lateral components of tight junctions in human intestinal cells. J Cell Sci. 2005; 118(Pt 17): 4049–57. PubMed Abstract | Publisher Full Text\n\nBrinkmann BF, Steinbacher T, Hartmann C, et al.: VE-cadherin interacts with cell polarity protein Pals1 to regulate vascular lumen formation. Mol Biol Cell. 2016; 27(18): 2811–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDe Maio F, Lo Cascio E, Babini G, et al.: Improved binding of SARS-CoV-2 Envelope protein to tight junction-associated PALS1 could play a key role in COVID-19 pathogenesis. Microbes Infect. 2020; 22(10): 592–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTeoh KT, Siu YL, Chan WL, et al.: The SARS coronavirus E protein interacts with PALS1 and alters tight junction formation and epithelial morphogenesis. Mol Biol Cell. 2010; 21(22): 3838–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRajendran P, Rengarajan T, Thangavel J, et al.: The vascular endothelium and human diseases. Int J Biol Sci. 2013; 9(10): 1057–69. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVioli F, Oliva A, Cangemi R, et al.: Nox2 activation in Covid-19. Redox Biol. 2020; 36: 101655. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDiNicolantonio JJ, McCarty M: Thrombotic complications of COVID-19 may reflect an upregulation of endothelial tissue factor expression that is contingent on activation of endosomal NADPH oxidase. Open Heart. 2020; 7(1): e001337. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchmitt FCF, Manolov V, Morgenstern J, et al.: Acute fibrinolysis shutdown occurs early in septic shock and is associated with increased morbidity and mortality: results of an observational pilot study. Ann Intensive Care. 2019; 9(1): 19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGuan Wj, Ni Zy, Hu Y, et al.: Clinical Characteristics of Coronavirus Disease 2019 in China. N Engl J Med. 2020; 382(18): 1708–1720. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHirano T, Murakami M: COVID-19: A New Virus, but a Familiar Receptor and Cytokine Release Syndrome. Immunity. 2020; 52(5): 731–733. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHu B, Huang S, Yin L: The cytokine storm and COVID-19. J Med Virol. 2020; 10: 1002. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVazquez-Garza E, Jerjes-Sanchez C, Navarrete A, et al.: Venous thromboembolism: thrombosis, inflammation, and immunothrombosis for clinicians. J Thromb Thrombolysis. 2017; 44(3): 377–385. PubMed Abstract | Publisher Full Text\n\nBucciarelli P, Martinelli I, Artoni A, et al.: Circulating microparticles and risk of venous thromboembolism. Thromb Res. 2012; 129(5): 591–7. PubMed Abstract | Publisher Full Text\n\nMagro C, Mulvey JJ, Berlin D, et al.: Complement associated microvascular injury and thrombosis in the pathogenesis of severe COVID-19 infection: A report of five cases. Transl Res. 2020; 220: 1–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nQi X, Keith KA, Huang JH: COVID-19 and stroke: A review. Brain Hemorrhages. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBikdeli B, Madhavan MV, Jimenez D, et al.: COVID-19 and Thrombotic or Thromboembolic Disease: Implications for Prevention, Antithrombotic Therapy, and Follow-Up: JACC State-of-the-Art Review. J Am Coll Cardiol. 2020; 75(23): 2950–2973. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMecca AP, Regenhardt RW, O’Connor TE, et al.: Cerebroprotection by angiotensin‐(1–7) in endothelin‐1‐induced ischaemic stroke. Exp physiol. 2011; 96(10): 1084–96. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSampaio WO, Nascimento AAS, Santos RAS: Systemic and regional hemodynamic effects of angiotensin-(1– 7) in rats. Am J Physiol Heart Circ Physiol. 2003; 284(6): H1985–94. PubMed Abstract | Publisher Full Text\n\nSweid A, Hammoud B, Bekelis K, et al.: Cerebral ischemic and hemorrhagic complications of coronavirus disease 2019. Int J Stroke. 2020; 15(7): 733–742. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFraiman P, Junior CG, Moro E, et al.: COVID-19 and Cerebrovascular Diseases: A Systematic Review and Perspectives for Stroke Management. Front Neurol. 2020; 11: 574694. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "77660",
"date": "24 Feb 2021",
"name": "Ismail Setyopranoto",
"expertise": [
"Reviewer Expertise Neurologist"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article is very good and relevant to the conditions of the COVID-19 pandemic, maybe it only needs to add differences in imaging examinations, namely head CT scans between intracerebral hemorrhages due to degenerative diseases, for example due to hypertension and those caused by COVID-19 infection.\nI agree that this article can be indexed with some additions that I have written above.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Yes\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes",
"responses": []
},
{
"id": "80518",
"date": "04 Mar 2021",
"name": "Guilherme Welter Wendt",
"expertise": [
"Reviewer Expertise Quantitative research methods",
"Systematic reviews and meta-analyses",
"Epidemiology",
"Public health."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is the report of a registered systematic review (SR; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=224470). The authors correctly deployed the methodology proposed, in addition to adding a few more objectives - such as the chronological order of symptoms. In summary, the study not only updated a previous SR on this specific topic but largely expanded the mechanisms underpinning both ischemic and hemorrhagic stroke. The authors also followed the PRISMA guidelines and assessed risk of bias using a well-known scale. The statistical analyses are sound and based on best practices for conducting SR and meta-analyses.\nSome questions and suggestions that emerged while reading the paper include:\nIt would be sensible to adopt American (i.e., hemorrhagic) or British (i.e., haemorrhagic) English in the manuscript, including subheadings and figures.\n\nThe flowchart presents “excluded studies” in the last stage, in which authors are supposed to present studies that were included. I suggest a revision of this last step to avoid confusion.\n\nIt is not clear if all patients were diagnosed by RT-PCR as suggested in the registration (CRD42020224470). I assume this was done at some stage during the screening/study eligibility stages. Nonetheless, if suspected cases were also included, this information needs to be clearly stated.\n\nWas there any cut-off score or other criteria for including/excluding studies due to risk of bias assessments?\n\nFigures 2 to 4 are misspelling “Hernández”.\n\nIn the discussion, I think it is always a good idea to be precise as possible. As such, I would always refer to COVID-19 patients when making comparisons (i.e., “This result is lower than a previously reported incidence”, which appears in the first paragraph of the discussion).\n\nIn the second paragraph of the discussion, when the authors mention 216 COVID-19 patients, are we really relying on RT-PCR confirmed cases? In other passages, the authors are more clear about the method of diagnosis (“Meanwhile, 5.33% were admitted owing to neurological symptoms related to ischemic stroke and had a positive result on real-time polymerase chain reaction (RT-PCR) testing for SARS-CoV-2 after screening; there was no clear explanation for hospital admission in the remainder (89.15%)”.\n\nThe authors could briefly expand or give examples of neurological symptoms. For instance, the next sentence says altered mental status, albeit I do believe more objective symptoms would be of extreme value.\nOther than that, I believe the contribution is original, relevant and has substantial contribution to the field.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Yes\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes",
"responses": []
},
{
"id": "80516",
"date": "24 May 2021",
"name": "Mahir Gachabayov",
"expertise": [
"Reviewer Expertise Clinical outcomes and Evidence Synthesis"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a well-conducted systematic review and a well-written manuscript evaluating the prevalence, clinical presentations, and mortality rates of hemorrhagic and ischemic stroke among COVID-19 patients. In addition, the authors have attempted to describe the pathogenesis behind stroke in such patients.\nThe research question makes sense. The objective of this systematic review was to shed light on a question with limited evidence in the current literature. The Introduction is well written and details the gap and controversy in the current literature.\nThe protocol of the review was prospectively registered in PROSPERO. No deviations from the original protocol were observed. The report complies with the PRISMA guidelines. The search strategy is comprehensive and was reported in detail. Major databases were utilized as data sources. The methodology of eligibility screening and study selection, quality assessment, and data extraction is adequate. Statistical methods are adequate as well.\nResults are well-presented in tables summarizing included studies and forest plots. The prevalence of hemorrhagic and ischemic strokes were found to be low at 0.5% and 1%, respectively. Nonetheless, the mortality rates of these entities were quite high at 45% and 36%. These findings are clinically relevant and add to the literature.\nDiscussion is comprehensible. Interpretation of the statistical findings is logical and the conclusions are justified.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Yes\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-34
|
https://f1000research.com/articles/10-30/v1
|
18 Jan 21
|
{
"type": "Case Report",
"title": "Case Report: Fatal myocarditis after combination of immune checkpoint inhibitor and platinum doublet chemotherapy",
"authors": [
"Vilde Drageset Haakensen",
"Solfrid M.H. Thunold",
"Geeta Gulati",
"Solfrid M.H. Thunold",
"Geeta Gulati"
],
"abstract": "Myocarditis is a rare but feared adverse event of treatment with immune checkpoint inhibitors. The incidence is about 1%, while the mortality rate is up to 50%. Many physicians treating lung cancer do not have experience with this serious adverse event, and few hospitals in Scandinavia have routines for baseline assessments that may help detect early signs of inflammation of the myocard. Early onset of anti-inflammatory treatment is associated with favourable outcome.\n\nWe present a case report of fatal myocarditis after treatment with check-point inhibitor. The patient was admitted with severe dyspnoea, general body ache, weakness, dysphagia, palpitations and diplopia two weeks after her second infusion of systemic treatment combining chemotherapy and immunotherapy. She had presented vague symptoms at time of the second infusion that were not identified as related to myocarditis/myositis. Upon aggressive treatment with methylprednisolone, mycophenolate mofetil, abatacept and plasmapheresis, her troponins and pro-BNP were reduced, but clinically she deteriorated and her life could not be saved. We present this case report to increase awareness of the condition and to raise discussion about the role of routine baseline assessments to aid early diagnosis and anti-inflammatory treatment to prevent treatment-related deaths.",
"keywords": [
"myocarditis",
"myositis",
"immunotherapy",
"checkpoint inhibitor",
"adverse events"
],
"content": "Introduction\n\nMyocarditis is a rare but feared adverse event of treatment with immune checkpoint inhibitors (ICI). The incidence is about 1%, while the mortality rate is up to 50%. The incidence of lung cancer in Norway is 117 per 100 000 with 3200 patients a year. 30% of patients are treated with surgery or stereotactic radiation. The remaining patients are eligible to treatment with ICI in first or later lines. ICI has been used for lung cancer in Norway since 2016. Still, most physicians in Norway treating lung cancer have not experienced ICI-induced myocarditis. An informal query to six university hospitals and eight local hospitals in the South-East Health Region in Norway revealed no routine examination (blood or imaging) at baseline to help detect development of ICI-induced myocarditis later. An electrocardiogram (ECG) is usually part of the diagnostic work-up before cancer treatment, but only two hospitals specify taking ECG before start of immunotherapy in later lines. Our impression is that the situation is similar in the other Nordic countries (after personal communication through the Nordic Cardio-Oncology Board). The literature indicates the importance of early diagnosis, as a retrospective study found favourable outcome with onset of corticosteroid treatment within 24 hrs of admission to hospital1. We present this case to increase awareness, to advocate early diagnosis and treatment of ICI-related myocarditis.\n\n\nCase\n\nA woman in her late sixties, working freelance without any exposure and never smoking, with paroxysmal atrial fibrillation and a subvalvular aortic stenosis had a right upper lung lobectomy for an epidermal growth factor receptor (EGFR) positive adenocarcinoma. Post-surgery, she was treated with adjuvant chemotherapy and radiotherapy. A relapse two years later was treated with erlotinib and paused after a year due to gradual progression. After 18 months, upon progression, carboplatin/pemetrexed/pembrolizumab was started. When seen for the second infusion, she had slightly increased dyspnoea and was feeling tender on palpation of the chest wall medial of her right breast and the food felt as if it stopped in her oesophagus if she ate too quickly. There was no cough or fever. Clinical examination, c-reactive protein (CRP), haematology and electrolytes were normal. Troponins and creatine kinase (CK) were not measured. Computerized tomography (CT) showed no signs of pneumonitis and carboplatin/pemetrexed/pembrolizumab treatment was continued.\n\nTwo weeks later she was admitted to the oncology department with increased dyspnoea, general body ache, muscular weakness, dysphagia, episodes of palpitations and diplopia. Clinical examination revealed ptosis. Her blood work-up showed troponin T 651 (<14) ng/L, NT-proBNP 6761 (<760) ng/L, CRP 6.4 (<5) mg/L, aspartate aminotransferase (AST) 453 (<35) U/L (grade 3 toxicity), alanine aminotransferase (ALT) 186 (<45) U/L (grade 2), lactate dehydrogenase (LDH) 1111 (<255) U/L and creatine kinase (CK) 5845 (<210) U/L. Her ECG showed sinus rhythm and a new left bundle branch block (Figure 1A). CT excluded pulmonary embolism. Three days later she developed a third-degree atrioventricular block (Figure 1B) and her clinical symptoms progressed. Echocardiography showed no regional wall motion abnormalities. Cardiac magnetic resonance imaging (CMR) was attempted, but aborted as the patient was unstable because of her third-degree atrioventricular block. A two-chamber pacemaker was implanted and a coronary angiography was done and showed no signs of acute coronary syndrome. Later, repeated echocardiography was unchanged from the initial echocardiography, the left ventricular ejection fraction (LVEF) continued to be normal of 58%. With the presenting symptoms, troponin elevation, third degree atrioventricular block and no significant stenosis on the coronary angiogram, the patient fulfilled the criteria of myocarditis2 and it was concluded the patient had ICI-induced myocarditis with myositis and myasthenia gravis-like syndrome.\n\nElectrocardiogram at admission after the second infusion of combination therapy (A) and three days after admission (B).\n\nIntravenous methylprednisolone 2 mg/kg/day was initiated on day 3 and increased to 1000 mg/day for four days as her troponins increased (Figure 2). As there was a lack of clinical and biochemical response, abatacept 1000 mg once every second week and oral mycophenolate mofetil 750 mg b.i.d was initiated. This improved the biochemical parameters, but the clinical symptoms were unchanged. On day 14 plasmapheresis with exchange of 2l plasma was performed, but could not be repeated as the patient did not tolerate the exchange. However, the biochemical values improved further and her ptosis, dyspnoea and general muscle weakness improved. Unfortunately, her dysphagia and dyspnoea kept deteriorating. On day 25, abatacept and mycophenolate mofetil was stopped and the patient died five days later.\n\nSerum levels of troponin T increase during the first 10 days. Normalization of creatine kinase and myoglobin may reflect improvement of myositis, while the myocarditis did not resolve. Daily treatment regimens are indicated as follows: methylprednisolone X= 2 mg/kg, #= 1000 mg × 1; abatacept X=20 mg/kg; mycophenolate mofetil X= 750 mg × 2, #=500 mg x 2, +=500 mg × 1.\n\n\nDiscussion\n\nMyocarditis is a rare incidence (0.06-1.14%)3,4 off-target effect of ICI with mortality rates up to 50%5. The symptoms are often non-specific as fatigue, dyspnoea or chest pain. In severe cases cardiac arrhythmias as atrioventricular block or ventricular arrhythmias can occur. The myocarditis can be combined with myositis and/or myasthenia gravis like syndrome6. With increasing use of immunotherapy alone and in combination with other therapies, an increase in immune mediated myocarditis can be expected. Combination immunotherapy has a higher incidence than distribution of a PD-1/PD-L1 inhibitor alone7. There are few studies reporting the incidence of myocarditis after combination of ICI and chemotherapy. Keynote 189 reported one case of myocarditis in the treatment arm with pembrolizumab, carboplatin and pemetrexed (of 405 patients)8.\n\nAs the incidence of myocarditis is rare, the symptoms non-specific and because in the acute setting, the patients are often seen by medical staff with less experience of immunotherapy off-target effects, the diagnosis may be missed unless there are routines for baseline examinations and pre-defined early diagnostic work-ups. As a minimum, before initiation of immunotherapy, ECG should be taken and troponins and NT-proBNP measured. There should be a low threshold for repeating these measures during treatment, particularly if the patient is admitted to the hospital. After onset of symptoms suspicious of myocarditis, coronary angiography can rule out acute coronary syndrome, while CMR can often detect myocardial oedema along with late gadolinium contrast enhancement (LGE) as a sign of inflamed myocardium. However, in a recent study of 103 patients with ICI-associated myocarditis LGE was only present in 48%, other signs of oedema as elevated T2-weighted short tau inversion recovery (STIR) were present in 28%. The presence of LGE or T2-weighted STIR were not associated with major adverse cardiovascular events (composite of cardiovascular death, cardiogenic shock, cardiac arrest and complete heart block)9. In our patient CMR was not performed as she was unstable because of a third-degree atrioventricular block and implantation of a pacemaker was prioritized. Diagnosis was based on the diagnostic criteria for clinically suspected myocarditis as recommended by the European Society of Cardiology (ESC)2. Early immune suppression is of the essence, and hence, clinical awareness and diagnostic routines are important. In a recent retrospective study, high dose methylprednisolone (1000 mg/d) and early initiation (<24h) were associated with improved cardiac outcomes1. Combination of several anti-inflammatory therapies and plasmapheresis may reverse the serious adverse event6,10.\n\n\nConclusion\n\nEducation of medical staff in various hospital departments and in the community is necessary and may increase awareness of ICI-associated myocarditis. Clear routines for standard investigations prior to and during treatment with immunotherapy may reduce time to diagnosis, increase awareness and hopefully reduce mortality of myocarditis.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nConsent\n\nWritten informed consent for publication of their clinical details and/or clinical images was obtained from the patient.",
"appendix": "References\n\nZhang L, Zlotoff DA, Awadalla M, et al.: Major adverse cardiovascular events and the timing and dose of corticosteroids in immune checkpoint inhibitor-associated myocarditis. Circulation. 2020; 141(24): 2031–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCaforio ALP, Pankuweit S, Arbustini E, et al.: Current state of knowledge on aetiology, diagnosis, management, and therapy of myocarditis: A position statement of the European Society of Cardiology Working Group on Myocardial and Pericardial Diseases. Eur Heart J. 2013; 34(33): 2636–48, 2648a–2648d. PubMed Abstract | Publisher Full Text\n\nMartins F, Sofiya L, Sykiotis GP, et al.: Adverse effects of immune-checkpoint inhibitors: epidemiology, management and surveillance. Nat Rev Clin Oncol. 2019; 16(9): 563–80. PubMed Abstract | Publisher Full Text\n\nHaanen JBAG, Carbonnel F, Robert C, et al.: Management of toxicities from immunotherapy: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2017; 28(suppl_4): iv119–iv142. PubMed Abstract | Publisher Full Text\n\nSalem JE, Manouchehri A, Moey M, et al.: Cardiovascular toxicities associated with immune checkpoint inhibitors: an observational, retrospective, pharmacovigilance study. Lancet Oncol. 2018; 19(12): 1579–1589. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHu J, Florido R, Lipson EJ, et al.: Cardiovascular toxicities associated with immune checkpoint inhibitors. Cardiovasc Res. 2019; 115(5): 854–68. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMahmood SS, Fradley MG, Cohen J V, et al.: Myocarditis in Patients Treated With Immune Checkpoint Inhibitors. J Am Coll Cardiol. 2018; 71(16): 1755–64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGadgeel S, Rodríguez-Abreu D, Speranza G, et al.: Updated Analysis From KEYNOTE-189: Pembrolizumab or Placebo Plus Pemetrexed and Platinum for Previously Untreated Metastatic Nonsquamous Non-Small-Cell Lung Cancer. J Clin Oncol. 2020; 38(14): 1505–17. PubMed Abstract | Publisher Full Text\n\nZhang L, Awadalla M, Mahmood SS, et al.: Cardiovascular magnetic resonance in immune checkpoint inhibitor-associated myocarditis. Eur Heart J. 2020; 41(18): 1733–43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSalem JE, Allenbach Y, Kerneis M: Abatacept for severe immune checkpoint inhibitor-associated myocarditis. N Engl J Med. 2019; 380(24): 2377–2379. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "79530",
"date": "15 Feb 2021",
"name": "Panagiotis T. Diamantopoulos",
"expertise": [
"Reviewer Expertise Melanoma",
"Immunotherapy",
"myelodysplastic syndrome"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIt was a pleasure to review this interesting case report by Haakensen et al. on the emergence of a complex neurologic/cardiac irAE in a patient with lung adenocarcinoma treated with pembrolizumab. Although the case is interesting, there are already several case reports in the literature on the emergence of such irAEs in patients treated with checkpoint inhibitors. Moreover, the investigation of the patient is somewhat incomplete, thus a definitive diagnosis cannot be made with accuracy. You will find my comments in detail below.\nMajor points\nIntroduction, paragraph 1 lines 8-20. The information provided here is mainly hypothetical and the methodology of the authors does not follow any scientific research rules. I suggest it to be removed. In any case it does not add to the manuscript.\n\nEnglish editing by a native speaker is highly recommended.\n\nCase, paragraph 1, lines 9-13. The description of the symptoms and signs of the patient should be more accurate. Moreover, medical terms such as dysphagia/odynophagia etc. should be used. Moreover, the findings of a detailed neurologic examination should be provided. Since there are clear signs of neurologic deterioration, did the treating doctors investigate the patient with a brain CT/MRI or with a lumbar puncture?\n\nCase. I suggest that the authors provide figures of CT scans, echocardiography, and coronary angiography.\n\nWhy did the doctors stop immunosuppression 5 days before the death of the patient? More information is needed on the decision.\n\nDid the treating doctors administer pyridostigmine to the patient, especially since the authors correctly identified the condition as myasthenia gravis with cardiac involvement? How did they support the diagnosis? Did they measure anti-AChR antibodies? I believe that the diagnosis of myasthenia gravis, although obvious, should be further supported.\n\nDoes the very low incidence of cardiac irAEs justify the measurement of troponin and NT-proBNP before treatment initiation, as suggested by the authors (Discussion, paragraph 2, lines 7-8)? The authors should provide relevant references if available.\n\nData availability: the authors should state that upon reasonable request the presented data is available.\n\nSince this is a case of concomitant development of neurologic and cardiac irAEs, the authors should reference some relevant studies such as those by Suzuki S et al. “Nivolumab-related myasthenia gravis with myositis and myocarditis in Japan\"1 or my own paper, Diamantopoulos PT et al. “Concomitant development of neurologic and cardiac immune-related adverse effects in patients treated with immune checkpoint inhibitors for melanoma”2.\nMinor points\nIntroduction, paragraph 1, line 4. “3200 new cases per year” instead of “3200 patients a year”.\n\nIntroduction, paragraph 1, line 5. The sentence should not start with a number – please rephrase.\n\nCase, paragraph 1, line 1. What does “working freelance without any exposure” mean?\n\nCase, paragraph 1, line 7. “And paused” seems like referring to the relapse. Please rephrase.\n\nCase, paragraph 1, line 8. What does “gradual progression” mean? Please provide more information and CT scan pictures at baseline and at disease progression.\n\nIs the background of the case’s history and progression described in sufficient detail? No\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? No\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? No\n\nIs the case presented with sufficient detail to be useful for other practitioners? No",
"responses": [
{
"c_id": "6421",
"date": "08 Mar 2021",
"name": "Vilde Drageset Haakensen",
"role": "Author Response",
"response": "We thank Dr Diamantopoulos for reviewing our case report. The first aim of this report was to increase clinicians’ awareness of immune checkpoint inhibitor (ICI) induced myocarditis/myositis. Even though case reports of ICI myocarditis are found and recommendations on how to follow-up and treat these patients are published, clinically the routine of cardiac baseline examination and follow-up could be improved. The introduction is based on statistics. The experiences gathered from Scandinavian colleagues are used to emphasize the lack of cardiovascular baseline stratification of these patients in the clinical setting. The second aim was to discuss adequate measures to prevent fatal outcomes by discussing the use of baseline and follow-up troponin and proBNP. The use of these biomarkers and other surveillance tools have already been justified in a Position statement from the Cardio-Oncology Study Group of the Heart Failure Association and the Cardio-Oncology Council of the European Society of Cardiology (Pudil et al. Eur J Heart Fail. 2020). The patient was not examined by brain CT/MRI or lumbar puncture or anti-AChR measurements and pyridostigmine was not used. We agree that neurologists should be involved in such cases and that reference from the neurology field would improve the discussion. Steroids were stopped five days before the patient died as she was clinically deteriorating and not tolerating the aggressive immunosuppressive therapy. We agree that additional imaging of the heart and tumor would enhance the quality of this case."
},
{
"c_id": "6495",
"date": "25 Mar 2021",
"name": "Vilde Drageset Haakensen",
"role": "Author Response",
"response": "We do not wish for additional reviews of this case report."
}
]
}
] | 1
|
https://f1000research.com/articles/10-30
|
https://f1000research.com/articles/8-2094/v1
|
12 Dec 19
|
{
"type": "Research Article",
"title": "The relationship between Helicobacter pylori infection and intestinal parasites in individuals from Khartoum state, Sudan: a case-control study",
"authors": [
"Yasir Yousif Abd Elbagi",
"Ahmed Bakheet Abd Alla",
"Mohammed Baha Eldin Saad",
"Yasir Yousif Abd Elbagi",
"Mohammed Baha Eldin Saad"
],
"abstract": "Background: In developing countries, Helicobacter pylori infection is common, as are intestinal parasites. Socioeconomic circumstances and low personal hygiene lead to the spread of these infections. This research aimed to evaluate the relationship between intestinal parasites and H. pylori in Khartoum, Sudan. Methods: This study was conducted in various hospitals in Khartoum between June and October 2018. The study involved 200 individuals: 100 patients with H. pylori as a case group and 100 healthy individuals as a control group. A stool sample was taken from each individual, and wet preparation, saturated sodium chloride flotation and formal ether concentration were used to detect intestinal parasites. Results: The results showed that 23% of H. pylori patients and 10% of healthy individuals had gastrointestinal parasites; Entamoeba histolytica was found in 12% of H. pylori cases followed by Entamoeba coli (7%) and Giardia lamblia (4%). Control group: Entamoeba histolytica in 5% followed by G. lamblia in 3% and E. coli in 2% of individuals. There was a significant difference in the prevalence of intestinal parasites between groups (P = 0.013). The prevalence rate of intestinal parasites among men and women was 24% and 22%, respectively, in the case group, and 9% and 11%, respectively, in the control group. In the case group, the highest prevalence rates (40% and 38%) were found among the age groups 1-15 and 46-60 years old, respectively, while the lowest rate (10.7%) was found among the 31-45 age group. In the control group, the highest prevalence rate (15%) was among the 31-45 age group, the lowest prevalence rate (8%) was found among the 16-30 age group. Conclusion: Together, we found that intestinal parasites are more common in patients with H. pylori. We also noticed that the rate of infection was not affected by gender while the age group was affected.",
"keywords": [
"H.pylori",
"Intestinal parsaite",
"Khartoum",
"Alsaha",
"Yastabsheron."
],
"content": "Introduction\n\nAmong the most common diseases in the world are intestinal parasite infections; an estimated 3.5 billion people are affected, and 450 million people are infected (Jayalakshmi & Dharanidevi, 2016). These infections are considered a serious public health problem as they cause anaemia with iron deficiency, retardation of growth in children, and other physical and mental health problems (Okyay et al., 2004; Tandukar et al., 2013; Wongstitwilairoong et al., 2007). One example is a pathogenic intestinal protozoon that infects the small and/or large intestine (Farthing & Kelly, 2005), or an intestinal worm, such as Ascaris lumbricoides, Trichuris trichiura, Enterobius vermicularis, and hookworms, which affect people in tropical countries (Smyth, 1990).\n\nHealth impacts differ with age: the small intestinal protozoa Giardia lamblia and Cryptosporidium spp. have a serious impact on children (Harhay et al., 2010), while the large intestine pathogen Entamoeba histolytica has a higher morbidity among adults of all ages (Mortimer & Chadee, 2010). Some protozoa, in particular, Cryptosporidium and Isospora belli, cause significant morbidity in individuals with immunodeficiency (Bachur et al., 2008), for example giardiasis and amoebiasis are opportunistic parasites (Biggs & Brown, 2001). Helicobacter pylori is the most common chronic human bacterial infection, infecting 70–90% of the population of developing countries and 25–50% of the people of developed countries (Gillespie & Hawkey, 2006). H. pylori colonizes the stomach’s mucus layer and induces chronic active gastritis inflammation (Konturek, 2003). It is a major cause of peptic ulcers and a risk factor for gastric malignancies (Lesbros-Pantoflickova et al., 2007).\n\nH. pylori can be easily identified in all microbiology laboratories using simple techniques (Guerrant et al., 2011). Numerous serological diagnostic tests used for the detection of H. pylori include bacterial agglutination, complement fixation, indirect immunofluorescence test, enzyme immunoassay, and enzyme-linked immunosorbent assay (Kim, 2016). Since H. pylori and intestinal parasites are prevalent in developing countries, this study aimed to determine the prevalence of intestinal parasites in patients with H. pylori in Khartoum state.\n\n\nMethods\n\nThis was a case-control study carried out in Khartoum state, Sudan, at Alsaaha Specialized Hospital and Yastabshiron Hospital between 1 June and 27 October 2018. The study was conducted in 100 patients who were H. pylori positive as a case group and 100 individuals without H. pylori as a control group. The H. pylori test had already been performed in hospitals using immuno-chromatographic test (ICT) for identification of H. pylori Ag in stool sample. Participants were divided into groups according to gender and age (see below).\n\nApproval for the study was obtained from the Ethical Authorization Committee (number, MLS-IEC-10-17) of the College of Medical Laboratory Sciences, Sudan University of Science and Technology. Written informed consent for participation and disclosure of data was obtained from each participant in the study and in the case of children (<18 years) written informed consent was obtained from their guardians.\n\nIndividuals in this study who were already being screened for presence of H. pylori using ICT for antigen in stool were asked to participate. Individuals with a positive ICT for H. pylori antigen were included in the case group after they agreed to participate in the study, while the control group were those with negative H. pylori antigen, who were also only included in the study after they agreed to participate.\n\nIn total, 200 stool samples were collected from the participants in the study. Samples were collected immediately after participants agreed to partake in the study. Each participant was provided with a labelled stool container (transparent and clean) and was instructed to collect a faecal sample.\n\nEvery stool sample was examined for the detection of the intestinal parasite by wet preparation, saturated sodium chloride floatation and formal ether concentration. If one detection method was positive, then the sample was counted as positive for intestinal parasites, even if other methods were negative.\n\nWet preparation. A small portion of stool was mixed with a drop of normal saline with a wooden applicator stick and deposited on a slide. This was covered with a cover slip and routinely examined under a microscope using 10X and a high magnification 40X to detect more parasites, as per the World Health Organization protocol (WHO, 2001).\n\nFormal ether concentration. Approximately 1 g of faeces from various parts of the stool was collected and emulsified in a glass beaker in 5 ml of formal saline. There a further 5 ml of saline was added and mixed. The resulting suspension was strained using a sieve with small pores. The filtered sample was poured into a centrifugal tube, and an equal volume of ether was added. For one minute, the tube was mixed and then centrifuged at 2000 rpm for 5 minutes. The upper three layers were discarded, and the sediment was moved to a slide, covered with a cover slip and analysed under a microscope using magnifications 10X and 40X. This was as per the protocol by Smith & Mank (2011).\n\nSaturated sodium chloride floatation. Approximately a 0.5 g of faeces was collected from different parts of the stool and emulsified in a long glass tube half-filled with saturated sodium chloride solution. Then the container was filled with sodium chloride until the top of tube. Carefully, a cover glass was put on the top of the tube avoiding air bubbles. After 30 to 45 minutes, the cover glass was removed from the top of the tube and put on a clean and dry slide and examined under the microscope using 10X and 40X magnifications. This was as per the protocol by Dryden et al. (2005).\n\nStatistical analysis was performed using SPSS version 20.0. The Chi-square method was used to compare variables. P values < 0.05 were considered statistically significant.\n\n\nResults\n\nThe results showed that 23 of the 100 patients with H. pylori were infected with gastrointestinal parasites (23%). Of the 100 control individuals, 10 were found to be infected with gastrointestinal parasites (10%). Between the case and control groups, there was a statistically significant difference in prevalence of intestinal parasites (P = 0.013).\n\nAmong H. pylori patients, the occurrence of intestinal parasites in men and women was similar (24% and 22%, respectively; P = 0.841; Table 1). On the other hand, the prevalence of gastrointestinal parasites in men and women in the control group was found to be 9% and 11%, respectively, but this difference was not statistically significant (P = 0.789; Table 1).\n\nCase difference between groups: P = 0.841; Control difference between groups: P = 0.789.\n\nIn the case group, the highest occurrence rates (40% and 38%) were reported among the 1–15 and 46–60 age groups, while the lowest rate (10.7%) was reported among the 31–45 age group. These differences were not statistically significant (P. value= 0.132; Table 2). For the control group, the highest prevalence rate (15%) was reported among the 31–45 age group, while the lowest prevalence rate was among the 16–30 age group (8%). This difference was not statistically significant (P = 0.528; Table 2).\n\nCase difference between age groups: P = 0.132; Control difference between age groups: P = 0.528.\n\nThe results showed that Entamoeba histolytica was seen in 12% of H. pylori cases followed by Entamoeba coli in 7% and G. lamblia in 4% of cases (Table 3). Among the control group E. histolytica was reported at 5%, followed by G. lamblia at 3% and E. coli at 2% (Table 3).\n\n\nDiscussion\n\nFrom the study, it is evident that the gastrointestinal parasite overall occurrence among H. pylori patients is relatively high (23%). It was found that this rate was higher than the published rate by Uğraş & Miman (2014) in Turkey (7.61%). As far as the control group is concerned, the overall occurrence rate reported was 10%. This rate is lower than the rate among H. pylori patients and higher than the rate reported by Uğraş & Miman (2014). The difference in rates between the control group and patient with H. pylori was significant. This, in our opinion, might mean that there is an association between the establishment of gastrointestinal parasites and H. pylori.\n\nThe difference in prevalence rates between men and women in H. pylori patients and control individuals was not statistical difference. This finding did not agree with Yakoob et al. (2005), who considered G. lamblia occurrence in Pakistan. That study found a higher rate in men (72%) than in women (28%).\n\nIn our study, the highest occurrences (40% and 38%) were reported among the 1–15 and 46–60 year age groups, respectively, in the H. pylori patients, and the 31–45 year age group (15%) for the control group. Our finding disagreed with the finding of Fadul et al. (2016), who reported the highest occurrence rate (50%) in the age group >66 years old. Our results also showed that E. histolytica was seen in 12% of the H. pylori cases followed by E. coli in 7% of cases and G. lamblia in 4%. Lower rates were reported among the control group where E. histolytica was seen in 5% followed by G. lamblia in 3% and E. coli in 2%. Our result are not in line with the findings of Gökşen et al. (2016) who reported 14.8% for G. lamblia in the H. pylori-positive group, which was in agreement with Escobar-Pardo et al. (2011) who also found a significant association between H. pylori and G. lamblia. However, our conclusion was in total disagreement with the finding of Uğraş & Miman (2014), who reported no significant association between H. pylori and intestinal parasites in Turkey. This may be due to differences in the study areas with different food and life style of individuals.\n\n\nConclusions\n\nGastrointestinal parasites are more common among H. pylori patients compared to individuals without H. pylori; but this infection rate was not affected by gender. The highest infection rate was reported in the 1–15 and 46–60 age group among H. pylori patients and 31–45 years of age group among the control patients.\n\n\nData availability\n\nFigshare: yasir and ahmed.sav, https://doi.org/10.6084/m9.figshare.10315769.v2 (Abd Alla & Yousif, 2019).\n\nThis project contains the following underlying data:\n\n- Raw data file.sav\n\n- Data dictionary\n\nData are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).",
"appendix": "References\n\nAbd Alla A, Yousif Y: yasir and ahmed.sav. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.10315769.v2\n\nBachur TP, Vale JM, Coêlho IC, et al.: Enteric parasitic infections in HIV/AIDS patients before and after the highly active antiretroviral therapy. Braz J Infect Dis. 2008; 12(2): 115–122. PubMed Abstract | Publisher Full Text\n\nBiggs B, Brown G: Principles and Practice of Clinical Parasitology. Principles and Practice of Clinical Parasitology. Edited by Stephen H.gillespie and R. D. Pearson. 2001. Publisher Full Text\n\nDryden MW, Payne PA, Ridley R, et al.: Comparison of common fecal flotation techniques for the recovery of parasite eggs and oocysts. Vet Ther. 2005; 6(1): 15–28. PubMed Abstract\n\nEscobar-Pardo ML, de Godoy AP, Machado RS, et al.: Prevalence of Helicobacter pylori infection and intestinal parasitosis in children of the Xingu Indian Reservation. J Pediatr (Rio J). 2011; 87(5): 393–8. PubMed Abstract | Publisher Full Text\n\nFadul N, Ahmed M, Elamin T, et al.: Prevalence Rate Of Giardia Lamblia / Helicobacter pylori Co-Infections In Khartoum State, Sudan. International Journal of Scientific & Technology Research. 2016; 5(3): 181–190. Reference Source\n\nFarthing MJ, Kelly P: Protozoal gastrointestinal infections. Medicine. 2005; 33(4): 81–83. Publisher Full Text\n\nGillespie SH, Hawkey PM: Principles and Practice of Clinical Bacteriology Second Edition. Principles and Practice of Clinical Bacteriology. 2006. Publisher Full Text\n\nGökşen B, Appak YÇ, Girginkardesler N, et al.: Coexistence of Helicobacter pylori and Intestinal Parasitosis in Children with Chronic Abdominal Pain. Turkiye Parazitol Derg. 2016; 40(1): 32–6. PubMed Abstract | Publisher Full Text\n\nGuerrant RL, Walker DH, Weller PF: Tropical infectious diseases: principles, pathogens and practice. third edit. Elsevier Inc. 2011. Publisher Full Text\n\nHarhay MO, Horton J, Olliaro PL: Epidemiology and control of human gastrointestinal parasites in children. Expert Rev Anti Infect Ther. 2010; 8(2): 219–234. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJayalakshmi S, Dharanidevi S: The Prevalence of Intestinal Parasitic Infections in a tertiary care hospital in southern India-A retrospective study. Int J Curr Microbiol App Sci. 2016; 5(10): 718–23. Publisher Full Text\n\nKim N: Helicobacter pylori. Edited by N. Kim. Springer Nature. 2016. Publisher Full Text\n\nKonturek JW: Discovery by Jaworski of Helicobacter pylori and its pathogenetic role in peptic ulcer, gastritis and gastric cancer. J Physiol Pharmacol. 2003; 54 Suppl 3: 23–41. PubMed Abstract\n\nLesbros-Pantoflickova D, Corthésy-Theulaz I, Blum AL: Helicobacter pylori and probiotics. J Nutr. 2007; 137(3 Suppl 2): 812S–8S. PubMed Abstract | Publisher Full Text\n\nMortimer L, Chadee K: The immunopathogenesis of Entamoeba histolytica. Exp Parasitol. 2010; 126(3): 366–380. PubMed Abstract | Publisher Full Text\n\nOkyay P, Ertug S, Gultekin B, et al.: Intestinal parasites prevalence and related factors in school children, a western city sample--Turkey. BMC Public Health. 2004; 4: 64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmith HV, Mank TG: Diagnosis of human giardiasis. In Giardia. Springer, Vienna. 2011; 353–377. Publisher Full Text\n\nSmyth JD: Foundations of parasitology (4th edn). 8th edn, Parasitology Today. 8th edn. Edited by P. E. Reidy. Americas, New York: Janice Roerig-Blong. 1990. Publisher Full Text\n\nTandukar S, Ansari S, Adhikari N, et al.: Intestinal parasitosis in school children of Lalitpur district of Nepal. BMC Res Notes. 2013; 6(1): 449. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUğraş M, Miman O: [The prevalence of intestinal parasites in children with Helicobacter pylori gastritis evaluated retrospectively] Turkiye Parazitol Derg. 2014; 37(4): 245–248. PubMed Abstract | Publisher Full Text\n\nWongstitwilairoong B, Srijan A, Serichantalergs O, et al.: Intestinal parasitic infections among pre-school children in Sangkhlaburi, Thailand. Am J Trop Med Hyg. 2007; 76(2): 345–350. PubMed Abstract | Publisher Full Text\n\nWorld Health Organization: Guidelines on standard operating procedures for laboratory diagnosis of HIV-oportunistic infections (No. SEA-HLM-332). WHO Regional Office for South-East Asia. 2001. Reference Source\n\nYakoob J, Jafri W, Abid S, et al.: Giardiasis in patients with dyspeptic symptoms. World J Gastroentrol. 2005; 11(42): 6667–6670. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "67157",
"date": "15 Jul 2020",
"name": "Elena Pomari",
"expertise": [
"Reviewer Expertise Infectious diseases"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAuthors describe the occurrence of H. pylori and intestinal parasites concomitant infections in individuals from a specific state in Sudan.\nMajor revision:\n.The rationale of the study is focused on the \"in developing countries\" and the discussion is mostly on studies conducted in Turkey. To the best of my knowlegde, the current (or recent) Turkey GDP is not so low. Thus, authors should compare their findings with other available data reported in studies conducted for example in Egypt, Ethiopia ... (doi: 10.12816/00108551; 10.1186/s13104-018-3246-42, etc).\n\nIn the discussion, authors conclude suggesting that different food and life style of individuals among different geo regions might determine various occurrence of infections. Did authors have any information about diet, life style of the included subjects? These variables could be added in tables and analysis.\nMinor revision:\nKeywords: check typing for \"Intestinal parsaite\" and \"Yastabsheron\"\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "67152",
"date": "21 Jul 2020",
"name": "Paweł Krzyżek",
"expertise": [
"Reviewer Expertise Microbiology",
"mostly focused on H. pylori"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article is a short report which aims to show the higher frequency of selected parasites in H. pylori-positive patients. I believe the manuscript is worth indexing, however, I suggest to apply two corrections:\nI think that it would be easier for readers to give p-values in tables as an independent column (this way it will be easy to read which results are statistically different from each other).\n\nI believe that the discussion should be extended to the immunological part, even though it was not the aim of this article. The results obtained by the Authors indicate a higher frequency of parasites in H. pylori-positive patients aged 1-15 and 46-60. These results indicate that an immature/weakened immune system may influence the parasite-H. pylori relationship in the host. Thus, I suggest that the Authors should discuss this topic and add relevant literature references.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "67158",
"date": "27 Jul 2020",
"name": "Asmaa Ibrahim",
"expertise": [
"Reviewer Expertise Infectious diseases",
"mostly focused on H. pylori and Cryptosporidium"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article aims to evaluate the prevalence rate of intestinal parasites in H. pylori positive patients in a specific state in Sudan.\nMajor revision:\nThe introduction needs to be more updated (ex. Haque (2007) and Nyantekyi et al., (2010) instead of (Smyth, 1990), both articles associated with helminthic infections in developing countries.\n\nIn the method of stool examination part, I prefer to make also examination with acid-fast stain (AF), in case of diarrhea or immunodeficiency to detect coccidian parasites.\n\nIn data analysis part, if there are any data about socio-demographic characteristics such as (residency in rural or urban areas, animal contact, and water source) and clinical symptoms such as (diarrhea, vomiting and fever) were collected, I prefer to analyze its association with the infection as estimated risk factors (multivariate logistic regression, 95% CI and OR).\n\nThe current study associated with the \" developing countries\" especially Sudan, while the discussion is focused on studies conducted a specific country “Turkey”, which mainly not considered as a developing country. I prefer to compare the present study with other studies in Egypt (10.1007/s12639-018-1075-y1), Iran (10.5812/pedinfect.152942) and Ethiopia (10.1186/s13104-018-3246-43). These countries more suitable for that topic.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "67155",
"date": "10 Aug 2020",
"name": "Yeong Yeh Lee",
"expertise": [
"Reviewer Expertise Helicobacter pylori",
"gut microbiota",
"gastrointestinal cancers"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study adds to a body of literature supporting the common presence of H. pylori and intestinal parasites co-infection. There are several concerns as below:\nBest if the study design involved age and gender-matched cohorts. Case-control might be biased especially selection bias.\n\nSample size calculation - how 100 samples each arm is decided?\n\nIdeally, two serological tests for H. pylori are needed to confirm infection.\n\nOnly limited parasites were detected. Please provide explanations.\n\nPlease discuss reasons for co-infection of detected parasites with H. pylori, the significance, and discuss the age group differences found between current study and others.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-2094
|
https://f1000research.com/articles/10-29/v1
|
18 Jan 21
|
{
"type": "Data Note",
"title": "Draft genome assembly of the biofuel grass crop Miscanthus sacchariflorus",
"authors": [
"Jose De Vega",
"Iain Donnison",
"Sarah Dyer",
"Kerrie Farrar",
"Iain Donnison",
"Sarah Dyer",
"Kerrie Farrar"
],
"abstract": "Miscanthus sacchariflorus (Maxim.) Hack. is a highly productive C4 perennial rhizomatous biofuel grass crop. M. sacchariflorus is among the most widely distributed species in the genus, particularly at cold northern latitudes, and is one of the progenitor species of the commercial M. × giganteus genotypes. We generated a 2.54 Gb whole-genome assembly of the diploid M. sacchariflorus cv. “Robustus 297” genotype, which represented ~59% of the expected total genome size. We later anchored this assembly using the chromosomes from the M. sinensis genome to generate a second assembly with improved contiguity. We annotated 86,767 and 69,049 protein-coding genes in the unanchored and anchored assemblies, respectively. We estimated our assemblies included ~85% of the M. sacchariflorus genes based on homology and core markers. The utility of the new reference for genomic studies was evidenced by a 99% alignment rate of the RNA-seq reads from the same genotype. The raw data, unanchored and anchored assemblies, and respective gene annotations are publicly available.",
"keywords": [
"Miscanthus",
"biofuel",
"C4",
"assembly",
"annotation"
],
"content": "Introduction\n\nMiscanthus is a genus of C4 perennial rhizomatous grasses native to East Asia and Oceania, and naturally adapted to a wide range of climate zones and land types. Miscanthus sacchariflorus is among the most widely distributed species within the genus. It originated in the Yellow Sea region of China and can be predominantly found in cool latitudes of East Asia with varying ploidy1. M. sacchariflorus occurs in both diploid (2n = 38) and tetraploid (2n = 76) forms, where tetraploid M. sacchariflorus genotypes originated by autopolyploidy2. M. sacchariflorus probably has the greatest winter hardiness among all the Saccharinae3.\n\nNatural interspecific Miscanthus hybrids are commonly observed, even between individuals of different ploidy. For example, introgression of M. sacchariflorus is often found among cultivated European M. sinensis ecotypes1,4. Furthermore, M. x giganteus, a sterile triploid hybrid resulting from the hybridization between M. sinensis and M. sacchariflorus, is the predominant commercially grown species owing to its high biomass productivity and low chemical input requirements. The common occurrence of hybridization events and variable ploidy are challenging to the improvement of these bioenergy grasses and increase the need for genomic resources from different Miscanthus species. A chromosomal-scale reference genome using a doubled-haploid M. sinensis line was recently published4.\n\nWe assembled, annotated and validated a draft genome from the diploid M. sacchariflorus cv. “Robustus 297” genotype, as well as generating rhizome, stem and leaf RNA-Seq data from the same genotype. This dataset was previously used to verify that both M. sinensis and M. sacchariflorus share the same A/B ancestral tetraploidy4. Here, we present the first draft genome of M. sacchariflorus, the second Miscanthus genome available after M. sinensis4.\n\n\nMethods\n\nDNA was extracted from leaves from the diploid M. sacchariflorus cv. “Robustus 297” genotype (Biosample SAMN08580354) using the Qiagen DNeasy kit. RNA was also extracted from leaf, stem and root tissues from the same plant. All samples were taken from a plant grown from seed in trays in a glasshouse in 2009. This genotype is established and used in breeding at IBERS (Wales, UK). The RNA-seq libraries were deposited as part of previous work in the BioProject PRJNA639832.\n\nWe obtained ~5.86e9 pairs of 100 bp paired-end reads from an Illumina paired-end library with a 560 bp insert-size that was sequenced on Illumina HiSeq 2500 machines in rapid run mode by the Earlham Institute. This represents approximately 50X coverage of the heterozygous content and 100X coverage of the homozygous content of the genome. Read quality was assessed, and contaminants and adaptors removed using Kontaminant5. These paired-end short-reads were assembled into 17M contigs with a total length of 3.27 Gb using ABySS6 version 1.5.1, with default options and a kmer size of 71.\n\nWe obtained ~141.1e6 pairs of reads from a Nextera 150 bp mate-pair library with approximately 7 Kb insert-size, which was used for scaffolding the previous contigs together with the paired-end reads, using SSPACE7 without “extension” step. Nextera mate-pair reads were required to include a fragment of the adaptor to be used in the scaffolding step5, and we filtered out sequences shorter than 500 bp. We obtained 589K scaffolds, a total length of 2.54 Gb with an N50 of 10.2 Kb. This whole-genome assembly was denominated “Msac_v2” and is deposited at NCBI in BioProject PRJNA679435.\n\nOur gene structure annotation pipeline8 used five sources of evidence that were provided to AUGUSTUS9 (version 2.7) for gene annotation: (1) Repetitive and low complexity regions of the scaffolds identified using RepeatMasker10 (version open-4.0.5) based on homology with the RepBase11 public database (Release 20140131) and a new database of repeat elements identified in the assembly with RepeatModeler12. The repeats annotation was deposited in Zenodo (See data availability); (2) exon-intron junctions identified by Tophat13 (version 2.1.0); (3) de novo and genome-guided ab initio transcripts assembled with Trinity14 (version 2.6.5 )and Cufflinks15 (version 2.2.1) from RNA-Seq reads obtained from several tissues from the same genotype; (4) ab initio gene models predicted by SNAP16 (version 29-11-2013) and GeneID17 (version 1.4.4); and (5) homology-based alignments of transcripts and proteins from Miscanthus sinensis and maize using Exonerate18 with a minimal identity of 0.7 and coverage of 0.7. Finally, AUGUSTUS9 was run with the options “genemodel=complete” and “alternatives-from-evidence=true” to ensure that the predicted genes were compatible with all the previous provided evidence.\n\nFor the functional annotation of these predicted genes, translated gene sequences were compared with the NCBI non-redundant (nr 20170116) proteins and EBI’s InterPro (version 5.22.61) databases, and the results were imported into Blast2GO19 to annotate the GO and GO slim terms, enzymatic protein codes and KEGG pathways. A similar GO annotation from translated gene sequences can be done with eggNOG-mapper20. These functional descriptors were deposited in Zenodo (See Underlying data).\n\nTo improve the genome contiguity, we anchored our M. sacchariflorus scaffolds to the Miscanthus sinensis genome4. However, no nucleotide content from M. sinensis was incorporated in the M. sacchariflorus assemblies.\n\nFirstly, scaffolds longer than 2 kbps from the whole genome assembly “Msac_v2” were scaffolded again using SSPACE7 and the M. sinensis mate-pairs reads, the gaps between scaffolds were filled in with Ns. This new whole-genome assembly was denominated “Msac_v3”, and was deposited at NCBI in Bioproject PRJNA435476, under the GenBank accession GCA_002993905. It contains 137,916 scaffolds for a total of 2.074 Gb with an N50 of 25.6 Kbps. The gene annotation was projected to the “Msac_v3” assembly using PASA21 (version 2.0.1): genes were aligned to the new assembly using GMAP, requiring a minimum identity of 0.85 and coverage of 0.55, and later validated using the default parameters in PASA.\n\nFinally, we obtained the chromosomal position in the M. sinensis chromosomes of the scaffolds from the “Msac_v3” assembly. Using Satsuma222 (version untagged-330e3341a1151a978b37), we identified every perfect-identify match between both assemblies (3,635,504 matches in total). The coordinates of these matches in BED 8 format were used as input to the “OrderOrientBySynteny” script from Satsuma2, which identifies the best chromosomal position for each scaffold. These position coordinates are available as an AGP file as part of GCA_002993905, which anchors our final whole-genome assembly to 19 chromosomes (accessions CM00959 to CM009609 in NCBI).\n\nRNA-seq cleaned reads from each tissue were independently aligned to both assembly versions using STAR23 (version 2.6.0c). BUSCO24 (version 4.1.4) was used to assess completeness with the single-copy orthologs database for green plants (Viridiplantae, version 2020-09-10). Orthologs were identified using Orthofinder225 (version 2.3.12) with default parameters and the option “-msa”, which directly provided comprehensive statistics comparing the provided proteomes. All the proteomes from the other species used (Table 1) were downloaded from Phytozome (v7.1 DOE-JGI). Genomes were aligned using Minimap226 (version 2.17) with the “asm10” parameter for related genomes, secondary alignments (tp:A:S) filtered out, and results visualised using dotPlotly27 (Github version, latest updated on 4 May 2018).\n\n*15 scaffolds from plastids were discarded during the deposit in NCBI resulting in 137,916 scaffolds. ** Only the longest transcript was considered in each projected locus. *** Cross-species alignments.\n\n\nResults\n\nWe produced two whole-genome assemblies for M. sacchariflorus that we named “Msac_v2” and “Msac_v3”, with total lengths of 2.54 Gbps and 2.074 Gbps, respectively (Table 1). The difference in size is mainly a result of filtering 402 Mb from sequences under 2 kb in the latter before anchoring to the M. sinensis genome. Our “Msac_v2” assembly covered ~59 % of M. sacchariflorus genome size, which is estimated to be 4.3 Gb28. Approximately 40% of the assembly was composed by transposable elements (987.3 Mb; Table 2), including 491 Mb (19.4%) and 154 Mb (6.1%) by copies of the Gypsy and Copia LTRs, respectively; and 180 Mb (7.1%) by several class 2 DNA transposons (MULE, CMC, Harbinger, etc.)\n\nWe identified 219,394 primary alignments longer than 2 kb between the unanchored M. sacchariflorus (“Msac_v2”) and M. sinensis. The resulting dotplot (Figure 1) shows the conserved synteny between both species, which diverged 1.6 Mya4. Figure 1 also shows the highly conserved synteny between the pairs of homoeologous chromosomes (e.g. green boxes in chromosomes one and two), and the fusion in chromosome 7 of the chromosome homeolog to chromosome 13; which was also reported in M. sinensis4. There are several large inversions between chromosomes 9 and 10, and 3 and 4 (cyan boxes in Figure 1). Our assembly of a heterozygous genotype resulted in multiple heterotigs (heterozygous contigs) containing the alternative or secondary haplotypes (e.g. pink boxes in Figure 1).\n\nThe plot shows the primary alignments longer than 2 kbps between both species. The M. sacchariflorus scaffolds (Y-axis) have been sorted by their coordinates in M. sinensis chromosomes (X-axis). Large homoeologous blocks and chromosomal rearrangements are highlighted in boxes.\n\nThe utility of our assemblies for genomic studies is evidenced by the proportion of RNA-seq from three different tissues from the same M. sacchariflorus genotype that aligned to the assemblies. On average 99% and 95% of the RNA-seq reads aligned in “Msac_v2” and “Msac_v3”, respectively (Table 1).\n\nWe estimated that we assembled more than 85% of the M. sacchariflorus genes. Furthermore, our assemblies include several alleles of genes in the heterozygous regions of the genome, while the M. sinensis reference was generated from a double-haplotyped genotype. The estimation of the proportion of assembled genes (~85%) was supported by (1) the results from BUSCO, which reported 86.4–87.7% of presented core genes, of which ~2/3rds were complete (Table 1); and (2) the difference in the number of proteins from related species for which we can identify an ortholog in M. sacchariflorus compared to M. sinensis, as control, using Orthofinder2 (Table 3).\n\nBased on the results from Orthofinder2 (Table 3), we found orthologs in M. sacchariflorus for 64.5% of the M.sinensis annotated proteins, so we estimated ~1/3rd of the Miscanthus proteins to be specific to each species. On the other hand, we estimated that ~3,000 genes may be missing in the “Msac_v2” annotation based on the number of Sorghum bicolor proteins with orthologues in M. sinensis but absent in M. sacchariflorus. Better estimations were obtained with the other four species, where the genes absent in Msac_v2 compared with M. sinensis were estimated to be 254, 579 and 1627 (Table 3). Additionally, ~6,000 genes could be missed in “Msac_v3” compared to “Msac_v2” based on the difference in the number of M. sinensis orthologues in each assembly (Table 3). This is likely from genes in the sequences shorter than 2 Kbps (totalling 402 Mbps) that were filtered out before anchoring. There was a large difference in the proportion of “fragmented” BUSCO genes found in the M. sacchariflorus (32.2%) and M. sinensis (1.6%) assemblies (Table 1). To assess if that difference had an effect on the quality of the annotation, we compared the number of proteins from M. sacchariflorus and M. sinensis for which we can identify an ortholog in another species (Table 3); we found the difference between both Miscanthus species ranged between 6,571 proteins when compared to sorghum (43,475 to 37,219; Table 2) to only 121 when compared to maize (39,986 to 38,478, Table 3).\n\nIn conclusion, our M. sacchariflorus genome can served as the basis for functional genetic analyses on Miscanthus, one of the main biofuel grass crops used in temperate latitudes. However, there are opportunities to improve it using new approaches, such as long-reads.\n\n\nData availability\n\nNCBI BioProject: Miscanthus sacchariflorus cultivar:Robustus 297. Accession number PRJNA435476; https://identifiers.org/bioproject:PRJNA435476.\n\nThis BioProject contains the raw paired-end and mate-pair reads.\n\nNCBI BioProject: RNA-seq Miscanthus hybrids with contrasting phenotypes. Accession number PRJNA639832; https://identifiers.org/bioproject:PRJNA639832.\n\nThis BioProject contains RNA-seq reads, deposited as part of a previous project29.\n\nNCBI BioProject: Miscanthus sacchariflorus cultivar:Robustus 297. Accession number PRJNA679435; https://identifiers.org/bioproject:PRJNA679435.\n\nThis Bioproject contains the unanchored “Msac_v2” assemblies and gene annotations under accession JADQCR000000000.\n\nThe anchored “Msac_v3” assemblies and gene annotations are deposited under accession accession GCA_002993905 under Bioproject PRJNA435476.\n\nThe chromosomal positions in the M. sinensis chromosomes of the scaffolds from the “Msac_v3” assembly are available in an AGP file as part of GCA_002993905, which places the scaffolds in 19 chromosomes (accessions CM009591 to CM009609 in NCBI).\n\nZenodo: Supplementary dataset to \"Draft genome assembly of the biofuel grass crop Miscanthus sacchariflorus\". http://doi.org/10.5281/zenodo.4270235.\n\nThis project contains the assemblies in FASTA format, gene annotations in GFF3 format, functional annotations in tabulated text format, and AGP file with anchoring information.\n\nData deposited with Zenodo are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "References\n\nClark LV, Jin X, Petersen KK, et al.: Population structure of Miscanthus sacchariflorus reveals two major polyploidization events, tetraploid-mediated unidirectional introgression from diploid M. sinensis, and diversity centred around the Yellow Sea. Ann Bot. 2019; 124(4): 731–48. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDwiyanti MS, Rudolph A, Swaminathan K, et al.: Genetic analysis of putative triploid Miscanthus hybrids and tetraploid M. sacchariflorus collected from sympatric populations of Kushima, Japan. Bioenergy Res. 2013; 6(2): 486–93. Publisher Full Text\n\nClark LV, Dzyubenko E, Dzyubenko N, et al.: Ecological characteristics and in situ genetic associations for yield-component traits of wild Miscanthus from eastern Russia. Ann Bot. 2016; 118(5): 941–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMitros T, Session AM, James BT, et al.: Genome biology of the paleotetraploid perennial biomass crop Miscanthus. Nat Commun. 2020; 11(1): 5442. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeggett RM, Ramirez-Gonzalez RH, Clavijo BJ, et al.: Sequencing quality assessment tools to enable data-driven informatics for high throughput genomics. Front Genet. 2013; 4: 288. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSimpson JT, Wong K, Jackman SD, et al.: ABySS: a parallel assembler for short read sequence data. Genome Res. 2009; 19(6): 1117–23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoetzer M, Henkel CV, Jansen HJ, et al.: Scaffolding pre-assembled contigs using SSPACE. Bioinformatics. 2011; 27(4): 578–9. PubMed Abstract | Publisher Full Text\n\nDe Vega JJ, Ayling S, Hegarty M, et al.: Red clover (Trifolium pratense L.) draft genome provides a platform for trait improvement. Sci Rep. 2015; 5: 17394. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStanke M, Keller O, Gunduz I, et al.: AUGUSTUS: ab initio prediction of alternative transcripts. Nucleic Acids Res. 2006; 34(Web Server issue): W435–W9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTarailo‐Graovac M, Chen N: Using RepeatMasker to identify repetitive elements in genomic sequences. Curr Protoc Bioinformatics. 2009; 25(1): Chapter 4: Unit 4.10. PubMed Abstract | Publisher Full Text\n\nJurka J, Kapitonov VV, Pavlicek A, et al.: Repbase Update, a database of eukaryotic repetitive elements. Cytogenet Genome Res. 2005; 110(1–4): 462–7. PubMed Abstract | Publisher Full Text\n\nSmit AF, Hubley R: RepeatModeler Open-1.0. 2008. Reference Source\n\nTrapnell C, Pachter L, Salzberg SL: TopHat: discovering splice junctions with RNA-Seq. Bioinformatics. 2009; 25(9): 1105–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrabherr MG, Haas BJ, Yassour M, et al.: Trinity: reconstructing a full-length transcriptome without a genome from RNA-Seq data. Nat Biotechnol. 2011; 29(7): 644–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTrapnell C, Roberts A, Goff L, et al.: Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nat Protoc. 2012; 7(3): 562–78. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBromberg Y, Rost B: SNAP: predict effect of non-synonymous polymorphisms on function. Nucleic Acids Res. 2007; 35(11): 3823–35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBlanco E, Parra G, Guigó R: Using geneid to identify genes. Curr Protoc Bioinformatics. 2007; Chapter 4: Unit 4.3. PubMed Abstract | Publisher Full Text\n\nSlater GSC, Birney E: Automated generation of heuristics for biological sequence comparison. BMC Bioinformatics. 2005; 6(1): 31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nConesa A, Götz S, García-Gómez JM, et al.: Blast2GO: a universal tool for annotation, visualization and analysis in functional genomics research. Bioinformatics. 2005; 21(18): 3674–6. PubMed Abstract | Publisher Full Text\n\nHuerta-Cepas J, Forslund K, Coelho LP, et al.: Fast Genome-Wide Functional Annotation through Orthology Assignment by eggNOG-Mapper. Mol Biol Evol. 2017; 34(8): 2115–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHaas BJ, Delcher AL, Mount SM, et al.: Improving the Arabidopsis genome annotation using maximal transcript alignment assemblies. Nucleic Acids Res. 2003; 31(19): 5654–66. PubMed Abstract | Publisher Full Text | Free Full Text\n\nClavijo B, Wright J, Yanes L. Reference Source\n\nDobin A, Davis CA, Schlesinger F, et al.: STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013; 29(1): 15–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSimão FA, Waterhouse RM, Ioannidis P, et al.: BUSCO: assessing genome assembly and annotation completeness with single-copy orthologs. Bioinformatics. 2015; 31(19): 3210–2. PubMed Abstract | Publisher Full Text\n\nEmms DM, Kelly S: OrthoFinder: phylogenetic orthology inference for comparative genomics. Genome Biol. 2019; 20(1): 238. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi H: Minimap2: pairwise alignment for nucleotide sequences. Bioinformatics. 2018; 34(18): 3094–100. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPoorten T. Reference Source\n\nRayburn AL, Crawford J, Rayburn CM, et al.: Genome Size of Three Miscanthus Species. Plant Mol Biol Report. 2009; 27(2): 184. Publisher Full Text\n\nDe Vega JJ, Peel N, Purdy SJ, et al.: Differential expression of starch and sucrose metabolic genes linked to varying biomass yield in Miscanthus hybrids. BioRxiv. 2020; 2020–08. Publisher Full Text"
}
|
[
{
"id": "77622",
"date": "25 Feb 2021",
"name": "Diego Mauricio Riaño-Pachón",
"expertise": [
"Reviewer Expertise Bioinformatics",
"genome assembly and annotation."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe Data Note, \"Draft genome assembly of the biofuel grass crop Miscanthus sacchariflorus\", introduces two Miscanthus sacchariflorus genome assemblies, which have been deposited in NCBI under the bioprojects: PRJNA679435 and PRJNA435476. Genome sequencing was carried out with Illumina paired end reads and mate-pairs, the assemblies are greatly fragmented, which is expected due to the sequencing technologies used. This is the first Miscanthus sacchariflorus genome assembly, which is of interested for the bioenergy community, and can be used to generate insigths with the genomes of other bioenergy crops, like sorghum and sugarcane.\n\nSuggestions:\nLook for contaminant organisms in the final assemblies using BlobPlots.\n\nProvide GenomeScope and Smudgeplots for the clean reads, to generate further statistics prior to assembly.\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? No\n\nAre the datasets clearly presented in a useable and accessible format? Yes",
"responses": []
},
{
"id": "78040",
"date": "01 Mar 2021",
"name": "Maria Stefanie Dwiyanti",
"expertise": [
"Reviewer Expertise Plant genetics and genomics"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe availability of Miscanthus sacchariflorus genome sequence will be useful for Miscanthus related research, particularly in bioenergy related topics.\n\n\"Better estimations were obtained with the other four species, where the genes absent in Msac_v2 compared with M. sinensis were estimated to be 254, 579 and 1627 (Table 3).\"\nI found that the difference between number of genes in \"Msac_v2\" compared to other four species is larger than 254, 579, and 1627; or the way I look into the table is wrong?\n\nPerhaps the sentence above can be reworded so we can easily compare with the Table 3 content.\nAlso, what are the predicted functions of genes absent in \"Msac_v2\" compared to M.sinensis?\nThis information may provide some clues to trait difference between M. sacchariflorus and M.sinensis.\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-29
|
https://f1000research.com/articles/10-28/v1
|
18 Jan 21
|
{
"type": "Brief Report",
"title": "Evolution of SARS-CoV-2 virus and assessment of the effectiveness of COVID-19 vaccine",
"authors": [
"Veljko Veljkovic",
"Vladimir Perovic",
"Isabelle Chambers",
"Slobodan Paessler",
"Vladimir Perovic",
"Isabelle Chambers",
"Slobodan Paessler"
],
"abstract": "A safe and effective vaccine is urgently needed to bring the current SARS-CoV-2 pandemic under control. The spike protein (SP) of SARS-CoV-2 represents the principal target for most vaccines currently under development. Despite the presence of a CoV proof-reading function in viral replication, SP protein from SARS-CoV still extensively mutates, which might have an impact on current and future vaccine development. Here, we present analysis of more than 1600 SP unique variants suggesting that vaccine candidates based on the Wuhan-Hu-1 reference strain would be effective against most of currently circulated SARS-CoV-2 viruses, but that further monitoring of the evolution of this virus is important for identification of other mutations, which could affect the effectiveness of vaccines.",
"keywords": [
"COVID-19",
"SARS-CoV-2",
"mutations",
"vaccine effectiveness"
],
"content": "Introduction\n\nThe current Coronavirus Disease 2019 (COVID-19) pandemic, caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), represents an unprecedented health threat resulting in over 1.1 million deaths globally (www.worldometers.info/coronavirus) and significant economic damage. There is a consensus that development of an effective and safe vaccine represents the best strategy to control the COVID-19 pandemic, and consequently, this endeavor remains a top public health priority. Researchers and drug companies around the world are working hard to develop a COVID-19 vaccine, and currently 282 vaccine candidates are in clinical trials or will enter clinical investigation soon (clinicaltrial.gov). The spike glycoprotein (SP) of SARS-CoV-2 is the principal target for most these vaccine candidates. Despite the presence of a CoV proof-reading function in viral replication1, SARS-CoV still extensively mutates, which might have an impact on current and future vaccine development. In the GISAID database more than 1,600 SARS-CoV-2 strains with unique combinations of mutations in SP are deposited. In order to fully understand the impact of the mutations on the SP properties (pathogenesis, virulence, transmissibility, immunogenicity), advanced laboratory studies are required. These investigations take time and are done in close collaboration between different research groups, institutions and academia. However, some important information about the biological effects of mutations could and should also be obtained by the in silico analysis of SP, which can be performed quickly and with a minimal experimental foreknowledge at the time of analysis.\n\nPreviously, a novel bioinformatics approach, which is based on electronic biology, was used in the assessment of the effect of mutations on the vaccine effectiveness (VE) for seasonal flu vaccines. Most recently, this approach allowed successful prediction of VE for two successive flu seasons2,3. Here, we used this bioinformatics approach for the analysis of the effect(s) of mutations in SP from SARS-CoV-2 virus on the effectiveness of the COVID-19 vaccines and/or vaccine candidates. The presented results showed that VE of vaccine candidates that are based on SP from the reference SARS-CoV-2 strain YP_009724390 remains between 76% and 94%, despite an extensive mutation rate in this important and likely protective antigen. Our analysis also suggests that further close monitoring of SP mutations might be essential in order to be able to respond to changes that may influence VE in the future.\n\n\nMethods\n\nWe analyzed the non-redundant subunit 1 of S proteins (SP1) from human SARS-CoV-2 viruses deposited in the GISAID (https://www.epicov.org/epi3/cfrontend#18c7c7) from January to October, 2020, and SP1 from viruses isolated from mink that are presented in the same database. Prototype virus Wuhan-HU-1 (YP_009724390) is used in the analysis as the wild type (WT) and as the vaccine virus.\n\nThe ISM is the virtual spectroscopy method for the analysis of the protein biological properties4. This bioinformatics approach encompasses three basic steps: (i) the representation of the primary structure of the protein as the numerical sequence by the assignment to each amino acid of the corresponding value of the electron-ion interaction potential (EIIP), (ii) the transformation of the obtained numerical sequence into the informational spectrum (IS), and (iii) the calculation of the cross-spectrum (CS) between interacting proteins.\n\nThe EIIP is the physical parameter that determines the long-range interactions of biological molecules (interactions at distances >5Å)5. This molecular descriptor is defined by the following equation6,7:\n\n\n\nwhere Z* is the average quasivalence number (AQVN):\n\n\n\nwhere N is the total number of atoms, ni is the number of atoms of the i-th component, Zi is the valence number of the i-th atomic component in the molecule and m is the number of components. The EIIP values calculated according to the Equation (1) are given in Rydbergs (Ry).\n\nThe numerical sequence, representing the primary structure of protein, is transformed into the IS by the discrete Fourier transformation\n\n\n\nwhere x(m) is the m-th member of a given numerical series, N is the total number of points in this series, and X(n) are discrete Fourier transformation coefficients. In this way, the information defined by the sequence of amino acids is represented as the series of frequencies and their amplitudes. The frequencies in IS correspond to the distribution of the structural motifs with defined physicochemical properties determining a biological characteristics of a protein. When comparing proteins, which share the same biological or biochemical function, the ISM technique allows detection of code/frequency pairs that are specific for their common biological properties, or which correlate with their specific interactions.\n\nThe algorithm of the ISM-based phylogenetic analysis, which was used for the assessment of the biological effect of mutations in SP1 proteins from SARS-CoV-2, was previously described in detail8. Figure 1 provides the schematic presentation of this algorithm. Here, we used an ISM distance measure d defined on the ratio between specific frequencies F(0.257) and F(0.479) in IS of SP1 protein, which characterize its interaction with host factors9.\n\n\nResults and discussion\n\nFigure 2 presents the ISM-based phylogenetic tree (the high resolution tree is given in Extended data: Figure 1S10) calculated for 1643 nonredundant SP1 proteins from SARS-CoV-2 viruses collected from January to October, 2020. Analyzed SP1s in this tree are grouped into four separated clusters. The largest cluster A, encompassing 78% of analyzed proteins, contains two sub-clusters, which includes the Wuhan-Hu-1 reference sequence for the Spike protein that is also the basis for different vaccine candidates (sub-cluster A2) and viruses with SP1 with the most abundant mutation D614G (sub-cluster A1). According to the IS concept, proteins that are grouped in the same cluster in the ISM-base phylogenetic tree have similar interacting and immunological profiles9,11–13. This indicates that a vaccine based on SP from the Wuhan-Hu-1 reference strain will be effective against viruses that are grouped together with this virus in the sub-cluster A2. It was suggested that a single vaccine candidate based on the reference SARS-CoV-2 strain would likely match most of currently circulating variants, including the mutant D614G14. Recently, it also showed that sera from D614-infected hamsters exhibit modestly higher neutralization titers against G614 virus than against D614 virus, indicating that this mutation may not reduce the ability of vaccines in clinical trials to protect against COVID-1915. This suggests that viruses which are grouped together with the mutant D614G in the sub-cluster A1 would also be responsive to the vaccine candidates that are based on the Wuhan-Hu-1 reference strain. In contrast to the cluster A, viruses that are grouped into other three small clusters B, C and D do not perfectly match the vaccine virus and might be resistant to the vaccine.\n\nIn Table 1, the distribution by months of non-redundant SP1 from potentially vaccine responsive SARS-CoV-2 is reported. These data suggest that VE of the vaccine candidates, estimated on the base of the IS parameters would be between 76% and 95%. This estimation of VE is limited with the assumption that viruses from the cluster A dominate among currently circulating SARS-CoV-2. Despite this limitation, the predicted VE is in accord with the results of the recently completed Phase-3 clinical trial of the candidate COVID-19 vaccine. This study, which enrolled 43,538 subjects, showed that the tested vaccine candidate to be more than 90% effective in preventing COVID-19 in participants.\n\nAlthough potentially vaccine non-responsive viruses in the clusters B, C and D represent only 22% of all 1,643 viruses with non-redundant SP1, these viruses could negatively affect VE. Mass vaccination, which is expected in the near future, creates an immune pressure in the population that would lead to increase in mutations in SP1 as well as in natural selection and preservation of escape mutants. Moreover, this fraction of mutant viruses that can escape vaccine protection and be efficiently transmitted could significantly increase among circulating SARS-CoV-2 once vaccination starts. For this reason, real-time monitoring of virus evolution with different methods, including the ISM-based phylogenetic tool for detection of “functional” changes in the SP1 of SARS-CoV-2 is needed.\n\nIn September 2020 in Denmark, 12 human cases of COVID-19 have been identified with SARS-CoV-2 unique variants associated with farmed minks. Viruses isolated from these patients had a combination of mutations, or changes that have not been previously observed and the implications of the identified changes in this variant are not yet well understood. Preliminary findings indicate that this particular mink-associated variant identified in both minks and the 12 human cases could negatively affect VE because mutations in these viruses moderately decreased sensitivity to neutralizing antibodies. Danish authorities have undertaken different actions to limit the further spread of this variant of the virus among mink and human populations, including culling of all farmed mink in Denmark (more than 17 million).\n\nIn order to assess the effect of these mutations on VE we calculated the ISM-phylogenetic tree for 5 non-redundant SP1 sequences from SARS-CoV-2 viruses from mink deposited in GISAID (Figure 3). As presented, four of five viruses are grouped with the vaccine virus suggesting that these variants would not significantly affect VE of the COVID-19 vaccines that are based on the SP from the Wuhan-Hu-1 reference strain. The virus with the combination of mutations G261D and Y453F is out of this cluster and it could be potentially resistant to these vaccines. This suggests that further spread of this variant should be carefully monitored. Of note is that the virus with single mutation Y453F is grouped with the vaccine virus indicating that this variant will not significantly affect VE although this mutation is located in RBD. Contrary, this mutation in combination with G261D could decrease responsiveness to vaccine. This result also points out that the assessment of biological effect of mutations should be included in analysis all mutations that are present in protein.\n\n\nConclusions\n\nThe presented analysis of SP1 proteins, performed by the electronic biology tool, suggest (i) that vaccine candidates based on the Wuhan-Hu-1 reference strain would be effective against most of currently circulated SARS-CoV-2 variants, and ii) further ISM-based monitoring of the evolution of SARS-CoV-2 is important for identification of other mutations, which could affect the effectiveness of vaccines against this virus. Therefore, the scientific community needs to be proactive in submitting genetic sequences into databases such as GSAID. For example, we have States such as Wisconsin, South and North Dakota with low mortality rate 0.8, 1.2, 1.2, respectively (GISAID accessed on November 39, 2020), but no ability to check genetic sequences of viruses from those States as none were deposited. It is very important to support genetic data collection and to share that data in real time in order to organize successful monitoring using different approaches including the one published in this paper.\n\n\nData availability\n\nSequence data of the viruses were obtained from the GISAID Database. To access the database each individual user should complete the “Registration Form For Individual Users”. This form, together with detailed instructions, are available on the website. After submission of the Registration form, the user will receive a password. There are no other restrictions for access to GISAID. Conditions of access to, and use of, the GISAID Database and Data are defined by Terms of Use.\n\nHarvard Dataverse: The ISM-based phylogenetic tree of the non-redundant SP1 proteins from SARS-CoV-2 viruses deposited in GISAID from January to October 2020, https://doi.org/10.7910/DVN/JLXWWP10.\n\nThis project contains the following extended data:\n\n- Figure 1S\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).",
"appendix": "Acknowledgement\n\nThis work was supported by the research grant from Agricultural Department of North Dakota, USA, to Biomed Protection ND (Grant Number 20-342A2).\n\n\nReferences\n\nDenison MR, Graham RL, Donaldson EF, et al.: Coronaviruses: an RNA proofreading machine regulates replication fidelity and diversity. RNA Biol. 2011; 8(2): 270–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaessler S, Veljkovic V: Prediction of influenza vaccine effectiveness for the influenza season 2017/18 in the US [version 1; peer review: 2 approved]. F1000Res. 2017; 6: 2067. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaessler S, Veljkovic V: Using electronic biology based platform to predict flu vaccine efficacy for 2018/2019 [version 2; peer review: 2 approved, 1 approved with reservations]. F1000Res. 2018; 7: 298. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVeljkovic V, Cosic I, Dimitrijevic B, et al.: Is it possible to analyze DNA and protein sequences by the methods of digital signal processing? IEEE Trans Biomed Eng. 1985; 32(5): 337–41. PubMed Abstract | Publisher Full Text\n\nVeljkovic V: A theoretical approach to preselection of carcinogens and chemical carcinogenesis. Gordon & Breach, New York, 1980. Reference Source\n\nVeljkovic V, Slavic I: Simple general-model pseudopotential. Phys Rev Let. 1972; 29(2): 105. Publisher Full Text\n\nVeljkovic V: The dependence of the Fermi energy on the atomic number. Phys Lett. 1973; 45(1): 41–42. Publisher Full Text\n\nPerovic VR, Muller CP, Niman HL, et al.: Novel phylogenetic algorithm to monitor human tropism in Egyptian H5N1-HPAIV reveals evolution toward efficient human-to-human transmission. PLoS One. 2013; 8(4): e61572. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVeljkovic V, Vergara-Alert J, Segales J, et al.: Use of the informational spectrum methodology for rapid biological analysis of the novel coronavirus 2019-nCoV: prediction of potential receptor, natural reservoir, tropism and therapeutic/vaccine target. F1000Res. 2020; 9: 52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPerovic V: \"The ISM-based phylogenetic tree of the non-redundant SP1 proteins from SARS-CoV-2 viruses deposited in GISAID from January to October 2020.\" Harvard Dataverse, V1. 2020. http://www.doi.org/10.7910/DVN/JLXWWP\n\nVeljkovic V, Veljkovic N, Muller CP, et al.: Characterization of conserved properties of hemagglutinin of H5N1 and human influenza viruses: possible consequences for therapy and infection control. BMC Struct Biol. 2009; 9: 21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVeljkovic V, Niman HL, Glisic S, et al.: Identification of hemagglutinin structural domain and polymorphisms which may modulate swine H1N1 interactions with human receptor. BMC Struct Biol. 2009; 9: 62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVeljkovic V, Veljkovic N, Paessler S, et al.: Predicted enhanced human propensity of current avian-Like H1N1 swine influenza virus from China. PLoS One. 2016; 11(11): e0165451. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDearlove B, Lewitus E, Bai H, et al.: A SARS-CoV-2 vaccine candidate would likely match all currently circulating strains. bioRxiv. 2020. Publisher Full Text\n\nPlante JA, Liu Y, Liu J, et al.: Spike mutation D614G alters SARS-CoV-2 fitness. Nature. 2020. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "77630",
"date": "20 Jan 2021",
"name": "Ursula Dietrich",
"expertise": [
"Reviewer Expertise Virology",
"HIV-1",
"Influenza Viruses",
"vaccine development",
"virus entry",
"antigenicity"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article by Veljkovic et al. describes the application of an in silico method previously developed by the group to monitor the \"functional\" evolution of viral sequences based on the informational spectrum method (ISM). This method is based on the primary amino acid sequences of proteins, from which informational spectra are being derived that ultimately inform about protein-protein interactions. This method allows a rapid assessment of functional alterations in evolving proteins like changes in antigenicity or receptor usage and has been validated in previous studies of the Veljkovic group. A key study was the validation of mutations in the hemagglutinin sequences of H5N1 Influenza viruses predicted by ISM to alter receptor specificity in functional in vitro studies by Schmier et al. (2015), which should be mentioned in the introduction.1 Here this method is applied to analyze more than 1600 non-redundant sequences of the spike glycoprotein of the SARS-CoV-2 actually circulating and deposited in the GISAID database over time to derive their immunological profiles and to detect potential escape from the actual vaccine strains. The derived phylogenetic tree uncovers 2 major clusters corresponding to the Wuhan-Hu1 reference group and a derived D614G mutant covered by the actual vaccines and 3 minor clusters representing 22% of sequences, of which vaccine coverage is not known. These results nicely show how potential candidates for vaccine escape mutants can rapidly be identified by the ISM method, which can then be proven experimentally. The authors also identified by ISM a subgroup of mink SARS-CoV-2 sequences transmitted to humans in Denmark, which represent outliers compared to the actual vaccine strains and should thus be watched carefully in the future. Overall, this paper nicely underlines the value of the ISM method, which can rapidly be performed on a huge set of related sequences to identify potentially \"dangerous\" outliers, which may be functionally altered in terms of virulence or antigenic escape. This sets the ground to focus experimentally on these outlier strains for functional implications of the predicted mutations, thus limiting long-lasting experimental work to essential strains. As such, the manuscript potentially contributes essentially to the control of the most acute viral health problem we are actually facing. The paper is generally well written, minor linguistic issues should be corrected in the manuscript.\nMinor issues:\n\nAbstract: \"....most vaccines currently in clinical applications or under development\"; \"Here we present computational analysis of more than....\"; \"effective against most of currently circulating SARS-CoV-2 viruses. However, further monitoring....\".\n\nIntroduction: \"...in silico analysis of SP, which can be performed much more rapidly and with minimal experimental knowledge at the time of analysis. Furthermore, in a proof of principle study, it was shown that HA mutations predicted in silico to alter avian receptor specificity of H5N1 Influenza A viruses indeed result in enhanced specificity for human receptors (Schmier et al., Scientific Reports 2015).\n\nResults: In Fig. 2 it should be indicated, what percentage of sequences has been assigned to the different groups (A1, A2, B, C, D). Also in Table 1, it should be indicated in an additional column, what percentage of sequences belongs to the minor groups (B, C, D) over the months to detect a potential increase over time.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Not applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "148965",
"date": "20 Sep 2022",
"name": "Marwa Osama Elgendy",
"expertise": [
"Reviewer Expertise Clinical Pharmacy",
"Respiratory system diseases and COVID-19"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nDear Editor,\nI want to inform you that this article is an exciting piece of work. The study presented the analysis of SP unique variants using the electronic biology tool whereas SP is an important target for developing most of the SARS-CoV-2 vaccines. The authors found that the vaccine candidates based on the Wuhan-Hu-1 reference strain would be effective against most of the currently circulated SARS-CoV-2 variants and that ISM-based monitoring of the evolution of SARS-CoV-2 is important for the identification of other mutations, which could affect the effectiveness of vaccines against this virus.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-28
|
https://f1000research.com/articles/10-27/v1
|
18 Jan 21
|
{
"type": "Software Tool Article",
"title": "The Paramak: automated parametric geometry construction for fusion reactor designs.",
"authors": [
"Jonathan Shimwell",
"John Billingsley",
"Rémi Delaporte-Mathurin",
"Declan Morbey",
"Matthew Bluteau",
"Patrick Shriwise",
"Andrew Davis",
"John Billingsley",
"Rémi Delaporte-Mathurin",
"Declan Morbey",
"Matthew Bluteau",
"Patrick Shriwise",
"Andrew Davis"
],
"abstract": "During the conceptual design process of fusion reactors it is useful to rapidly prototype different design concepts and assess their suitability against a range of high level requirements. Rapid prototyping allows the 'fail early' mantra of other fields to be applied to engineering design. Furthermore, the rapid generation of low fidelity analysis allows fast exploration of design space, which enables better decisions to be made during concept selection and the detailed design phase. The Paramak is an open-source tool that aims to provide automated parameter driven 3D CAD models for fusion reactor components and magnetic fusion reactors. The geometry produced is compatible with several analysis workflows and this allows iterative automated model building and analysis to help steer the design concept optimisation process. The Paramak uses CadQuery 2 to create the 3D CAD model. The Paramak framework is used to create a few example reactor configurations including: a spherical reactor, a regular large radius tokamak and a compact submersion tank reactor. Input parameters for the various reactors that the Paramak can generate generally fall into three categories: continuous ranges such as blanket thickness, integer ranges such as number of toroidal field coils and categorical parameters such as type of divertor. The Paramak facilitates parameter studies where users can investigate the impact of input design parameters on the reactor performance. The use of modern software practices allows the geometry to be continuously tested in analysis workflows to ensure it is fit for purpose. The generation of output metrics from input parameters lends itself to the use of data science and machine learning approaches in order to steer the design. The Paramak provides rapid construction of analysis ready CAD in a manner that allows the designer to save time when exploring the design space for design studies and facilitate automated generative design.",
"keywords": [
"parametric",
"design",
"automated",
"fusion",
"energy",
"CAD"
],
"content": "Introduction\n\nWhen assessing the suitability of a fusion reactor design, one of the stages is the construction of a 3D model. This tends to be a digital 3D CAD model which is then adapted for use in different analysis disciplines, for example, engineering and neutronics. Following the conclusion of analysis, feedback can be provided and the design cycle can be iterated to refine and optimise the design. Automating the analysis can help to rapidly develop a design as shown in 1. While some automated analysis remains a challenge in certain disciplines, progress is being made on the automation of a number of domains used in the analysis of fusion reactors. In domains such as aerospace2, the automatic generation of simplified representative CAD geometries has given great benefit to the concept design process. Fusion reactor design processes involve analysis being carried out and fed back into the creation of the next CAD model; this process is usually a manual GUI based operation. The interpretation of results from analysis and geometric modifications are necessarily decoupled in this mode of operation, which slows the speed of iteration. The situation can occur where the models are updated several times as different analysis streams feedback into the design. In the case where different analysis tasks take different amounts of time, there may be situations where the analysis streams therefore get different versions of geometry, before it is further modified for their respective analysis. The scripts that users create to generate the CAD model are compatible with version control and therefore the method of creating the CAD can be version controlled and traced across a design process. Having this automated model creation for simple space reserving could be considered the first stage in creating a more efficient, automated, rapid and reproducible design cycle. Automated model creation can reduce the risk of geometry creation becoming a bottleneck in the design cycle. While complex model construction might be difficult to automate with the current software, there is utility in automating simpler models and allowing the analysis of specific geometry details to be filled in at a later stage. Additionally, there is also some utility in the use of automated CAD in conjunction with automated analysis at early stages in the design, where simple models are more appropriate.\n\nA key advantage of creating a 3D reactor geometry from parameters is that the produced model then becomes easy to quantify in terms of a small set of values. Being able to describe a 3D model with a series of parameters allows direct linking between an optimiser, input parameters and output metrics. The designer’s input is still required to make the parameterisation rules that allow components to be varied in ways that impact their performance. A designer’s skill is required to ensure the layout of components interface correctly and do not overlap. A benefit of the parametric model construction process is that when one parametric model is made this results in many perturbations that can be generated by scripts, while a static model remains a single static model. A disadvantage is that creating a parameterised model layout is more complex than a single model.\n\n\nApproach\n\nThe efficient use of CAD in the design process requires a well thought out and coherent approach to utilise all the potential benefits. While there are many possible approaches to the challenge, the Paramak focuses on an automated, parameter driven, permissive and open-source approach. This section aims to justify that decision.\n\nCadQuery 23 offers a potential solution for the creation of automated parametric CAD. CadQuery 2 is an open-source Python library that binds to OpenCascade (OCCT 7.5 release)4 and has some unique features among the possible open-source candidates. One such capability is the ability to search, filter and then operate on the CAD solids during construction. This allows components to be linked and built from each other without the user having to be concerned with redefining related solids when a linked solid is modified. This is already possible with proprietary CAD software but these capabilities are now emerging into the open-source area. CadQuery development is ongoing and the specific version used for this publication is 5.\n\nWhen using a complete CAD and analysis solution that includes a parametric CAD modelling package, the transfer of parametric models from the CAD engine to the analysis software can be achieved. Within the ecosystem of commercial PLM (Product Lifecycle Management) systems it is entirely possible to generate parametric CAD and use it within parametric analysis workflows. When wishing to utilise the CAD geometry in external analysis workflows that are not included within the ecosystem of proprietary software solutions, the use of parametric CAD becomes more challenging. The provision of CAD models via open formats such as STP format AP2146 do not support the encoding of parametric components within the CAD file. This lack of parametric support in the file formats that are used for transferring the model results in non parametric CAD being received at the analysis level for certain types of analysis. There are several possible solutions for this such as incorporating more analysis into the PLM software or developing new open CAD formats that support parametric components such as STP AP2427. The approach taken by the Paramak is to provide a parametric creation of non parametric STP files with a permissive licensing model. The combination of permissive licensing and parametric studies allows automated geometry creation and analysis to be carried out on potentially tens of thousands of designs in parallel. Cloud bursting together with Cloud computing can provide the computing resources for such a study. Traditional licensing models where the costs scale with number of parallel sessions can result in significant costs implications for such a spike in compute capacity. Depending on the number of parallel sessions required these licensing costs can become significant. Permissively licensed open-source software offers a solution to such a scenario. Since the Paramak is distributed with a permissive usage MIT open-source licence, it is therefore compatible with cases where parallel sessions are desired without incurring any licensing costs. As cloud computing grows both in popularity and market penetration, the licensing of software becomes an important factor in the software’s utility. This is reflected by the growing popularity of permissive open-source licensing8.\n\nThe source-code is under version control and openly available via Github9 under a permissive MIT licence. The Paramak Python package is distributed via PyPi10 and there are plans to incorporate a Conda distribution in the future. A containerised build environment is distributed via Dockerhub11 containing a pre-built environment with all the required dependencies. The code is documented with diagrams and examples on ReadTheDocs12 which makes use of extensive Docstrings within the code. The code has been internally reviewed by a Research Software Engineer internal to UKAEA and also undergoing a professionally reviewed by an external company PullRequest13. Continuous integration has been implemented using CircleCI14 to run a broad range of unit tests and integration tests. The test suite also covers use of the parametrically generated CAD in neutronics simulations using DAGMC15 and OpenMC16. This helps ensure the geometry made is suitable for use in neutronics analysis. Github Actions have been utilised from an early stage for automating several aspects of the code distribution, packaging and static code analysis. Github Actions have been used for employing code style guides (PEP8), updating the PyPi package distribution and automatically building and uploading new Dockerhub images with each new version of the code. The decision to open-source the Paramak code was a key enabler that allowed use of the previously mentioned platforms and in turn allowed the code to grow and improve rapidly. Additionally, the open-source nature of the project has facilitated contributions from outside the organisation, as demonstrated by the wide author list and contributors on to the Github repository.\n\n\nCode structure and examples\n\nThe Paramak consists of three main groups of classes: Shapes, Components and Reactors (see Figure 1).\n\nThe Parametric Shapes provide profiles from a combination of straight edges, circular edges and Bezier spline edges. These shapes can represent a wide range of basic shapes and are made from a series of 2D coordinates. Shapes can be operated on to create 3D volumes using extrude, revolve, sweep and rotate operations (see Figure 2). Boolean operations such as cut, intersect and union are also available to Shapes. To build Shapes the class must be provided with coordinates or points and edge connection types to connect each coordinate.\n\nThe Parametric Components inherit from Shape and build upon these basic families of shapes to create volumes that more closely resemble components found in fusion reactors. Parametric Components generally have particular methods of finding the coordinates that make up the shape and are thus able to provide the coordinates needed to make a Shape class. The methods of finding points differ from component to component and are encoded within the component’s class.\n\nFor the simplest Parametric Components such as CenterColumnCylinder() the points coordinates are found based on a hollow cylinder. This requires just four points and uses straight lines to connect the points followed by a rotation around the Z axis. This is then abstracted for the user so that only the height, inner radius and outer radius are required. The Component class then finds the points from the internal rules and applies any CAD operations or Boolean operations. More complicated shapes such as the BlanketFP() (see Figure 6) finds points on the front surfaces using a variable offset from the plasma. A variable thickness between the interior and exterior surface is then used to find the rear surface points. The front and rear surface points are connected with a series of splines with straight connections between the two surfaces. The variable offsets and thicknesses can be provided as a function of poloidal angle and the component is therefore able to construct more complex 3D objects. Some components (e.g. InnerFirstwallFCCS()) are constructed entirely from other components, in this case finding their coordinates is not necessary as a surface offset and Boolean cut is sufficient to find the 3D volume.\n\nThere are currently over 34 Parametric Components available (see Figure 3) and many additional shapes are planned. When these components are combined then the variety of 3D volumes available is sufficient to start constructing simple fusion reactors as shown in Figure 4.\n\nNote that because these shapes are all customisable with parameters they can appear differently to their default view pictured in the diagram. These inherit from the Shape class and have encoded methods of calculating the points required.\n\nNote that because these reactors are all customisable with parameters they can appear differently to their default view pictured in the diagram. From left to right and up to down the reactor class names are BallReactor(), SingleNullBallReactor(), SegmentedBallReactor(), SingleNullSubmersionTokamak(), SubmersionTokamak() and CenterColumnStudyReactor().\n\nParametric Reactors allow users to create a 3D reactor model by combining Parametric Components and Shapes with linkage that describes how they fit together. The models are not exact reproductions of any particular design but reflective of different reactor configurations that are available. There are currently six Parametric Reactors available in the Paramak (see Figure 4).\n\nTwo examples models created entirely from parameters are presented (see Figure 7 and Figure 8). In the case of the SegmentedBallReactor() the model has no inboard breeder zone and has divertors in the upper and lower positions. There are also single-null varieties of the BallReactor(). The main user inputs required are the radial thicknesses of components. The reactors require less user inputs than the individual components that make up the reactor would require. This is due to the radial build process that helps component inputs be derived from other components. The reactor design has the order of components encoded and therefore from this user information it is possible to know where each component starts and ends in the radial direction.\n\nThe vertical build for the SegmentedBallReactor() is largely based on the radial build which greatly minimises the number of user inputs required for a 3D model. The user inputs for the plasma elongation and triangularity, combined with the radial build parameters for the plasma, allow the coordinates of the top of the plasma to be calculated. The vertical offset from the firstwall to the plasma defaults to the same value as the outboard plasma gap radial thickness but can be specified independently using the plasma gap vertical thickness parameter. The blanket thickness is constant all around the reactor both in radial and vertical directions. The Parametric Component for the blanket accepts a variable thickness as a function of angle (see Figure 6) however this particular reactor design has been programmed to have constant thickness blankets throughout. This means the users will not be asked for a vertical blanket thickness but have less control over the reactor. The blanket is also segmented by another Parametric Component (BlanketStarCutter()) to create banana segments. CadQuery’s inbuilt filter methods are then used to select the front edges of the firstwall and breeder zone so that they can be filleted. A Boolean cut between the firstwall block and the breeder zone results in a wrap around design. Positioning of poloidal field coils is a user controllable argument, however if (R,Z) coordinates are not specified then they are equispaced vertically behind the blanket. Four types of toroidal field coils exist as Parametric Components: rectangle, coat hanger, Princeton-D and triple arc. However, simple rectangular toroidal field coils are used for the current BallReactor() design. The SegmentedBallReactor() inherits from the BallReactor() so it also uses rectangular magnets by default. However other magnets shapes are also avaialble as parametric components (see Figure 5). When inheriting from a base design it is possible to overwrite any of the components. Due to this system the number of variations on the base design can rapidly increase. The BallReactor() design has inbuilt assumptions regarding the connections and shapes of components, this has disadvantages in terms of the flexibility but also the advantage of having reduced inputs for the user to specify.\n\nAdditionally the blanket has a variable thickness and variable offset from the plasma.\n\nThe example also exports the SVG image used in this Figure and CAD files (STP) used when making Figure 4.\n\nThe example also exports the SVG image used in this Figure and CAD files (STP) used when making Figure 4.\n\nThe SubmersionTokamak() requires a few more inputs from the users and offers more flexibility when creating the models. Additional inputs are required for the radial thickness of the supports and the radial thickness of the inner blanket. The computational time to generate the 3D volumes and export CAD files in STP format once the input dimensions have been specified varies from around 20 seconds for a simple BallReactor() to around 40 seconds for a SegmentedBallReactor() on a desktop computer (i5 Intel processor). In this case the time difference is due to segmenting the blanket and filleting the edges of the blanket. Currently the entire construction process is a serial operation so there is scope to speed up the construction by parallelising parts of the construction process.\n\nThe CenterColumnStudy() reactor is designed for a specific use case. When study the impact of geometric parameters on the center column it is possible to simplify the design to only include components that significantly impact the simulation result. For example, the outboard TF and PF coils have little impact on the simulation results in this case. This reduces the time needed for model creation and reduces model initialisation in analysis use cases.\n\nWhile the existing Parametric Reactors are not a full representation of magnetic fusion reactors, the framework established can be used to create more detailed components with more complex relationships between components.\n\nAll the various reactor classes allow operations such as exporting the volume(s) to CAD files (STP and STL format) and 2D images (SVG) of the geometry as used in the documentation12. Other properties of the geometry can easily be obtained such as the volume of each shape or component in the reactor. This can be useful for cost estimates in systems codes or mass calculations in remote maintenance strategies. The utility of CAD models goes beyond visualisation and basic properties in assessing a design’s suitability and can be used as part of an automated parameter study. The Paramak knows the extent of the x, y, z dimensions for the geometry and therefore can automatically create thin shell bounding boxes (referred to as Graveyard volumes) for use in CAD based neutronics with DAGMC16. While this paper aims to focus on the geometry creation within the Paramak there are future papers planned where utilisation within neutronics and engineering workflows will be demonstrated.\n\n\nConclusion\n\nThe Paramak code has been introduced and the motivation for facilitating a data science approach to geometry construction has been discussed. Several benefits of the open-source approach have been realised during the project. The number of Parametric Components has grown to the level where simplified reactor models can be constructed. Reactor models can be encoded to encapsulate design decisions which allow the required user inputs to be limited. This is demonstrated by the three example models presented in the paper and reinforced by additional Parametric Reactor models contained in the documentation12. There are currently six different Parametric Reactors for users to create. Due to the structure of the code, it is straightforward to inherit existing reactors and modify specific parts of their design to extend the reactor family to accommodate additional features or parameters of Parametric Reactors.\n\nThe current parametric models provided in the Paramak are relatively simple but it is also possible to make more complex models when provided with a design.\n\nThe Paramak has been used within UKAEA to create models of several spherical tokamak configurations and has also been used to reproduce a SPARC like design based on the diagrams in 17. The outputs of the Paramak are CAD models which are useful in fusion analysis disciplines such as Finite Element Analysis, neutronics, visualisation and even cost models which often require CAD files as an input.\n\nDue to the use of modern software practices (continuous integration and containerisation), the software is able to test the CAD generated in neutronics analysis and demonstrate the compatibility of the geometry in use. The software employs modern software practices such as automatic documentation generation (ReadTheDocs)12, package distribution services (PyPi)10 and can be containerised11. Consequently, the learning time, installation time and time to first results are minimal.\n\nThe use of these models in automated workflows has yet to be demonstrated in a publication but this would be the next logical stage in the process and the authors plan to publish a range of use cases for the parametric geometry in the future. Future work will, amongst other improvements, incorporate detailed parametric blanket models which have previously been created1.\n\n\nSoftware availability\n\nParamak source code available from: https://github.com/ukaea/paramak\n\nArchived source code as at time of publication: http://doi.org/10.5281/zenodo.438426918\n\nLicense: MIT",
"appendix": "Acknowledgements\n\nThe authors would also like to thank Dr Lidija Pasuljevic Shimwell, Helen Gale, Stephanie Ellis, Colin Billingsley and Linda Billingsley for their support. The authors would like to thank Simon McIntosh for the provision of examples when calculating the coordinates of the TF coils. The authors would like to thank Lyal Avery and the rest of the team at PullRequest for initiating the software review. The authors would also like to thank all the CadQuery developers and in particular Adam Urbańczyk, Jeremy Wright and Dave Cowden.\n\n\nReferences\n\nShimwell J, Delaporte-Mathurin R, Jaboulay JC, et al.: Multiphysics analysis with cad-based parametric breeding blanket creation for rapid design iteration. Nuclear Fusion. 2019; 59(4): 046019. Publisher Full Text\n\nMukundakrishnan B, Rajmohan N, Rajnarayan DG, et al.: A Script-Based CAD System for Aerodynamic Design. 2019. Publisher Full Text\n\nUrba´nczyk A, Wright J, Cowden D, et al.: bsilvereagle, jwhevans, xix xeaon, Cadquery/cadquery 2.0.1, 2020.\n\nOpen CASCADE Technology 7.5.0 released. 2020; (accessed November 30, 2020). Reference Source\n\nAdam Urbanczyk DC, Wright J: CADQuery, A python parametric CAD scripting framework based on OCCT, Git SHA cf275b0. 2020; (accessed November 30, 2020). Reference Source\n\nISO 10303-214: 2010 Industrial automation systems and integration — Product data representation and exchange — Part 214: Application protocol: Core data for automotive mechanical design processes. 2020; (accessed Novem-ber 30, 2020). Reference Source\n\nISO 10303-242: 2014 Industrial automation systems and integration — Product data representation and exchange — Part 242: Application protocol: Managed model-based 3D engineering. 2020; (accessed November 30, 2020). Reference Source\n\nThe Complete Guide to Open Source Licenses 2020. 2020; (accessed November 30, 2020). Reference Source\n\nShimwell J, Billingsley J, Delaporte-Mathurin R, et al.: Paramak source code Github repository, v0.2.0, Git commit c41dc4c2e68183869556544ee7a72deb1d16a8dc. 2020; (accessed January 4, 2021). Reference Source\n\nPyPi - The Python Package Index: Paramak v0.2.0 PyPi distribution. 2020; (accessed January 4, 2021). Reference Source\n\nDockerhub, containerized distribution via Dockerhub of the Paramak. 2020; (accessed November 30, 2020). Reference Source\n\nReadTheDocs: Paramak API documentation and example usage,2020; (accessed November 30, 2020). Reference Source\n\nPullRequest: Code Review as a Service,2020; (accessed November 30, 2020). Reference Source\n\nCircleCI: CI pipeline for the Paramak,2020; (accessed November 30, 2020). Reference Source\n\nRomano PK, Horelik NE, Herman BR, et al.: Openmc: A state-of-the-art monte carlo code for research and development. 2014. Publisher Full Text\n\nWilson PP, Tautges TJ, Kraftcheck JA, et al.: Acceleration techniques for the direct use of cad-based geometry in fusion neutronics analysis. Fusion Engineering and Design. 2010; 85(10–12): 1759–1765. Publisher Full Text\n\nCreely AJ, Greenwald MJ, Ballinger SB, et al.: Overview of the sparc tokamak. J Plasma Phys. 2020; 86(5): 865860502. Publisher Full Text\n\nShimwell J, Billingsley J, Delaporte-Mathurin R, et al.: ukaea/paramak: v0.2.0 (Version 0.2.0). Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4384269"
}
|
[
{
"id": "77632",
"date": "16 Feb 2021",
"name": "Yuefeng Qiu",
"expertise": [
"Reviewer Expertise nuclear physics",
"fusion neutronics"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSection \"Integration in workflows\": this section discusses two topics, one is on the preservation of the parametrization in the STEP format, another one is on software licensing. The author is suggested to discuss them in a separate paragraph to provide a better understanding. The licensing issues are not particularly concerned by the reader, and the discussion on licensing for a open-source software is confusing.\n\nThe ordering of the figures needs to be improved so that the reader doesn't need to jump over pages. For example, Figures 5/6 are mentioned before figures 3/4.\n\nSection \"parametric reactors\": \"The Parametric Component for the blanket accepts a variable thickness ... however this particular reactor design has been programmed to have constant thickness blankets throughout.\" The reviewer is not clear about this statement, whether the blanket thickness can vary or not. Because the blanket is essential to the fusion reactor design, providing users the necessary freedom of modifying the blanket is a key capability for this program. Please provide some examples of how the blanket thickness can vary in Figure 6, using thickness = [150,70,70] as an example.\n\nFigure 4. the reviewer has seen the D shape of TF coil in Figure 1 but it hasn't been used for any of the six reactors in Figure 4. Please explain the reason.\n\nFigure 7. Since the users are responsible for providing all the offsets and constraints of those components, the reviewer is interested in how the program can handle design failures, such as geometry collisions. This could possibly occur in the vertical direction because the geometry relation in the vertical direction seems not provided.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "80436",
"date": "19 Mar 2021",
"name": "Jingang Liang",
"expertise": [
"Reviewer Expertise nuclear reactor physics",
"Monte Carlo particle transport",
"reactor benchmarks",
"reactor core design"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article describes an open-source python package Paramak which facilitates rapid production of 3D CAD models of fusion reactors and helps parametric studies on geometry configurations for reactor prototyping. This tool, built upon an open-source CadQuery library and integrated with fusion reactor domain knowledge, provides an efficient framework for automated geometry creation and analysis of fusion reactor designs. It will be of interest to a wide range of researchers in nuclear engineering fields. In addition, Paramak employs modern software practices and open-source principles in its development, making it friendly to both users and developers.\nThe manuscript is well-written. The methodology, implementation, and demonstrations are clear and reasonable. The source code of Paramak is available on GitHub, together with complete guides, examples, and other documentation. There is also an explanatory video that shows how to use the tool. The reviewer tried to install Paramak and tested some functionality. It worked well and as expected.\nI have no suggestions for improvement of the manuscript.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-27
|
https://f1000research.com/articles/9-1117/v1
|
10 Sep 20
|
{
"type": "Research Article",
"title": "Impact of sin taxes on consumption volumes of sweetened beverages and soft drinks in Saudi Arabia",
"authors": [
"Rania Megally",
"Ayoub Al-Jawaldeh",
"Ayoub Al-Jawaldeh"
],
"abstract": "Background: The consumption of sweetened drinks plays a major role in increasing the burden of diseases such as obesity and cardiovascular diseases. The prevalence rate of obesity and overweight individuals in Saudi Arabia has increased alarmingly among children over the past decade, revealing a crucial need the initiate and monitor effective prevention measures of obesity. Hence, this paper aims to measure the impact of sin taxes of sugar-sweetened beverages on the consumption level of such beverages in Saudi Arabia. Building enough evidence to guide other countries in the Eastern Mediterranean Region (EMR) on how to reduce the level of sugar intake consumption to reduce obesity and malnutrition has an impact on the economy as a whole. Methods: The excise taxes on sugar-sweetened beverages were applied in 2017. The impact of this was measured using a time series data set that covered sales volumes of soft drinks in Saudi Arabia from 2010 to 2020. Results: Sin taxes had a significant negative impact on the sales volume over the years. A significance level of <1% was found as sales volume of soft drinks decreased by 57.64% from 2010 to 2017 when sin taxes were applied on energy and soft drinks. Conclusions: Sin tax policy is recommended to other countries in the EMR to reduce obesity levels among children. These recommendations advocate the recommended priority actions by the World Health Organization for the strategy on nutrition for the EMR 2020-2030.",
"keywords": [
"Saudi Arabia",
"Health and Economic Development",
"Behavioral Economics",
"Health Behavior",
"Welfare Economics",
"Household Behavior and Family Economics",
"Development Planning and Policy",
"Economy wide Country Studies"
],
"content": "Introduction\n\nSugar-sweetened beverages are well-known to be related to adverse health outcomes as they are considered a significant source of excess caloric intake (Fidler Mis et al., 2017). The consumption of such sweetened drinks, especially those containing fructose, plays a major role not only in increasing the calories, but also in increasing the burden of diseases such as obesity and cardiovascular diseases. Multiple problems can occur in obese children, such as the high risk of having sleep disordered breathing (Paglia et al., 2019). Further, consuming large quantities of sugar is associated with adverse health outcomes such as attention deficit disorder and hyperactivity disorder (Del-Ponte et al., 2019; Johnson et al., 2011). Children are categorized as overactive when they cannot sit still, cannot concentrate, cannot keep silent, and leave one unfinished activity and move to the next (Bellisle, 2004).\n\nThe impact of sugar-sweetened beverages has been also measured on children’s cognitive level using the Kaufman Brief Intelligence Test. The results showed that as the consumption of sugar-sweetened beverages increases, the verbal scores of the mid-childhood Kaufman test decreases. On the contrary, there was a positive association between fruit consumption and higher cognitive scores (Cohen et al., 2018). These facts demonstrate the negative impact of sugar-sweetened beverage consumption on the cognition level of the children, showing that it is directly related to their productivity in schooling performance, which is eventually reflected in their productivity and income level in the future.\n\nThe increased risk of gaining weight caused by the sugar contained in many drinks, such as soft drinks and energy-rich drinks, has been proved in many longitudinal studies and clinical trials. It should be noted that there were no significant differences in gaining weight recorded when sucrose was replaced with low-calorie artificial sweeteners. However, some of the artificially sweetened drinks can increase the risk of diabetes and may lead to weight gain and obesity (Al-Jawaldeh et al., 2018).\n\nEastern Mediterranean Region (EMR) policymakers have been advised to measure the levels of current free-sugar intake in drinks and foods given the emphasized importance of reducing sugar intake in areas where there is malnutrition. Also, the adoption of measures and policies became a necessity given that 60% of our daily energy intake is constituted from carbohydrates, such as refined cereals and sugars (Al-Jawaldeh et al., 2018).\n\nThe high prevalence of obesity and overweight adults and children has been observed in the EMR, with the highest prevalence of regional diabetes rates worldwide. These facts spotlight the high rates of overweight children in the region as it has been observed in some countries that more than 15% of children are affected (World Health Statistics, 2016).\n\nSaudi Arabia is among the Gulf Cooperation Council (GCC) countries that has experienced a high prevalence rate of obesity and overweight; reaching 13.4% and 18.2% of overweight and obesity, respectively (Alqarni, 2016). The prevalence rate of obesity and overweight individuals has increased alarmingly among children in Saudi Arabia over the past decade, revealing a crucial need to initiate and monitor effective prevention measures of obesity (Al-Hussaini et al., 2019).\n\nMexico reached the highest consumption level of sugar-sweetened beverages worldwide in 2012 (Valadez, 2013), which was linked to the high prevalence rates of obesity and overweight individuals – 30% of children and 71% of adults (Barquera et al., 2013; Encuesta Nacional de Salud y Nutrición (2012). Evidence showed that 71% of the consumption of added sugars was derived from sugar-sweetened beverages (Sánchez-Pimienta et al., 2015). The Mexican government reacted by setting sugar-sweetened beverages taxes in order to reduce their consumption, leading to a reduction in the high rate of obesity/overweightness. In addition, studies showed that such taxes decreased the consumption of sugar-sweetened beverages in Mexico (Colchero et al., 2016; Pan American Health Organization, 2015).\n\n\nWorld Health Organization recommendations and fiscal policies\n\nResilient and sustainable food systems for healthy diets are one of the main nutrition strategies of the United Nations Action on Nutrition. The EMR of the World Health Organization (WHO) has developed an action plan and policy statement for sugar reduction based on the guidelines of the WHO, considering the energy intake per person that exceeded 2000 kcals/day in all regional countries (Alwan et al., 2017). Hence, the average intake of sugar should be decreased by more than 50% for adults and children (WHO, 2020).\n\nThe WHO recommends the use of well-designed subsidies and taxes in order to incentivize the consumption and production of healthier drinks and foods. One of the vital Eastern Mediterranean regional initiatives that have been developed to support the actions for obesity prevention 2019–2023 is the implementation of fiscal measures. These fiscal measures include the implementation of applying taxes on sugar-sweetened drinks, as well as other subsidies and taxes that promote healthier diets (WHO, 2019).\n\nThe objectives of this study were as follows:\n\n1. Provide an overview of the impact of interventions to discourage sugar intake and reduce the consumption level;\n\n2. Measure the impact of sin taxes on the sales and consumption level of sugar sweetened beverages in Saudi Arabia, one of the countries in the Mediterranean region who applied excise taxes on these products.\n\n\nMethods\n\nThis paper measured the impact of sin taxes on sugar-sweetened beverages using a time series data set that covered sales volumes of soft drinks in Saudi Arabia from 2010 to 2020. The data were secondary data collected by Global Company Intelligence (GCI), which is a company that specializes in collecting data from national governments and international industrial companies.\n\nThe authors requested GCI to create a report with the following variables concerning Saudi Arabia for the period 2010 to 2020: value of soft drinks in million dollars and local currency of Saudi Arabia per year, consumption of soft drinks in million liters per year, percentage growth from previous period to current period in million liters, and percentage growth from previous period to current (PP Growth %). The dataset created by GCI can be found in Underlying data. The impact of sin taxes has been tested via the following model:\n\n\n\nWhere SalesVolt refers to the sales volume in million liters and Pricet refers to the price of soft drinks in million US dollars.\n\nSTATA 16.0 was used to conduct descriptive statistics and data anlaysis. Subsequently autocorrelation of the sales volume trends over the years was tested. Finally, the impact of sin taxes on sales volumes was been tested via regression analysis after testing for normal distribution of the time series of both dependent and independent variables using the Shapiro-Wilk test.\n\n\nResults\n\nIn Saudi Arabia, the obesity rate has doubled over the past decade (Al-Hussaini et al., 2019). This lead to vital actions taken by Saudi policymakers, such as imposing 50% excise taxes on sweetened-soft beverages in 2017, as a response to the proposed policy priorities that have been recommended by the EMR- WHO to prevent diabetes and obesity in the region (Alwan et al., 2017).\n\nTable 1 shows the decline in percentage change of sugar-sweetened drink consumption and percentage change in sweetened juice consumption due to imposed excise taxes, from 2016 to 2019.\n\nFrom 2010 to 2017, sales volumes of soft drinks decreased by 57.64%; there was a decrease from 7694.6201 to 12129.507 million liters annually during the period when sin taxes have been applied to energy and soft drinks. Saudi Arabia started advocacy and communication campaigns before introducing sin taxes as part of their national action plan for obesity prevention, guided by the National Food Based Dietary Guidelines, which had key messages to social media, TV advertisements, and direct communication through schools (WHO, 2019). This led to slight reduction up to 2016; however, there was a sharp significant reduction after imposing sin taxes in 2017. Overall, an increasing trend in sales volume of soft drinks over the last decade has occurred (Figure 1a); however, the percentage change of the sales volume (PP growth) started to decrease sharply in 2017, the year that sin taxes were applied to the prices of soft drinks (Figure 1b).\n\n(a) Sales volume in million liters; (b) percentage growth from previous period (PP growth %).\n\nIn Table 2, autocorrelation shows that the time series has been divided into three lags indicating three stages of PP growth. The results show that the impact of sin taxes in reducing sales volume over time and the trend between the three lags is statistically significant (P<0.05).\n\nThe normal distribution of sales volume, the growth rate of sales volume, as well as the value of soft drinks have been tested before estimating a regression model using the Shapiro Wilk test. H0 assumes a normal distribution of the variables. Table 3 shows that the time series of all variables are normally distributed, which qualifies them to be used in the regression models.\n\nThe results show that sin taxes have a significant negative impact on the percentage volume growth over the years with a significance level <1% and high R2 of around 68%, which reflects that the model estimates the impact of sin taxes at 68% (Table 4).\n\n\nDiscussions\n\nThe prevalence rate of obesity and overweightness has increased alarmingly among children in Saudi Arabia over the past decade, leading to a crucial need for intervention (Al-Hussaini et al., 2019). Evidence shows that there is a positive relationship between the consumption of sugar-sweetened beverages and the prevalence rate of obesity and overweightness (Paglia et al., 2019). Hence, EMR of the WHO recommended governments to introduce a sustainable reduction of sugar intake over the coming 3–4 years. Considerable reductions of sugar intake should be 50% or more to terminate the increase in obesity and diabetes to decrease the burden of premature deaths, as non-communicable diseases are expected to reach 25% by 2025 (WHO, 2020). These facts were alarming enough for the WHO to set their action plans for the nutrition strategy 2020–2030 (WHO, 2019), which recommended the implementation of taxes on sugar-sweetened beverages to reduce obesity rates in various nations.\n\nThese facts were inspiring enough to study the impact of sin taxes that have been applied to sugar-sweetened beverages in one EMR country, Saudi Arabia, in 2017. The current study is consistent with the regional WHO recommendations to apply the United Nations Political Declaration on Non-Communicable Diseases, the priority legal interventions that target the prevention of non-communicable diseases in the EMR (Gostin et al., 2017), as well as the recommendations of the Commission on Ending Obesity (Food and Agriculture Organization of the United Nations/World Health Organization, 2017; Report of the Commission on Ending Childhood Obesity, 2016).\n\nThe results showed that the rate of change in the sales volume over the last decade in Saudi Arabia started to decrease sharply in 2017, the year the sin taxes have been applied to the prices of soft drinks. Sales volumes from 2017 were increasing but at decreasing rates, and the sin tax had a significant negative impact on the change of sales volumes over the past 10 years. These results are in line with the nutrition strategies of the United Nations of Action on Nutrition that was based on evidence and experimental studies, which observed effective fiscal measures of taxes and subsidies in shifting habits of purchases and promotion of dietary change (Thow et al., 2014; WHO, 2015a; WHO, 2016). This is expected to decrease the obesity levels among Saudi children over the coming years. Improvements in children’s health are expected to be reflected in better cognition, intelligence, and schooling performance of the children. Moreover, this study showed a decrease of around 57% of the percentage change of sales volume, which aligns with previous evidence from countries that applied taxes and reduced the purchases of sugar-sweetened beverages in a range of 20–50% (Colchero et al., 2016; Ells et al., 2015; Mozaffarian et al., 2012; Powell et al., 2013; Thow et al., 2014; WHO, 2015b; WHO, 2016).\n\nLimitations of this research were not knowing the nutrition status of the children after imposing the sin taxes, and limited analysis using simple linear regression that does not consider any confounding factors.\n\n\nConclusions and recommendations\n\nRecently, the WHO shed light on the importance of reducing obesity among children in the EMR given the high rate that has been observed lately. One intervention that was suggested is to impose taxes to sugar-sweetened beverages (Lobstein, 2014). The impact of applying such an intervention has been previously shown to have a positive impact in reducing the level of consumption of sugar-sweetened beverages in GCC countries. Hence, such a policy is recommended to extend to cover other countries in the EMR. In addition, public health education should be considered using social marketing campaigns and restriction of media advertisements about sugar-sweetened beverages, which should support the imposition of sin taxes, as recommended by Lobstein (2014). These recommendations advocate the recommended priority actions by the WHO for the strategy on nutrition for the EMR 2020-2030 (WHO, 2019).\n\n\nData availability\n\nHarvard Dataverse: Soft Drinks Volumes, https://doi.org/10.7910/DVN/9SE4V3 (Megally, 2020).\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).",
"appendix": "References\n\nAlwan A, McColl K, Al-Jawaldeh A, et al.: Proposed Policy Priorities for Preventing Obesity and Diabetes in The Eastern Mediterranean Region. World Health Organization Regional Office for Eastern Mediterranean. 2017. Reference Source\n\nAl-Hussaini A, Bashir MS, Khormi M, et al.: Overweight and obesity among Saudi children and adolescents: Where do we stand today? Saudi J Gastroenterol. 2019; 25(4): 229–235. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAl-Jawaldeh A, El-Mallah C, Obeid O: Regional Policies on Sugar Intake Reduction at Population Levels to Address Obesity in The Eastern Mediterranean. JSM Nutrional Disorders. 2018; 2(1): 1006. Reference Source\n\nAlqarni SSM: A Review of Prevalence of Obesity in Saudi Arabia. Journal of Obesity and Eating Disorders. 2016; 2(2: 25): 1–6. Publisher Full Text\n\nBarquera S, Campos-Nonato I, Hernádez-Barrera L, et al.: [Prevalence of obesity in Mexican adults 2000-2012]. Salud Publica Mex. 2013; 55 Suppl 2: S151–60. PubMed Abstract\n\nBellisle F: Effects of diet on behaviour and cognition in children. Br J Nutr. 2004; 92 Suppl 2: S227–S232. PubMed Abstract | Publisher Full Text\n\nCohen JFW, Rifas-Shiman SL, Young J, et al.: Associations of Prenatal and Child Sugar Intake With Child Cognition. Am J Prev Med. 2018; 54(6): 727–735. PubMed Abstract | Publisher Full Text | Free Full Text\n\nColchero MA, Popkin BM, Rivera JA, et al.: Beverage Purchases from Stores in Mexico under The Excise Tax on Sugar Sweetened Beverages: Observational Study. BMJ. 2016; 352: h6704. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDel-Ponte B, Anselmi L, Assuncao MCF, et al.: Sugar consumption and attention-deficit/hyperactivity disorder (ADHD): A birth cohort study. J Affect Disord. 2019; 243: 290–296. PubMed Abstract | Publisher Full Text | Free Full Text\n\nElls LJ, Roberts K, McGowan VJ, et al.: Sugar Reduction: The Evidence for Action. Annexe 2: Review of Behaviour Changes Resulting from Experimental Studies of Fiscal Methods. A Mixed Method Review of Behaviour Changes Resulting from Experimental Studies that Examine The Effect of Fiscal Measures Targeted at High Sugar Food and Non-Alcoholic Drink. London: Public Health England; 2015. Reference Source\n\nEncuesta Nacional de Salud y Nutrición: Estado de Nutrición, Anemia, Seguridad Alimentaria en La Población Mexicana. Instituto Nacional de Salud Pública, Mexico, 2012. accessed 29 March 2016. Reference Source\n\nFidler Mis N, Braegger C, Bronsky J, et al.: Sugar in Infants, Children and Adolescents: A Position Paper of the European Society for Paediatric Gastroenterology, Hepatology and Nutrition Committee on Nutrition. J Pediatr Gastroenterol Nutr. 2017; 65(6): 681–696. PubMed Abstract | Publisher Full Text\n\nFood and Agriculture Organization of the United Nations/World Health Organization: Second International Conference on Nutrition, Rome, 19– 21 November 2014. Conference Outcome Document: Framework for Action. 2017. accessed 1 February 2017. Reference Source\n\nGostin LO, Abou-Taleb H, Roache SA, et al.: Legal priorities for prevention of non-communicable diseases: innovations from WHO's Eastern Mediterranean region. Public Health. 2017; 144: 4–12. PubMed Abstract | Publisher Full Text\n\nJohnson RJ, Gold MS, Johnson DR, et al.: Attention-deficit/hyperactivity disorder: is it time to reappraise the role of sugar consumption? Postgrad Med. 2011; 123(5): 39–49. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLobstein T: Reducing Consumption of Sugar-Sweetened Beverages to Reduce The Risk of Childhood Overweight and Obesity. 2014. Reference Source\n\nMegally R: Soft Drinks Volumes. Harvard Dataverse, V1. 2020. http://www.doi.org/10.7910/DVN/9SE4V3\n\nMozaffarian D, Afshin A, Benowitz NL, et al.: Population approaches to improve diet, physical activity, and smoking habits: a scientific statement from the American Heart Association. Circulation. 2012; 126(12): 1514–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaglia L, Friuli S, Colombo S, et al.: The Effect of Added Sugars on Children’s Health Outcomes: Obesity, Obstructive Sleep Apnea Syndrome (OSAS), Attention-Deficit/Hyperactivity Disorder (ADHD) and Chronic Diseases. Eur J Paediatr Dent. 2019; 20(2): 127–132. PubMed Abstract | Publisher Full Text\n\nPan American Health Organization: Taxes on Sugar-Sweetened Beverages as A Public Health Strategy: The Experience of Mexico. Mexico Representative Office, Mexico, D.F., 2015. accessed 26 March 2016. Reference Source\n\nPowell LM, Chriqui JF, Khan T, et al.: Assessing the Potential Effectiveness of Food and Beverage Taxes and Subsidies for Improving Public Health: A Systematic Review of Prices, Demand and Body Weight Outcomes. Obes Rev. 2013; 14(2): 110–28. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReport of the Commission on Ending Childhood Obesity. Geneva: World Health Organization. 2016. Reference Source\n\nSánchez-Pimienta T, Batis C, Lutter CK, et al.: Main Sources of Total and Added Sugars Intake in The Mexican Population. 16 Congreso de Investigación en Salud Pública; Instituto Nacional de Salud Pública, Cuernavaca, Mexico. 2015.\n\nThow AM, Downs S, Jan S: A Systematic Review of The Effectiveness of Food Taxes and Subsidies to Improve Diets: Understanding the Recent Evidence. Nutr Rev. 2014; 72(9): 551–65. PubMed Abstract | Publisher Full Text\n\nValadez B: Desplaza México a EU en Consumos de Refrescos De Cola, 2012. Milenio. (web document in Spanish only). 2013. Reference Source\n\nWorld Health Organization: Using Price Policies to Promote Healthier Diets. Copenhagen: World Health Organization Regional Office for Europe; 2015a. Reference Source\n\nWorld Health Organization: Taxes on Sugar-Sweetened Beverages as A Public Health Strategy: The Experience of Mexico. Washington DC :Pan American Health Organization; 2015b. Reference Source\n\nWorld Health Statistics: Monitoring Health for The SDGs. Geneva: World Health Organization; 2016. Reference Source\n\nWorld Health Organization: Fiscal Policies for Diet and The Prevention of Noncommunicable Diseases. Technical Meeting Report 5-6 May 2015, Geneva, Switzerland. Geneva: World Health Organization Regional Office for Europe; 2016. Reference Source\n\nWorld Health Organization: Strategy on Nutrition for the Eastern Mediterranean Region 2020-2030. Cairo: WHO Regional Officer for the Eastern Mediterranean. Licence: CC BY-NC-SA 3.0 IGO. 2019. Reference Source\n\nWorld Health Organization: Policy Statement and Recommended Actions for Lowering Sugar Intake and Reducing Prevalence of Type 2 Diabetes and Obesity in The Eastern Mediterranean Region. 2020. Reference Source"
}
|
[
{
"id": "71182",
"date": "21 Sep 2020",
"name": "Maha Hoteit",
"expertise": [
"Reviewer Expertise Public Health Nutrition",
"Food composition tables",
"food microbiology",
"autism and nutrition",
"community nutrition"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSin tax is a tool that contribute to enhancing consumer behavior and if implemented can have a positive health impact on the population. In general, the decisions taken at the governmental level in terms of the introduction of new taxes, should be supported by reliable analyses and this paper is a good evidence to show the improvement of the countries to reduce the intake of sugar through the decrease in consumption of soft drinks that is responsible of aggravating the prevalence of obesity in EMR. The results of this study highlighted the sharp decrease in 2017, the year when the sin taxes have been applied to the prices of soft drinks and the sin tax had a significant negative impact on the change of sales volumes over the past 10 years. This manuscript is well written that reflect background literature on soft drink consumption in Saudi Arabia and highlight the importance of implementing sin taxes in all GCC.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6143",
"date": "26 Nov 2020",
"name": "Rania Megally",
"role": "Author Response",
"response": "Dear Dr. Maha, I just do like to thank you for your sincere review and comments that we the authors do highly appreciate Many thanks, Best regards, Authors"
}
]
},
{
"id": "71181",
"date": "14 Oct 2020",
"name": "Ali Arabi",
"expertise": [
"Reviewer Expertise gastroenterology and nutrition"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe introduction:\nThere are many repeated phrases concerning obesity and overweight - the authors need to shorten.\n\nAn introduction should include other problems like caries and enamel erosion, short sleep duration, hyperactivity, increased blood pressure and non-alcoholic fatty liver disease. High sugar content and low pH render SSB is a real threat to the developing dentition.\nMethodology and results are okay.\n\nIn recommendations the authors should also include:\nAdvertising of SSBs to children should be banned.\n\nThe introduction of compulsory information on the front of pack labels of SSB on health risks should be considered.\n\nAt School:\nEducational and behaviour-changing programs on appropriate choices of beverages and encouraging water consumption should be implemented in pre-schools and schools.\n\nPre-schools and schools should offer unlimited access to drinking water, for example, by water fountains, whereas SSBs and other sweet beverages should not be made available.\n\nParents:\nParents should avoid the use of SSBs to please their children.\n\nSSBs should not be easily available at home.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6163",
"date": "04 Dec 2020",
"name": "Rania Megally",
"role": "Author Response",
"response": "Dear Dr. Ali, First of all, I do like to express my gratitude for your thorough review of our paper that was of a great benefit as we have considered all your comments in the updated version of the paper Thank you again for your sincere efforts Best regards, Rania"
}
]
}
] | 1
|
https://f1000research.com/articles/9-1117
|
https://f1000research.com/articles/9-90/v1
|
07 Feb 20
|
{
"type": "Research Article",
"title": "Crop production under nitrogen starvation conditions: relationships with applied organic matter and soil microbial biomass",
"authors": [
"Masato Oda",
"Uchada Sukchan",
"Uchada Sukchan"
],
"abstract": "Background: The application of organic matter with a high C/N ratio is effective for the prevention of soil degradation, although this can cause nitrogen starvation. However, some fields are highly productive under nitrogen-starvation conditions. The underlying mechanisms for this are unclear but the correlation between soil microbial biomass (SMB) and crop yield suggests that nitrogen flows from SMB to crops. We aimed to clarify this flow and the source of nitrogen. Methods: We achieved nitrogen starvation conditions by applying waste mushroom bed and repeated lettuce cropping with different crop management practices, such as watering and fertilizer application. We analyzed correlations among crop yield, SMB, and total nitrogen. Results: The order of the lettuce yield stably corresponded with the management practice used. The SMB increased remarkably by the time of the second lettuce cropping and showed a strong correlation with crop yield. The nitrogen from the waste mushroom bed was lost by denitrification. The rate of decomposition showed no correlation with yield or SMB. Conclusions/Discussions: The crop yield corresponded with the management practice earlier than SMB. Namely, no nitrogen flow from SMB affected the crops. Furthermore, most applied nitrogen was denitrified, so the nitrogen flows of applied organic matter, SMB, and crops are independent. Therefore, the nitrogen source of both SMB and crops is biological fixation. The correlation between SMB and crop yield is not a causal relationship. The nitrogen source for both is nitrogen fixation. The application of organic matter enhances this by occurring nitrogen starvation but not providing a nitrogen source.",
"keywords": [
"Organic farming",
"Sandy loam",
"Tropics",
"Northeast Thailand",
"Taguchi method"
],
"content": "Introduction\n\nThe application of synthetic nitrogen is a key technology used in agriculture; however, it decreases soil microbial biomass (SMB) and results in soil degradation (Mulvaney et al., 2009). Conversely, the application of organic matter with a high C/N ratio increases SMB and helps degraded soil to recover (Chaudhary et al., 2009; Goyal et al., 1999). The problem that remains is nitrogen starvation (van Iterson, 1904). However, some fields are highly productive under conditions of nitrogen starvation (Nakatsuka et al., 2016; Oda et al., 2014).\n\nThe possible mechanism underlying this high productivity under nitrogen starvation conditions is the large flow of low concentrations of nitrogen. Possible sources of this nitrogen include decomposed organic matter, turnover of SMB, and microorganisms in the root zone (Geisseler et al., 2010). Previous studies have mentioned a correlation between SMB and crop yield (Entry et al., 1996; He et al., 1997); however, these studies also found a correlation between soil total nitrogen (TN) and crop yield at the same extent. Although there is also a study that did not find a similar correlation (Holt & Mayer, 1998). Logically, the correlation between SMB and crop yield occurs in one of three ways: 1) both SMB and crop yield are related to other factors such as TN; 2) SMB is related to crop yield; or 3) crop yield is related to SMB.\n\nIn this study, we created nitrogen starvation conditions by applying organic matter with a high C/N ratio (waste mushroom bed) to fields and performed repeated lettuce cropping with different crop management practices. We analyzed correlations among crop yield, SMB, and TN to clarify the flow and source of nitrogen.\n\n\nMethods\n\nThis study was conducted at a site with a tropical savanna climate in a lateritic loamy sand field in Tha Phara village, Khon Kaen Province, Thailand (16° 34′ N, 102° 83′ E), during 2011 and 2012. The site had ideal conditions for this study because sandy soils at a high temperature have a low level of soil organic matter and high microbial activity. Conducting the study during the dry season avoided any effects from rainwater.\n\nWe chose waste mushroom bed (gifted by a local farmer) as an organic material because of its adequate C/N ratio (40) for producing nitrogen starvation and its homogeneity. We applied the material to the fields at a fresh-weight rate of 1 kg m−2 (equivalent to a C application rate of 300 g m−2) and mixed it well with the topsoil (about 14 cm) at each time of planting. This ensured that the nutritional input was the same for all plots. We generated a gradient of SMB and crop yield by altering crop management practices, including watering (1.5 mm twice per day, 3 mm once per week, or none); urea application (at a rate of 10, 0.1, or 0 g m−2, the 0.1 g m−2 application performed as a solution in 1.0 l of water); waste mushroom bed application method (incorporated, applied to the soil surface, or incorporated after killing the fungi by packing the material in plastic mulch film and exposing it to sunlight for 1 d); and planting density (standard, double, or none). No other materials were used.\n\nThe four factors of the three levels of practice (Table 1) were assigned to an L18 orthogonal array (Taguchi, 1986). Such a design enables each effect to be evaluated at an accuracy of six replications. The plot (3 × 3 m) locations were first randomized then fixed.\n\nThe practice of plant density 0 was used to verify the independence of SMB from crop growth in the experiment, because an increase in the size of the root zone is associated with an increase in microbial activity (Alam et al., 2014). The yield of plant density 0 reflects no effect of any practice on the yield, but the problem is evenly allocated to all factors. It should be noted that the aim of the present study was not to evaluate the effects of each factor.\n\nSequential cropping is thought to be an essential condition for achieving high SMB (Oda et al., 2014). Therefore, we planted water spinach during the rainy season (seeded Aug 25, harvested Sep 26), then planted lettuce during the dry season (transplanted and harvested on Oct 20 and Dec 7; Dec 8 and Jan 17; and Jan 23 and Mar 1, respectively). Total precipitation for the first to fourth crop seasons was 248, 7, 0, and 0 mm, respectively. The plants were free from disease and insect pests; no plant protection procedures were used. The field was kept free of weeds by hand weeding.\n\nWe harvested the whole crop of lettuce and immediately oven-dried and weighed it to obtain the dry weight. Topsoil (to a depth of about 14 cm, bulk density 1.22) was sampled from each plot just after the crops were harvested. A composite sample from ten sampling points was collected from each plot. Each soil sample was sieved through a 2-mm sieve (the mushroom waste could pass through it) while moist, and 500 g of each sample was stored at 2°C until the SMB-N content was measured. The SMB-N content was measured using the fumigation–extraction method (Amato & Ladd, 1988). The inorganic nitrogen concentration of each sample was determined by extracting the sample with 2 M KCl and performing NH4+ and NO3− assays on the extract (Keeney & Nelson, 1982). The remaining portion of each sample was air-dried, and the total nitrogen and total carbon content of the soil was determined using an NC analyzer (SUMIGRAPH NC 200F; Sumitomo Chemical, Tokyo, Japan) using the dry combustion method.\n\nWe analyzed simple correlations (Pearson product-moment) among the crop yield, SMB, and TN using the mean values of the practices (excluding the values of plant density = 0). To do this, we used the CORREL function in Microsoft Excel 2016.\n\n\nResults\n\nWe achieved nitrogen-starved soil by applying waste mushroom bed followed by repeated lettuce cropping under different crop management practices, then analyzed correlations among crop yield, SMB, and TN.\n\nThe NO3-N content of the soil during the lettuce crop season was very low (2.3–3.1 μg g−1) compared with the threshold of fertilizer application used for conventional cultivation (20 μg g−1) (Breschini & Hartz, 2002; Fox et al., 1989). Table 2 shows the changes in soil properties. The maximum lettuce yield (45 g DM m–2) was higher than the average obtained through conventional farming in Thailand (33 g DM m–2, Department of Agricultural Extension 1996–2001; calculated as 4.1% DM, Food composition table ver. 7). No correlation between soil NO3-N content and yield was found. The order of the yield corresponded with the practices and resulted in correlations among the dry seasons in terms of yield (Figure 1).\n\nn=6\n\na n=4, b Soil microbial biomass nitrogen, c initial value, d dry season (water spinach).\n\nFresh waste mushroom bed was applied to every crop. Lettuce was grown in the dry season.\n\n◯Season 1, ●Season 2, △Season 3, ▲Season 4 Mean values (n=4) for different crop management practices are shown. Season 1 was during the rainy season (water spinach) and the others were during the dry season (lettuce). W: Watering (0: none, 1/w: once per week, 2/d: twice per day); M: Material position (S: surface, A: incorporated, D: incorporated following disinfection by the sun); P: Plant density (1: standard, 2: double); N: Nitrogen application (0: none, 0.1: 0.1g m-2, 10: 10g m–2).\n\nThe SMB increased remarkably during seasons 3 and 4 (Figure 2). The maximum SMB-N content (424 μg g−1) was an order of magnitude larger than that found in previous studies (Entry et al., 1996; He et al., 1997; Holt & Mayer, 1998). The SMB was lower in season 4 than season 3. The low soil moisture content during the later dry season affected the SMB because of the large effect of watering practice (Table 2). The SMB changed largely according to the practice used to input the same quantity of material. SMB showed a strong correlation with crop yield. The correlation (r = 0.977, p < 0.01) with SMB was stronger than that with TN (r = 0.588, p = 0.06). The SMB was independent of the crop yield because the SMB-N of plant density = 0 was approximately the average for that of the practices (Figure 3).\n\n◯Season 1, ●Season 2, △Season 3, ▲Season 4 Mean values (n=4) for different crop management practices are shown. Season 1 was during the rainy season (water spinach) and the others were during the dry season (lettuce). SMB-N: soil microbial biomass – nitrogen; TN: total nitrogen.\n\nMean values for different crop management practices for seasons 3 and 4 (n=8). W: Watering (0: none, 1/w: once per week, 2/d: twice per day); M: Material position (S: surface, A: incorporated, D: incorporated following disinfection by the sun); P: Plant density (0: none, 1: standard, 2: double); N: Nitrogen application (0: none, 0.1: 0.1g m-2, 10: 10g m–2); SMB-N: soil microbial biomass – nitrogen.\n\nThe waste mushroom bed applied decomposed within the crop season (Table 2). This rapid decomposition of organic material in the sandy soil of the tropics is consistent with the findings of a previous study (Chivenge et al., 2011). The amount of nitrogen applied was 7.5 g m–2 crop–1 (equivalent to 10% of the initial topsoil). The 0.6 g (average DM 24 g * 0.023; Food composition table ver. 7) of nitrogen removal by the crop was negligible compared with the input of nitrogen. There was no leaching because there was no rainfall. Finally, the same amount of nitrogen input was lost to the air by denitrification. We used the 3.1 of the K factor (Amato & Ladd, 1988) multiplied for the calculation of SMB-N and non-fumigation nitrogen; however, that overestimated the nitrogen because the total was larger than TN. The K factor does not affect the conclusion from the correlation analysis.\n\nThe rate of decomposition differed among the different practices. The rate was slow when there was no fertilizer, when surface application, and no watering (Table 3). The application of even a small amount of nitrogen remarkably enhanced the rate of decomposition. There was a very weak correlation between the release (decrease in TN) of nitrogen during the latter crop period and yield (r = 0.168) or SMB (r = 0.163).\n\nn=6. Change in soil TN in season 4 from the harvesting time of season 3 on days 25 and 38. Waste mushroom bed (89 μg N g soil−1) was applied on day 0. First half: day 25 – 0, Latter half: day 38 – 25. TN: total nitrogen.\n\n\nDiscussion\n\nA conventional yield was achieved under conditions of nitrogen starvation. The order of the yield stably corresponded with the management practice used. Later, SMB increased remarkably; the quantity corresponded with the practice in the similar order as the crop yield. As a result, SMB showed a strong correlation with crop yield. The applied nitrogen contained in the waste mushroom beds was lost via denitrification. The rate of decomposition showed no correlation with yield or SMB.\n\nWe achieved a conventional yield with a very low NO3-N level. This is inconsistent with nitrogen stock. The same quantity of material was applied each time and the nitrogen equivalent to the applied amount was lost with every crop. In addition, the difference in released nitrogen did not affect either crop yield or SMB. These findings led to the conclusion that the stock-based nutrition balance is no longer effective under conditions of nitrogen starvation. On the other hand, crop management practices made a remarkable difference.\n\nA robust correlation was seen between yield and SMB. However, this was the same response as seen with the different practices. The response was seen in yield earlier than SMB. Therefore, it is unlikely that the nitrogen flow occurs from SMB to crops.\n\nThe reason why the SMB suddenly increased from season 3 is unclear. This may have been affected by changes in the soil, such as structural changes due to aggregation, changes in microbial flora, or the accumulation of substances. In the present work, non-fumigation nitrogen increased monotonously.\n\nApplied nitrogen was lost by denitrification, but SMB was still seen to increase considerably. This means that these phenomena are independent of one another. SMB obtains nitrogen via biological fixation. Crops also obtain nitrogen by biological fixation. The application of organic matter with a high C/N ratio enhances the biological nitrogen fixation in the crop root zone by occurring nitrogen starvation but not providing a nitrogen source.\n\n\nConclusions\n\nWe examined nitrogen flow under conditions of nitrogen starvation through correlation analyses among crops, SMB, and TN. The correlation between SMB and crop yield showed no causal relationship. The nitrogen source was nitrogen fixation. The application of organic matter enhanced nitrogen fixation by occurring nitrogen starvation but not acting as a nitrogen source. Crop management practices largely affect the crop yield. The effect of applying a small amount of nitrogen should be studied to investigate how it can enhance microbial activity.\n\n\nData availability\n\nFigshare: SMB&Yield.xlsb, https://doi.org/10.6084/m9.figshare.11760963.v1 (Oda & Sukchan, 2020).\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).",
"appendix": "Acknowledgements\n\nThe authors sincerely thank Srisuda Thippayarugs for performing the SMB-N analyses. The authors would also like to thank Dr. Yasukazu Hosen for his helpful comments.\n\n\nReferences\n\nAlam MZ, Braun G, Norrie J, et al.: Ascophyllum extract application can promote plant growth and root yield in carrot associated with increased root-zone soil microbial activity. Can J Plant Sci. 2014; 94(2): 337–348. Publisher Full Text\n\nAmato M, Ladd JN: Assay for microbial biomass based on ninhydrin-reactive nitrogen in extracts of fumigated soils. Soil Biol Biochem. 1988; 20(1): 107–114. Publisher Full Text\n\nBreschini SJ, Hartz TK: Presidedress Soil Nitrate Testing Reduces Nitrogen Fertilizer Use and Nitrate Leaching Hazard in Lettuce Production. HortScience. 2002; 37(7): 1061–1064. [Accessed: 23 January 2014] Publisher Full Text\n\nChaudhary VB, Bowker MA, O’Dell TE, et al.: Untangling the biological contributions to soil stability in semiarid shrublands. Ecol Appl. 2009; 19(1): 110–122. PubMed Abstract | Publisher Full Text\n\nChivenge P, Vanlauwe B, Six J: Does the combined application of organic and mineral nutrient sources influence maize productivity? A meta-analysis. Plant Soil. 2011; 342(1–2): 1–30. Publisher Full Text\n\nEntry IA, Mitchell CC, Backman CB: Influence of management practices on soil organic matter, microbial biomass and cotton yield in Alabama’s “Old Rotation”. Biol Fertil Soils. 1996; 23(4): 353–358. Publisher Full Text\n\nFox RH, Roth GW, Iversen KV, et al.: Soil and tissue nitrate tests compared for predicting soil nitrogen availability to corn. Agronomy Journal. 1989; 81(6): 971–974. Publisher Full Text\n\nGeisseler D, Horwath WR, Joergensen RG, et al.: Pathways of nitrogen utilization by soil microorganisms - A review. Soil Biol Biochem. 2010; 42(12): 2058–2067. Publisher Full Text\n\nGoyal S, Chander K, Mundra MC, et al.: Influence of inorganic fertilizers and organic amendments on soil organic matter and soil microbial properties under tropical conditions. Biol Fertil Soils. 1999; 29(2): 196–200. Publisher Full Text\n\nHe Z, Yao H, Chen G, et al.: Relationship of crop yield to microbial biomass in highly-weathered soils of China. Plant Nutrition for Sustainable Food Production and Environment. 1997; 78: 751–752. [Accessed: 8 July 2013]. Publisher Full Text\n\nHolt JA, Mayer RJ: Changes in microbial biomass and protease activities of soil associated with long-term sugar cane monoculture. Biol Fertil Soils. 1998; 27(2): 127–131. Publisher Full Text\n\nvan Iterson G: Die Zersetzung vo Celulose durch aerobe Mikroorganimen. Zbl Bakt. 1904; 11(2): 689.\n\nKeeney DR, Nelson DW: Nitrogen—inorganic forms. 1982. Reference Source\n\nMulvaney RL, Khan SA, Ellsworth TR: Synthetic nitrogen fertilizers deplete soil nitrogen: a global dilemma for sustainable cereal production. J Environ Qual. 2009; 38(6): 2295–2314. PubMed Abstract | Publisher Full Text\n\nNakatsuka H, Oda M, Hayashi Y, et al.: Effects of fresh spent mushroom substrate of Pleurotus ostreatus on soil micromorphology in Brazil. Geoderma. 2016; 269: 54–60. Publisher Full Text\n\nOda M, Sukchan U: SMB&Yield.xlsb. figshare. Dataset. 2020. http://www.doi.org/10.6084/m9.figshare.11760963.v1\n\nOda M, Tamura K, Nakatsuka H, et al.: Application of high carbon:nitrogen material enhanced the formation of the soil A horizon and nitrogen fixation in a tropical agricultural field. Anthropological Science. 2014; 05(12): 1172–1181. Publisher Full Text\n\nTaguchi G: Introduction to quality engineering: designing quality into products and processes. 1986; [Accessed: 18 March 2014]. Reference Source"
}
|
[
{
"id": "69290",
"date": "08 Sep 2020",
"name": "Jianwei Li",
"expertise": [
"Reviewer Expertise Soil biogeochemistry"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript lacks solid foundation in describing soil nitrogen cycling in general and some of the key nitrogen processes were either ignored or just mentioned without qualification in the studies system. With organic amendments, many biogeochemical processes relevant to nitrogen cycling were not well described. Soil quality is usually improved with organic matter amendments and it may favor nutrient uptake, nitrogen mineralization, diffusion with improved water condition; also nitrogen use efficiency may also increase, that is, crop yield may be higher per unit of nitrogen taken up by the root. I think the current work is too superficial with very poor foundation and justification though it intends to target a very important research question. At the present form, I don't recommend it for indexing.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "5938",
"date": "15 Sep 2020",
"name": "Masato Oda",
"role": "Author Response",
"response": "Thank you for considering our manuscript. I guess that you were surprised to see the manuscript.\"Soil quality is usually improved with organic matter amendments and it may favor nutrient uptake, nitrogen mineralization, diffusion with improved water condition.\"That's true; however, it is widely known that an organic matter also has negative impacts on plant growth such as nitrogen starvation. Another point of this work is that the work was done in the tropical area of which organic materials are rapidly decomposed. In addition, the soil is sandy soil.To tell the truth, I was shocked to see the characteristic of the sandy soil in Northeast Thailand for the first.By the way, what do you think about the data? May I have your comment on the data?"
}
]
},
{
"id": "74754",
"date": "16 Nov 2020",
"name": "Oskar Franklin",
"expertise": [
"Reviewer Expertise Process-based ecological modeling and theory"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study quantifies the effects of additions of organic matter (SOM) with high C:N ratio, N fertilizer, and water on lettuce crop yield, soil microbial biomass (SMB), and soil N. There is a strong positive correlation observed between SMB and yield. Still it is argued that SOM addition induces N starvation for the crop, and that N does not flow via SMB to crops, which instead get their N from N-fixation.\nI am not an expert in this field at all, so I cannot comment on the practical experimental aspects, only on the analysis. While I appreciate concise and short papers, this paper is too short. Many things are not well described, such as the frequency of N application, the significance of the decomposition results, and the definition and estimates of N fixation. These aspects seem to be important for the conclusions so they need to be well explained. Without better explanations I cannot understand the conclusion that SMB is not involved in the provision of N to the crop despite the strong correlation between the two. For example, would it not be possible that the activity of SMB releases inorganic and organic N that is intercepted by plants and is responsible for the yield increase? Rates of microbial N cycling does not necessarily correlate with the microbial biomass.\n\nThe estimates of soil microbial N content also seem very high in comparison to total N. In many cases actually higher than total N, which should not be possible.\n\nIn conclusion, it is an interesting topic, but this study requires better descriptions of all assumptions and conclusions, and an explanation or re-evaluation of incompatible N values in soil and microbes.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": [
{
"c_id": "6249",
"date": "13 Jan 2021",
"name": "Masato Oda",
"role": "Author Response",
"response": "Thank you so much for the precious comments. We made the following improvements. frequency of N application N application is a starter of decomposition (Added in the “treatments”). the significance of the decomposition results TC contents of the crop soil were added in the “Fate of waste mushroom bed”. the definition and estimates of N fixation N fixation is estimated at most 32.1 g m −2 from the difference of the max and min of SMB in season3 (Added in the “Source of nitrogen”). Would it not be possible that the activity of SMB releases inorganic and organic N that is intercepted by plants and is responsible for the yield increase? The rate of decomposition (amount of released nitrogen) showed no correlation with yield or SMB. Rates of microbial N cycling does not necessarily correlate with the microbial biomass. That’s right! Our conclusion is that SMB and plant biomass need similar conditions but there are no strong relations between SMB and plant biomass. In many cases actually higher than total N, which should not be possible. The 3.1 of K factor could be different, e.g. 2.5, according to the soil (Joergensen & Brookes, 1990)."
}
]
}
] | 1
|
https://f1000research.com/articles/9-90
|
https://f1000research.com/articles/10-21/v1
|
13 Jan 21
|
{
"type": "Study Protocol",
"title": "Exclusive human milk diet for very preterm babies in England: protocol for a cost-effectiveness and budget impact analysis",
"authors": [
"Chris Sampson",
"Kyann Zhang",
"David Parkin",
"Grace Hampson",
"Kyann Zhang",
"David Parkin",
"Grace Hampson"
],
"abstract": "Introduction: Babies born before 30 weeks’ gestation are at increased risk of major clinical complications and have greater nutritional requirements. Where nutritional requirements cannot be sufficiently provided for by the mother’s own milk (MOM), routine care in England uses cow milk-derived fortifiers and formulas. However, the use of cow milk in the diets of preterm babies has been associated with adverse health outcomes. Clinical trials have shown that an exclusive human milk diet (EHMD) – where MOM is supplemented by donor human milk-derived formulas and fortifiers – has the potential to be clinically beneficial and reduce the risk of complications. Objectives: This study has two key objectives: 1) estimate the cost-effectiveness of an EHMD for babies born before 30 weeks’ gestation, relative to routine care; 2) estimate the budget impact of adopting EHMDs in practice in England. Methods: The analysis will use a modelling approach based on the most relevant data available. The population will consist of babies born in England before 30 weeks’ gestation. Babies in the intervention arm will be simulated to represent outcomes associated with babies fed an EHMD, and those in the comparator arm to receive routine care. Model parameters will be drawn from three sources: i) a recently completed randomised clinical trial, ii) the National Neonatal Research Database, and iii) published literature. The model will adopt a time horizon of two years following initial admission to a neonatal unit. The primary outcome for the cost-effectiveness analysis will be the incremental cost per life-year gained (if observed) associated with the intervention, relative to the comparator. We will also present disaggregated outcomes in a cost-consequence analysis. The primary outcome for the budget impact analysis will be the total cost associated with EHMD compared with current practice from the perspective of the English National Health Service (NHS).",
"keywords": [
"human milk",
"diet",
"neonatal",
"nutrition",
"cost-effectiveness",
"budget impact",
"preterm infants"
],
"content": "Background\n\nBabies born before 30 weeks’ gestation are at increased risk of major clinical complications including necrotising enterocolitis (NEC), sepsis, and mortality1. The clinical management of preterm babies is complicated by their having greater nutritional requirements than full-term babies. In many cases, the mothers’ own milk (MOM) is not sufficient – in volume or nutritional content – to meet preterm babies’ needs. Consequently, both preterm formulas and milk fortifiers are used to feed preterm infants.\n\nIn England, routinely used fortifiers and formulas are derived from cow milk. The use of cow milk-derived fortifier (CMDF) in the diet of preterm infants has been shown to be associated with several adverse health outcomes2. Clinical trials have demonstrated that an exclusive human milk diet (EHMD), based on a MOM alongside fortifiers and formulas manufactured from donor human milk, may be clinically beneficial3,4. An EHMD has been associated with reduced risk of negative sequelae such as NEC, sepsis, neurodevelopmental problems, and lung disease5,6.\n\nA randomised controlled trial was recently completed in England, sponsored by Newcastle Hospitals NHS Foundation Trust. The Interactions between the diet and gut microbes and metabolism in preterm infants (INDIGO) study sought to evaluate EHMDs in the English setting in terms of its impact on gut bacteria and body composition7. The INDIGO trial also recorded data relating to health care resource use and clinical endpoints.\n\nAn EHMD, where human milk-derived fortifier (HMDF) and formula are provided (where MOM is insufficient for the preterm infant’s nutritional needs), is likely to be associated with higher upfront costs for the provision of nutrition. However, the major cost of neonatal care in England is attributable to time spent in a neonatal unit (NNU). If an EHMD reduced the time spent in the NNU, it could reduce costs overall.\n\nPrevious studies have evaluated the cost-effectiveness of an EHMD for low birth weight babies in the United States and found that it is likely to reduce mortality and reduce costs by reducing adverse clinical events8–10. However, there are important differences between the United States and the National Health Service (NHS) context in England, which mean that the findings may not be applicable. No previous studies have estimated the cost-effectiveness of an EHMD for low birth weight babies in England.\n\n\nMethods\n\nThe aim of this analysis is to estimate the expected cost-effectiveness of an EHMD for preterm babies in England, and the budget impact of adopting its use in practice. The analysis will use a modelling approach based on the most relevant data available.\n\nThe population will be babies born in England before 30 weeks’ gestation, which aligns with the inclusion criteria used in the INDIGO trial. The population will represent a complete cohort of babies admitted to NNUs in England within one year.\n\nBabies in the intervention arm are fed with MOM, supplemented with HMDFs (Humavant®+6 human milk fortifier [human, pasteurized], Prolacta Bioscience) with or without human milk-derived ready-to-feed preterm formula (Humavant® RTF 26 human milk-based premature infant formula, Prolacta Bioscience). The intervention arm is henceforth referred to as EHMD.\n\nBabies in the comparator arm are fed with MOM, supplemented with CMDFs with or without cow milk-derived ready-to-feed formula. This comparator is intended to represent usual care in England, though usual care can vary between hospitals.\n\nThe cost-effectiveness analysis will estimate the cost per life-year associated with the intervention and comparator, using the best available evidence. If an EHMD is associated with improved outcomes and greater costs, its cost-effectiveness will be estimated as the cost per life year gained. This analysis will be conducted from the perspective of the NHS in England.\n\nA secondary analysis will consider disaggregated outcomes in the form of a cost-consequence analysis. These outcomes will include counts of key events including death and several diagnostic indicators as described below.\n\nAs with the cost-effectiveness analysis, the budget impact of an EHMD will be estimated from the perspective of the NHS in England. This will be summarised as the total incremental cost based on health care costs associated with nutritional provision, and complications that incur service use. Costs will also be presented in a disaggregated form to guide decision-making at different levels (e.g. national and local).\n\nThe time horizon for the analysis will be two years from baseline, where baseline is initial admission to an NNU. Costs will be discounted at an annual rate of 3.5% for the cost-effectiveness analysis in accordance with methodological guidance published by the National Institute for Health and Care Excellence (NICE). Discounting will not be applied for the budget impact analysis.\n\nThe overall approach for the analysis will be a model-based cost-effectiveness analysis. We will construct an individual sampling model to simulate clinical pathways and disease events for individual babies. The study is informed by published methods and reporting guidance, as set out in principles of good practice in state-transition modelling, budget impact analysis, and reporting for economic evaluations of health interventions11–14. The model will be developed using Microsoft Excel (Microsoft 365 version).\n\nWe will develop a probabilistic discrete-time state-transition microsimulation. The cycle length for the model will be one day. We will conduct 10,000 Monte Carlo simulations for the purpose of probabilistic sensitivity analysis. Each simulation will count the occurrence of events and sum costs over the time horizon.\n\nThe state-based transition model will have seven states, made up of four levels of neonatal care – intensive, high dependency, special, and transitional – inpatient hospital care, home, and death, as shown in Figure 1. Each state will be associated with a per-cycle cost. Each day in a neonatal care state will also be associated with a cost of nutrition.\n\nInformed by the modelling exercise reported by Seaton et al.15, we will assume that infants born before 30 weeks’ gestation are transferred to one of three levels of neonatal care: intensive care, high dependency care, or special care, and that subsequent transitions are to lower levels of dependency. While this may not always be the case in practice, the key driver of health care costs is likely to be length of stay, rather than the specific pathway, and so we do not anticipate that this simplifying assumption will introduce substantial bias to our cost estimates.\n\nTransitions are modelled from any neonatal care state to any post-discharge state. An unpopulated transition matrix is shown in Table 1.\n\nBlack cells represent transitions with zero probability. White cells represent transitions with positive probability. Grey cells represent the probability of no transition.\n\nA set of events can occur before a baby is discharged from neonatal care. Our model will include the following events:\n\nSurgical treatment for NEC\n\nDiagnosis of late-onset sepsis\n\nDiagnosis of short bowel syndrome\n\nDiagnosis of bronchopulmonary dysplasia (BPD)\n\nDiagnosis of retinopathy of prematurity (ROP)\n\nDiagnosis of neurosensory impairment\n\nThe probability of these events occurring will be assumed to be fixed across the different levels of care but to be potentially co-dependent on other events. For instance, the probability of short bowel syndrome and BPD will be associated with the occurrence and treatment of NEC. Stochastic occurrence of all possible events will be recorded within each cycle of each simulation. Each event will be associated with a cost, if relevant.\n\nTable 2 shows the list of parameters that will be required by the model and their candidate sources. Transition probabilities, event probabilities, and diet-specific costs will depend on treatment allocation.\n\nAbbreviations: NNRD – National Neonatal Research Database; INDIGO – Interactions between the diet and gut microbes and metabolism in preterm infants (study); NNU – neonatal unit; NHS – National Health Service; NEC – necrotising enterocolitis; BPD – bronchopulmonary dysplasia; ROP – retinopathy of prematurity; RTF – ready-to-feed\n\nAs part of the INDIGO trial, data were collected for participants, both directly and through the National Neonatal Research Database (NNRD). The variables available from the INDIGO trial are shown in Table 3.\n\nAbbreviations: RTF – ready-to-feed; NEC – necrotising enterocolitis; BPD – bronchopulmonary dysplasia; ROP – retinopathy of prematurity.\n\nCollection and analysis of variables as part of the INDIGO study was approved by the North East –Tyne & Wear South Research Ethics Committee (REC reference 17/NE/0169).\n\nThe key driver of total costs is likely to be the length of stay in the NNU. The INDIGO data will be used to estimate daily transition probabilities between different levels of care, assuming that babies are admitted to the highest level of care observed and are discharged from the lowest level of care observed, where intensive care > high dependency care > special care > transitional care. As described above, we do not anticipate that this assumption will introduce substantial bias to our cost estimates. Each transition probability will be derived from the rate at which babies leave each state.\n\nThe INDIGO data will also be used to estimate the cost of nutrition associated with each comparator, based on the quantity of Humavant+6 fortifier, Humavant RTF 26 premature infant formula, and other formula provided.\n\nKey clinical inputs for this project will be sought through collaboration with clinical experts and from existing publications of previous research. Published sources used will include studies focusing on the prevalence and prognosis of complications associated with very premature babies (for example, (e.g. 21), as well as the outcomes of procedures (e.g. surgery) used to address these complications (e.g. 16). We will source papers that report estimates that most closely correspond to parameters required by our model, will use evidence from England wherever available, and will also prioritise more recent data over older data.\n\nWe will use NNRD data to define the population and to support external validation of our model. The extracted data items will be at the individual level, as described in Table 4.\n\nAbbreviations: NNAP – National Neonatal Audit Programme; NEC – necrotising enterocolitis; BPD – bronchopulmonary dysplasia\n\nThe size of the population will be determined by the NNRD population, which we will assume to be equal to the number of eligible babies born in England for the one-year period from 1 January 2019 to 31 December 2019. Each baby simulated by the model will be attributed a birth weight and gestation length at birth, which will be used to determine the amount of feed required. The comparator group will be simulated to be of the same size and birth characteristics. The NNRD data will also define the proportion of babies allocated to intensive, high dependency, or special care at initial admission to the NNU.\n\nWe will compare our estimates with nationally representative data from NNRD to externally validate the estimates of our model with respect to clinical outcomes and resource use.\n\nAn application has been submitted to a national Research Ethics Committee for the use of NNRD data for the budget impact analysis. This study will involve analysis of data already collected by the NNRD, with no novel data collection or identifiable information used.\n\nThe time horizon for both the cost-effectiveness analysis and the budget impact analysis will be two years following admission to the NNU. Costs will be calculated from the perspective of the NHS using a combination of data from the INDIGO clinical trial and NHS Reference Costs.\n\nThe key outcome of the cost-effectiveness model will be the incremental cost per life-year gained for preterm babies fed with an EHMD, relative to those receiving standard care. Costs considered will include upfront costs associated with providing an EHMD, as well as costs of health care resource use associated with common clinical complications in preterm babies, including BPD and ROP. Only directly incurred costs associated with these clinical events will be included.\n\nThe budget impact will be calculated as the difference in total cost between a scenario where babies are fed an EHMD, and one in which CMDFs (with or without cow milk-derived read-to-feed formula) are used. Cost items included will be the same as those for the cost-effectiveness model.\n\nThe budget impact analysis will adopt a payer (NHS) perspective. The time horizon will be two years post-admission. The model will evaluate additional costs arising from the switch to a more expensive feeding regime against potential reductions in costs associated with lower health care resource use as a result of improved health outcomes and lower rates of complications (if observed).\n\nThe increase in costs associated with an EHMD consist of the additional (total) cost of human milk supplementation, which in turn will depend on the additional cost per day of human milk supplementation, the length of time supplementation is required, and the size of the target population. Cost reductions may arise from improved health outcomes for very preterm babies, with reductions in morbidity, surgical procedures (and associated complications), along with reduced length of stay in enhanced care facilities.\n\nThe overall budget impact will be presented as a net cost (or saving) to NHS England.\n\nAs a sensitivity analysis, we will conduct a within-trial analysis using only INDIGO trial data in combination with unit cost estimates.\n\nEstimates generated by our model will be compared to estimates from NNRD as a means of externally validating our model. We will compare the following between NNRD and our model’s estimates for the usual care arm:\n\nMean length of stay in NNU\n\nNumber of surgical NEC cases\n\nNumber of diagnoses of ROP\n\nNumber of sepsis diagnoses\n\nFindings of this study will be published in a peer-reviewed journal or other publishing platform.\n\nStudy status. The study is in the initial planning stages, with early development of the model. The researchers have not yet accessed any data to be analysed as part of the study. Funding for the study is secured and the study has provisional approval from the Neonatal Data Analysis Unit (subject to ethical approval) to access NNRD data.\n\n\nDiscussion\n\nThis study is the first economic evaluation of EHMD use for very preterm babies in England. Given the potential for EHMD to reduce the incidence of health complications associated with significant costs to the health system – as shown in a previous evaluation for the United States – it may represent significantly reduced costs for the NHS and alleviate pressure on neonatal care resources. Beyond cost considerations, this intervention has the potential to bring about significant improvements in quality of life for preterm babies and, by association, their carers.\n\nBy using the results of a recent clinical trial for an EHMD in England, as well as costs specific to the English setting, the findings here will be highly relevant to decision-making about whether to use EHMD in the NHS. The inclusion of both a cost-consequence and budget impact analysis will allow us to illustrate a more comprehensive picture of the overall impact of an EHMD on the NHS.\n\n\nData availability\n\nNo data are associated with article.",
"appendix": "Acknowledgments\n\nWe are grateful to Prof Nicholas Embleton for the expert opinion provided during the development of our study. Any errors or omissions are our own.\n\n\nReferences\n\nPatel RM, Kandefer S, Walsh MC, et al.: Causes and Timing of Death in Extremely Premature Infants from 2000 through 2011. N Engl J Med. 2015; 372(4): 331–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAbrams SA, Schanler RJ, Lee ML, et al.: Greater Mortality and Morbidity in Extremely Preterm Infants Fed a Diet Containing Cow Milk Protein Products. Breastfeed Med. 2014; 9(6): 281–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSullivan S, Schanler RJ, Kim JH, et al.: An Exclusively Human Milk-Based Diet Is Associated with a Lower Rate of Necrotizing Enterocolitis than a Diet of Human Milk and Bovine Milk-Based Products. J Pediatr. 2010; 156(4): 562-567.e1. PubMed Abstract | Publisher Full Text\n\nCristofalo EA, Schanler RJ, Blanco CL, et al.: Randomized Trial of Exclusive Human Milk versus Preterm Formula Diets in Extremely Premature Infants. J Pediatr. 2013; 163(6): 1592-1595.e1. PubMed Abstract | Publisher Full Text\n\nBuckle A, Taylor C: Cost and Cost-Effectiveness of Donor Human Milk to Prevent Necrotizing Enterocolitis: Systematic Review. Breastfeed Med. 2017; 12(9): 528–36. PubMed Abstract | Publisher Full Text\n\nHair AB, Peluso AM, Hawthorne KM, et al.: Beyond Necrotizing Enterocolitis Prevention: Improving Outcomes with an Exclusive Human Milk-Based Diet. Breastfeed Med. 2016; 11(2): 70–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEmbleton N: Interactions between diet and gut microbes in preterm infants. ISRCTN; 2017 [cited 2020 Apr 16]. Publisher Full Text\n\nHampson G, Roberts SLE, Lucas A, et al.: An economic analysis of human milk supplementation for very low birth weight babies in the USA. BMC Pediatr. 2019; 19(1): 337. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGanapathy V, Hay JW, Kim JH: Costs of Necrotizing Enterocolitis and Cost-Effectiveness of Exclusively Human Milk-Based Products in Feeding Extremely Premature Infants. Breastfeed Med. 2012; 7(1): 29–37. PubMed Abstract | Publisher Full Text\n\nJohnson TJ, Patel AL, Bigger HR, et al.: Economic Benefits and Costs of Human Milk Feedings: A Strategy to Reduce the Risk of Prematurity-Related Morbidities in Very-Low-Birth-Weight Infants. Adv Nutr. 2014; 5(2): 207–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSullivan SD, Mauskopf JA, Augustovski F, et al.: Budget Impact Analysis-Principles of Good Practice: Report of the ISPOR 2012 Budget Impact Analysis Good Practice II Task Force. Value Health. 2014; 17(1): 5–14. PubMed Abstract | Publisher Full Text\n\nCaro JJ, Briggs AH, Siebert U, et al.: Modeling Good Research Practices--Overview: A Report of the ISPOR-SMDM Modeling Good Research Practices Task Force-1. Value Health. 2012; 15(6): 796–803. PubMed Abstract | Publisher Full Text\n\nHusereau D, Drummond M, Petrou S, et al.: Consolidated Health Economic Evaluation Reporting Standards (CHEERS)--Explanation and Elaboration: A Report of the ISPOR Health Economic Evaluation Publication Guidelines Good Reporting Practices Task Force. Value Health. 2013; 16(2): 231–50. PubMed Abstract | Publisher Full Text\n\nSiebert U, Alagoz O, Bayoumi AM, et al.: State-Transition Modeling: A Report of the ISPOR-SMDM Modeling Good Research Practices Task Force--3. Value Health. 2012; 15(6): 812–20. PubMed Abstract | Publisher Full Text\n\nSeaton SE, Barker L, Draper ES, et al.: Modelling Neonatal Care Pathways for Babies Born Preterm: An Application of Multistate Modelling. PLoS One. 2016; 11(10): e0165202. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRees CM, Pierro A, Eaton S: Neurodevelopmental outcomes of neonates with medically and surgically treated necrotizing enterocolitis. Arch Dis Child Fetal Neonatal Ed. 2007; 92(3): F193-198. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCole CR, Hansen NI, Higgins RD, et al.: Very low birth weight preterm infants with surgical short bowel syndrome: incidence, morbidity and mortality, and growth outcomes at 18 to 22 months. Pediatrics. 2008; 122(3): e573-e582. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPanickar J, Scholefield H, Kumar Y, et al.: Atypical chronic lung disease in preterm infants. J Perinat Med. 2004; 32(2): 162–7. PubMed Abstract | Publisher Full Text\n\nO’Connor DL, Kiss A, Tomlinson C, et al.: Nutrient enrichment of human milk with human and bovine milk–based fortifiers for infants born weighing <1250 g: a randomized clinical trial. Am J Clin Nutr. 2018; 108(1): 108–16. PubMed Abstract | Publisher Full Text\n\nAdams GGW, Bunce C, Xing W, et al.: Treatment trends for retinopathy of prematurity in the UK: active surveillance study of infants at risk. BMJ Open. 2017; 7(3): e013366. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLui K, Lee SK, Kusuda S, et al.: Trends in Outcomes for Neonates Born Very Preterm and Very Low Birth Weight in 11 High-Income Countries. J Pediatr. 2019; 215: 32-40.e14. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "86762",
"date": "18 Jun 2021",
"name": "Hema Mistry",
"expertise": [
"Reviewer Expertise Health Economics"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this protocol, the authors want to explore the cost-effectiveness of exclusive human milk diet (EHMD) with routine care for babies born before 30 weeks' gestation. They will do this by building an economic model using available data from a randomised clinical trial, the National Neonatal Research Database, and published literature. The time horizon will be two years. They will adopt a National Health Service perspective and the results will be presented as cost per life year gained. They also aim to do a budget impact analysis, looking at the total cost associated with EHMD compared with current practice.\n\nThe protocol was well written and the objectives of the study are well defined. I have a few minor comments to help improve the study:\nIn the Abstract, the authors state that: “the primary outcome for the cost-effectiveness analysis will be incremental cost per life year gained (if observed)”. If this is not observed, what will the primary outcome be? This needs to be spelt out in the Abstract.\n\nIn the Abstract, it would also be useful to spell out some of the disaggregated outcomes that you plan to present in the cost-consequence analysis.\n\nIn the Methods section, the authors say that: \"the population will be babies born in England before 30 weeks' gestation\", as this aligns with the trial. Technically, babies born before 37 weeks are known as preterm, babies born before 32 weeks are very preterm, and babies born before 28 weeks are extremely preterm. Will the analysis miss out on costs/outcomes if we are excluding babies 31- or 32-weeks’ gestation?\n\nWould any of the babies after one year of age still be exclusively fed on EHMD diet or would they have cow’s milk?\n\nIn the transition matrix/model, would you not need a separate health state for any outpatient visits/A&E visits etc., seeing as the model cycle length is one day?\n\nAre contacts with other health professionals in the hospital, such as hospital dietician, included or excluded?\n\nAlso, the baby during the first year of life would have contacts with health professionals in the community, such as health visitor or a breastfeeding midwife, are these costs going to be included?\n\nApart from the trial and NHS reference costs, would any other sources be required to get unit cost information?\n\nIf the baby is included in the study until the age of two and they had a hospital visit between 6 months and 2 years, would they include any other costs of food/solids on top of the milk?\n\nIn terms of the formula and fortifier costs, would the costs of sterilisers, bottles and teats be included as well?\n\nFor babies that died, would post-mortem costs be included?\n\nI think you need a little more detail on how you will conduct the cost-consequence analysis.\n\nAre the any other planned sensitivity analyses apart from the one that was stated in the protocol?\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes",
"responses": []
},
{
"id": "87914",
"date": "09 Aug 2021",
"name": "Tuan T Nguyen",
"expertise": [
"Reviewer Expertise Nutritional Epidemiology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nOverall, this is an important, well-written research protocol. I have some specific comments.\nBackground\nParagraph 3/Intervention: Please compare different types of milk used for INDIGO. A table would be useful.\n\nPlease specify the difference between the context in the US and UK which might lead to the difference between findings from the UK and the US. The justification of literature gaps needs to be clearer.\n\nMethods\nParagraph 1: Please put the statement of goal at the end of the Background. It will be great to have specific objectives (i.e. to address key components of the studies).\n\nPlease indicate how each of the three datasets would contribute to the objectives? Description of the three data sets upfront and the use of the data. Could be in the form of a table.\n\nThe data analysis part is based heavily on the data from INDIGO. What about the other two datasets?\n\nA diagram to present the two arms of the INDIGO (given this has been done already).\n\nPlease specify how to estimate the cost before cost-effectiveness or cost-consequence analysis.\n\nWhy don't the authors perform a data analysis based on the real data available from the INDIGO? The description of the component cost might need a clearer description.\n\nThe simulation data analysis could be done at a later stage to show the variation across assumptions.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Partly\n\nAre the datasets clearly presented in a useable and accessible format? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-21
|
https://f1000research.com/articles/9-598/v1
|
12 Jun 20
|
{
"type": "Opinion Article",
"title": "Insulin Receptors and Intracellular Ca2+ Form a Double-Negative Regulatory Feedback Loop Controlling Insulin Sensitivity",
"authors": [
"Igor Pomytkin",
"Vsevolod Pinelis"
],
"abstract": "Since the discovery of insulin and insulin receptors (IR) in the brain in 1978, numerous studies have revealed a fundamental role of IR in the central nervous system and its implication in regulating synaptic plasticity, long-term potentiation and depression, neuroprotection, learning and memory, and energy balance. Central insulin resistance has been found in diverse brain disorders including Alzheimer’s disease (AD). Impaired insulin signaling in AD is evident in the activation states of IR and downstream signaling molecules. This is mediated by Aβ oligomer-evoked Ca2+ influx by activating N-methyl-D-aspartate receptors (NMDARs) with Aβ oligomers directly, or indirectly through Aβ-induced release of glutamate, an endogenous NMDAR ligand. In the present opinion article, we highlight evidence that IR and free intracellular Ca2+ concentration [Ca2+]i form a double-negative regulatory feedback loop controlling insulin sensitivity, in which mitochondria play a key role, being involved in adenosine triphosphate (ATP) synthesis and IR activation. We found recently that the glutamate-evoked rise in [Ca2+]i inhibits activation of IR and, vice versa, insulin-induced activation of IR inhibits the glutamate-evoked rise in [Ca2+]i. In theory, such a double-negative feedback loop generates bistability. Thus, a stable steady state could exist with high [Ca2+]i and nonactive IR, or with active IR and low [Ca2+]i, but no stable steady state is possible with both high [Ca2+]i and active IR. Such a circuit could toggle between a high [Ca2+]i state and an active IR state in response to glutamate and insulin, respectively. This model predicts that any condition leading to an increase of [Ca2+]i may trigger central insulin resistance and explains why central insulin resistance is implicated in the pathogenesis of AD, with which glutamate excitotoxicity is a comorbid condition. The model also predicts that any intervention aiming to maintain low [Ca2+]i may be useful for treating central insulin resistance.",
"keywords": [
"Insulin",
"insulin receptor",
"glutamate",
"NMDA receptor",
"Ca2+",
"double-negative feedback loop",
"mitochondria",
"ATP"
],
"content": "Introduction\n\nSince the discovery of insulin1 and insulin receptors (IR)2 in the brain in 1978, numerous studies have revealed a fundamental role of IR in the central nervous system (CNS). IR-mediated signaling is implicated in the regulation of diverse functions in the CNS, including synaptic plasticity, long-term potentiation and depression, neuroprotection, learning and memory, and energy balance3. Central insulin resistance has been found in neurodegenerative diseases such as Alzheimer’s disease (AD)4,5 and Parkinson’s disease (PD)6, stroke, and traumatic brain injury (TBI)7. Impaired insulin signaling in AD is evident in the activation states of IR and downstream signaling molecules5. Compared with control cases, insulin in AD brains induced 24–58% less activation at the level of IR and 90% less activation of insulin receptor substrate 1 (IRS-1)5. It has been presumed5 that the inhibition of IR activation is mediated by Aβ oligomer-triggered Ca2+ influx, in part by activating N-methyl-D-aspartate receptors (NMDARs)8, followed by a rise in Akt1 pS473 9, which can inhibit insulin-induced IR activation through Thr phosphorylation of the IR β subunit10. Aβ oligomers may activate the NMDAR-gated Ca2+ influx directly11 or indirectly through the intermediate release of glutamate, a ligand of NMDAR11–15. This suggests that the rise in intracellular free Ca2+ concentration ([Ca2+]i), evoked by either Aβ oligomers or glutamate, leads to dysfunctional activation of IR in AD. In the present opinion article, we highlight evidence that IR and [Ca2+]i form a double-negative regulatory feedback loop controlling insulin sensitivity, and mitochondria have a key role in this feedback loop, being involved in adenosine triphosphate (ATP) synthesis and IR activation.\n\n\nGlutamate-evoked rise in [Ca2+]i causes inhibition of IR signaling\n\nGlutamate serves as the major excitatory neurotransmitter in the CNS. Its excessive accumulation in a synaptic cleft can trigger excitotoxicity, a pathologic process leading to neuronal cell death. Glutamate-induced activation of the NMDAR-gated Ca2+ influx is generally considered central to the development of excitotoxicity16. Prolonged glutamate exposure causes a rapid initial increase in the [Ca2+]i, followed by a larger secondary [Ca2+]i increase concomitant with a decrease in the mitochondrial inner membrane potential (ΔΨm)17–19. We recently found that on Ca2+-induced mitochondrial depolarization, insulin induced 48% less activation of IR (assessed by pY1150/1151) compared with control20. Earlier, we showed that a decrease in ΔΨm can abrogate IR activation18, since the ΔΨm-dependent mitochondrial signal at complex II is involved in the activation of IR in neurons21–23. Thus, the glutamate-evoked increase in [Ca2+]i, followed by the drop in ΔΨm, leads to the inhibition of insulin-induced activation of IR.\n\n\nInsulin prevents glutamate-evoked rise in [Ca2+]i\n\nNormally, the NMDAR-gated Ca2+ influx is counterbalanced with Ca2+ efflux, which is governed by plasma membrane Ca2+ ATPase and the Na+/Ca2+ exchanger (NCX)24,25. NCX-mediated Ca2+ efflux is also ATP-dependent, since NCX exchanges one Ca2+ for three Na+, and the three Na+ are then pumped out by the Na+/K+ ATPase at the expense of one ATP. In excitotoxicity, prolonged stimulation with glutamate leads to ATP depletion and an abnormal rise in [Ca2+]i, since the massive Ca2+ influx is no longer counterbalanced by Ca2+ efflux26. Therefore, maintenance of ATP production is crucial for preventing the rise in [Ca2+]i in excitotoxicity. We found recently that pre-treatment with insulin prevents neurons from glutamate-evoked ATP depletion due to its protective effect on spare respiratory capacity (SRC), a measure that relates to the amount of extra ATP that can be produced via oxidative phosphorylation in case of increased energy demand19. The effect of insulin on SRC relates to its action on mitochondrial metabolism. It has long been known that the tricarboxylic acid cycle is the intracellular site of insulin action and that insulin acutely stimulates succinate oxidation at mitochondrial complex II26,27. Succinate oxidation at mitochondrial complex II has been identified recently as the main source of SRC28. In line with this, insulin prevented the glutamate-evoked rise in [Ca2+]i in our experiments with glutamate excitotoxicity19.\n\n\nIR and [Ca2+]i form a double-negative feedback loop controlling insulin sensitivity\n\nCollectively, this evidence suggests that a double-negative regulatory feedback loop exists between IR and [Ca2+]i. The glutamate-evoked rise in [Ca2+]i inhibits activation of IR and, vice versa, insulin-induced activation of IR inhibits the glutamate-evoked rise in [Ca2+]i (Figure 1a).\n\n(A) glutamate triggers NMDA receptor–gated Ca2+ influx, inhibiting IR activation, and insulin triggers activation of IR, inhibiting [Ca2+]i rise; (B) glutamate-triggered stable steady state with high [Ca2+]i and nonactive IR (pY-IR↓); (C) insulin-triggered stable steady state with active IR (pY-IR↑) and low [Ca2+]i.\n\nIn theory, a double-negative feedback loop generates bistability29. Thus, a stable steady state could exist with high [Ca2+]i and nonactive IR (Figure 1b), or with active IR and low [Ca2+]i (Figure 1c), but no stable steady state is possible with both high [Ca2+]i and active IR. Such a circuit could toggle between a high [Ca2+]i state and an active IR state in response to glutamate and insulin, respectively.\n\nThis double-negative feedback loop model predicts that any condition leading to an increase in [Ca2+]i may trigger insulin resistance. It appears to explain why central insulin resistance is implicated in the pathogenesis of disorders such as AD4,5, PD6, stroke, and TBI7, with which glutamate excitotoxicity is a comorbid condition30. The model also predicts that any intervention aiming to prevent Ca2+ influx of or enhance efflux of Ca2+ from neurons, thereby maintaining low [Ca2+]i, may be useful for treating central insulin resistance. Given that Ca2+ efflux is ATP-dependent, any intervention directed to enhance ATP production in neurons may be especially useful to improve insulin sensitivity in the brain.\n\n\nData availability\n\nNo data are associated with this article.",
"appendix": "References\n\nHavrankova J, Schmechel D, Roth J, et al.: Identification of insulin in rat brain. Proc Natl Acad Sci U S A. 1978; 75(11): 5737–41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHavrankova J, Roth J, Brownstein M: Insulin receptors are widely distributed in the central nervous system of the rat. Nature. 1978; 272(5656): 827–829. PubMed Abstract | Publisher Full Text\n\nPomytkin I, Costa-Nunes JP, Kasatkin V, et al.: Insulin receptor in the brain: Mechanisms of activation and the role in the CNS pathology and treatment. CNS Neurosci Ther. 2018; 24(9): 763–774. PubMed Abstract | Publisher Full Text | Free Full Text\n\nArnold SE, Arvanitakis Z, Macauley-Rambach SL, et al.: Brain insulin resistance in type 2 diabetes and Alzheimer disease: concepts and conundrums. Nat Rev Neurol. 2018; 14(3): 168–181. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTalbot K, Wang HY, Kazi H, et al.: Demonstrated brain insulin resistance in Alzheimer's disease patients is associated with IGF-1 resistance, IRS-1 dysregulation, and cognitive decline. J Clin Invest. 2012; 122(4): 1316–38. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAthauda D, Foltynie T: Insulin resistance and Parkinson's disease: A new target for disease modification? Prog Neurobiol. 2016; 145–146: 98–120. PubMed Abstract | Publisher Full Text\n\nZhou Q, Sheng M: NMDA receptors in nervous system diseases. Neuropharmacology. 2013; 74: 69–75. PubMed Abstract | Publisher Full Text\n\nAlberdi E, Sánchez-Gómez MV, Cavaliere F, et al.: Amyloid beta oligomers induce Ca2+ dysregulation and neuronal death through activation of ionotropic glutamate receptors. Cell Calcium. 2010; 47(3): 264–72. PubMed Abstract | Publisher Full Text\n\nPerkinton MS, Ip JK, Wood GL, et al.: Phosphatidylinositol 3-kinase is a central mediator of NMDA receptor signalling to MAP kinase (Erk1/2), Akt/PKB and CREB in striatal neurones. J Neurochem. 2002; 80(2): 239–54. PubMed Abstract | Publisher Full Text\n\nMorisco C, Condorelli G, Trimarco V, et al.: Akt mediates the cross-talk between beta-adrenergic and insulin receptors in neonatal cardiomyocytes. Circ Res. 2005; 96(2): 180–8. PubMed Abstract | Publisher Full Text\n\nTexidó L, Martín-Satué M, Alberdi E, et al.: Amyloid β peptide oligomers directly activate NMDA receptors. Cell Calcium. 2011; 49(3): 184–90. PubMed Abstract | Publisher Full Text\n\nZhao WQ, De Felice FG, Fernandez S, et al.: Amyloid beta oligomers induce impairment of neuronal insulin receptors. FASEB J. 2008; 22(1): 246–60. PubMed Abstract | Publisher Full Text\n\nKabogo D, Rauw G, Amritraj A, et al.: ß-amyloid-related peptides potentiate K+-evoked glutamate release from adult rat hippocampal slices. Neurobiol Aging. 2010; 31(7): 1164–72. PubMed Abstract | Publisher Full Text\n\nFindley CA, Bartke A, Hascup KN, et al.: Amyloid Beta-Related Alterations to Glutamate Signaling Dynamics During Alzheimer's Disease Progression. ASN Neuro. 2019; 11: 1759091419855541. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHascup KN, Hascup ER: Soluble Amyloid-β42 Stimulates Glutamate Release through Activation of the α7 Nicotinic Acetylcholine Receptor. J Alzheimers Dis. 2016; 53(1): 337–47. PubMed Abstract | Publisher Full Text\n\nNicholls DG: Mitochondrial dysfunction and glutamate excitotoxicity studied in primary neuronal cultures. Curr Mol Med. 2004; 4(2): 149–77. PubMed Abstract | Publisher Full Text\n\nVergun O, Keelan J, Khodorov BI, et al.: Glutamate-induced mitochondrial depolarisation and perturbation of calcium homeostasis in cultured rat hippocampal neurones. J Physiol. 1999; 519(Pt 2): 451–66. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAbramov AY, Duchen MR: Mechanisms underlying the loss of mitochondrial membrane potential in glutamate excitotoxicity. Biochim Biophys Acta. 2008; 1777(7–8): 953–64. PubMed Abstract | Publisher Full Text\n\nKrasil'nikova I, Surin A, Sorokina E, et al.: Insulin Protects Cortical Neurons Against Glutamate Excitotoxicity. Front Neurosci. 2019; 13: 1027. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPomytkin I, Krasil'nikova I, Bakaeva Z, et al.: Excitotoxic glutamate causes neuronal insulin resistance by inhibiting insulin receptor/Akt/mTOR pathway. Mol Brain. 2019; 12(1): 112. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPersiyantseva NA, Storozhevykh TP, Senilova YE, et al.: Mitochondrial H2O2 as an enable signal for triggering autophosphorylation of insulin receptor in neurons. J Mol Signal. 2013; 8(1): 11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStorozhevykh TP, Senilova YE, Persiyantseva NA, et al.: Mitochondrial respiratory chain is involved in insulin-stimulated hydrogen peroxide production and plays an integral role in insulin receptor autophosphorylation in neurons. BMC Neurosci. 2007; 8: 84. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPomytkin IA: H2O2 Signalling Pathway: A Possible Bridge between Insulin Receptor and Mitochondria. Curr Neuropharmacol. 2012; 10(4): 311–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCarafoli E, Santella L, Branca D, et al.: Generation, control, and processing of cellular calcium signals. Crit Rev Biochem Mol Biol. 2001; 36(2): 107–260. PubMed Abstract | Publisher Full Text\n\nBrittain MK, Brustovetsky T, Sheets PL, et al.: Delayed calcium dysregulation in neurons requires both the NMDA receptor and the reverse Na+/Ca2+ exchanger. Neurobiol Dis. 2012; 46(1): 109–17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBessman SP, Mohan C, Zaidise I: Intracellular site of insulin action: mitochondrial Krebs cycle. Proc Natl Acad Sci U S A. 1986; 83(14): 5067–70. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBessman SP, Mohan C: Insulin as a probe of mitochondrial metabolism in situ. Mol Cell Biochem. 1997; 174(1–2): 91–6. PubMed Abstract\n\nPfleger J, He M, Abdellatif M: Mitochondrial complex II is a source of the reserve respiratory capacity that is regulated by metabolic sensors and promotes cell survival. Cell Death Dis. 2015; 6(7): e1835. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFerrell JE Jr: Self-perpetuating states in signal transduction: positive feedback, double-negative feedback and bistability. Curr Opin Cell Biol. 2002; 14(2): 140–8. PubMed Abstract | Publisher Full Text\n\nWang R, Reddy PH: Role of Glutamate and NMDA Receptors in Alzheimer's Disease. J Alzheimers Dis. 2017; 57(4): 1041–1048. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "64837",
"date": "01 Jul 2020",
"name": "Kevin N Hascup",
"expertise": [
"Reviewer Expertise glutamate signaling",
"Alzheimer's disease",
"Parkinson's disease",
"insulin signaling",
"gerontology"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors put forth an opinion article in regard to a double negative feedback loop that controls cerebral insulin receptor signaling during times of either high glutamate or insulin levels. They briefly go onto hypothesize that any disease state resulting in elevated intracellular calcium will result in reduced cerebral insulin resistance, particularly when excitotoxicity is a comorbid factor. The authors attempt to simplify an extremely complicated phenomenon. However, this oversimplification neglects to take into account several factors that need to be addressed. Reduced insulin signaling as a result for glutamate induced calcium influx has been discussed in other literature (see Zhao et al., 20081). This work should be taken into consideration when authors discuss their model.\nAlthough insulin receptors are ubiquitously expressed in different cell types of the CNS, these cell types rely on different mechanisms for ATP production. Glial cells predominantly use glycolytic pathways in the cytoplasm whereas neurons rely on oxidative phosphorylation in the mitochondria. (For review see Pellerin and Magistretti, 20122). Accordingly, the cellular localization of ATP production should be taken into consideration. Furthermore, do the authors believe different cell type specific mechanisms exist that their model should take into consideration? In several places, the authors discuss the role of amyloid on activation of NMDA receptors and the subsequent intracellular Calcium increase. However, the authors have not taken into account how amyloid binding can also elicit glutamate release from either α7nAChRs (see Hascup and Hascup, 20163 and Mura et al., 20124) or mGLUR5 receptors (Renner et al., 20105). In the latter case, several laboratories have shown mGLUR5 acts as a scaffolding complex for amyloid accumulation causing receptor clustering at the membrane surface and results in elevated intracellular calcium levels.\n\nAdditionally, the manuscript could be improved if a discussion on how their model might vary across different CNS disorders such as AD, Parkinson’s TBI, etc. in relation to healthy functional activity. The authors neglect to take into account factors that also play a role in modulating insulin signaling. These include:\nInflammation. This is prominent in numerous disease states and can negatively impact insulin signaling, while exacerbating mechanisms associated with multiple neurodegenerative disorders.\n\nInhibitory feedback regulation. Insulin signaling is tightly controlled to prevent perturbations in metabolism as well as control the specificity of the signal on multiple downstream effectors. Several phosphatases are responsible for this, not just at the receptor, but also on effector enzymes.The current model does not take into consideration this tightly controlled feedback loop.\n\nReceptor internalization. Upon insulin binding, the insulin receptor becomes internalized as another means to control the strength and duration of the signal. This internalization is more prominent in hyperinsulinemia and may account for the resulting insulin resistance. How would this model change during stages of insulin resistance, which are hypothesized to initiate the cognitive decline observed in AD?\n\nPeripheral insulin signaling.Insulin produced in the pancreas is able to enter the CNS.How does the proposed model take into consideration fluctuations during normal periods of food consumption?\nAre the authors proposing a similar mechanism for the structurally analogous insulin-like growth factor-1?\nMinor concerns:\nThe glutamate pathway in Figure 1 should have a different color scheme to make it easier to differentiate from Calcium concentration.\n\nFigure 1B is slightly confusing. The red line makes it seem that low levels of glutamate give rise to high levels of Calcium instead of showing just the rise in calcium. A way to incorporate glutamate activation of NMDA receptors in Fig 1B & C might help to conceptualize the model.\n\nThe abstract is lengthy in relation to the article. I would suggest this is shortened and made more succinct.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? No\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly",
"responses": [
{
"c_id": "5708",
"date": "10 Jul 2020",
"name": "Igor Pomytkin",
"role": "Author Response",
"response": "Our opinion article relates only to relationship between intracellular Ca2+ concentrations [Ca2+]i and insulin receptor activation state, and not between [Ca2+]i and insulin signaling system as a whole. Insulin resistance is the term that currently applied to any of the biological actions of insulin and, therefore, is too broad to be discussed in terms of models that may predict the system behavior. The proposed model in the opinion article is not oversimplified, but takes into consideration only link between [Ca2+]i and the stage of activation of insulin receptor kinase (i.e Tyr1150/1151 phosphorylation), the earliest step in insulin action that precedes all other signaling events and effects of insulin.In our opinion article we selected only two measurable parameters, namely [Ca2+]i and insulin receptor activation state defined as Tyr1150/1151 phosphorylation, but not downstream molecules of IR signaling pathway such as IRS-1 or others. Therefore, our opinion relates only to the activation of insulin receptor, and not to downstream molecules or events. This approach relates directly to insulin sensitivity, since the activation of the receptor with insulin is the only stage at which insulin sensitivity can be measured directly.The opinion about existence of regulatory double-negative feedback loop between [Ca2+]i and insulin receptor activation state is based on our experimental results, obtained at the same experimental conditions in two studies that have already been published [references 19 and 20].Results of Zhao et al. do not contradict our opinion about relationship between Ca2+ and insulin receptor activation state. Moreover, results of Zhao et al. support our opinion. Zhao et al. have found that glutamate stimulation reduced the insulin-stimulated tyrosine phosphorylation of the receptor beta-subunit at Tyr1150/1151 and this glutamate effect was completely inhibited by the cell-permeable Ca2+ chelator BAPTA-AM. Although authors did not measure intracellular Ca2+ concentrations, they concluded that IR inhibition is dependent on elevated intracellular Ca2+, given a well-known link between glutamate and Ca2+. It is the same conclusion that we made. The difference is only that our conclusion is made on our direct measurements of intracellular Ca2+.In our studies [references 19 and 20] that underlie the our opinion we used glia-free cortical neurons and directly measured ATP levels.Our opinion relates to link between [Ca2+]i and insulin receptor activation state, and amyloid-NMDA relationship are out of scope of our opinion.Our opinion is limited to only insulin receptor activation, but not to more broad “insulin signaling”. According to current knowledge, inflammation affects insulin signaling, but downstream of insulin receptor at IRS-1 level. This is out of scope of our opinion.Our opinion does not relate to any signaling events downstream of insulin receptor, such phosphatase action and receptor internalization. The insulin regulated internalization of insulin receptors has been shown to require autophosphorylation of all three regulatory tyrosines 1146, 1150, and 1151. Carpentier JL et al. Two steps of insulin receptor internalization depend on different domains of the beta-subunit. J Cell Biol. 1993 Sep;122(6):1243-52. doi: 10.1083/jcb.122.6.1243. Thus, the internalization occurs after the activation of IR and, therefore, is out of the scope of our opinion.According to the model, insulin inhibits rise of [Ca2+]i independently of the insulin source. Therefore, during periods of insulin transport to the brain, the rise of [Ca2+]i , e.g. glutamate-evoked, would be diminished, independently on whether it comes from pancreas or administered intranasally.We did not propose the same mechanism for IGF-1 in our opinion aticle, since we have no supportive evidence. However, it is likely that there is a link between [Ca2+]i and activation state of IGFR1.In a conclusion, the model proposed in the opinion article relates only to relationship beteen two measurable parameters, namely intracellular Ca2+ concentrations and insulin receptor activation state. Therefore, all other downstream elements of insulin signaling system and factors affecting the downstream elements are out of scope of this opinion. The opinion is completely based on our experimental results previously published, references 19 and 20 of the opinion article."
}
]
},
{
"id": "67002",
"date": "03 Aug 2020",
"name": "Zhen Deng",
"expertise": [
"Reviewer Expertise cerebral vascular disease"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis model predicts that any disease that causes elevated [Ca2+]i may trigger central insulin resistance, and explains why central insulin resistance is related to the pathogenesis of AD, and glutamate excitotoxicity is a comorbidity. The model also predicts that any intervention aimed at maintaining low [Ca2+]i can be used to treat central insulin resistance.\nIt is an interesting theory between brain insulin resistance and AD. Although not much experimental data support the theory directly, it is worth following.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": []
},
{
"id": "72629",
"date": "02 Nov 2020",
"name": "Venkatesh V. Kareenhalli",
"expertise": [
"Reviewer Expertise Systems biology",
"Liver metabolism",
"modeling signaling pathways"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors propose an existence of double negative feedback loop which may result in bistable response. It is clear that Ca signaling and insulin signaling play a role in the CNS and the related pathogenesis of neurogenerative diseases. The mitochondrial function and thereby ATP synthesis also plays a role in the stated phenotype. The proposed dual negative feedback on each other can toggle between two states, but it may not be bistable. Bistable is with respective to an input variation. It is unclear what is the input that the hypothesis is being discussed. At a given input, the existence of both steady state, with different history, yields bistability. A figure that demonstrates all the molecular components and action including mitochondrial state and metabolites is more helpful than what is illustrated.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? No",
"responses": [
{
"c_id": "6254",
"date": "13 Jan 2021",
"name": "Igor Pomytkin",
"role": "Author Response",
"response": "The manuscript has been revised in accordance with notes that (a) bistability is not a necessary consecuence of a double negative regulatory feedback loop and (b) figure 1 will be more useful when signal transduction pathways will be shown. Sentences related to \"bistability\" have been removed from the abstract, main text, and reference list. Figure 1 has been revised and includes now pathways described in the text of the first version."
}
]
}
] | 1
|
https://f1000research.com/articles/9-598
|
https://f1000research.com/articles/9-314/v1
|
30 Apr 20
|
{
"type": "Case Report",
"title": "Case Report: Isolated lingual dystonia",
"authors": [
"Zaland Ahmed Yousafzai",
"Wajeeha Qayyum",
"Sohail Khan",
"Mawara Iftikhar",
"Qazi Kamran Amin",
"Wajeeha Qayyum",
"Sohail Khan",
"Mawara Iftikhar",
"Qazi Kamran Amin"
],
"abstract": "Oromandibular dystonia is defined as a focal dystonia that manifests as forceful contractions of the face, jaw, and/or tongue. Lingual dystonia is a rare subtype of oromandibular dystonia that specifically affects the tongue. Multiple etiologies are thought to attribute to oromandibular dystonia, including brain damage, the use of neuroleptic medications, neurodegenerative disorders, metabolic disorders, neurodevelopmental disorders, and viral infections. Idiopathic cases of isolated lingual dystonia are rare and seldom reported in the literature. This report describes a 35-year-old female patient with lingual dystonia that was present at rest and aggravated during speech. Despite detailed history taking and a thorough examination, along with multiple imaging and laboratory studies, no cause could be established and her case was classified as being that of an idiopathic etiology.",
"keywords": [
"oromandibular dystonai",
"dystonia",
"neurology",
"isolated dystonia",
"lingual dystonia",
"focal dystonia"
],
"content": "Introduction\n\nDystonia refers to a movement disorder that causes abnormal muscles contractions and spams that are either sustained or intermittent, and can produce repetitive movements of the affected muscle group or abnormal posture. Dystonia is often elicited and aggravated by voluntary actions and is associated with overflow activation of the affected muscles. Dystonia can be further be classified on the basis of clinical characteristics and etiology1.\n\nOromandibular dystonia are characterized by involuntary movements involving masticatory, lingual and pharyngeal muscles. It can manifest as jaw clenching, jaw opening, jaw deviation and tongue protrusion, and can result in impaired speech, dysphagia and cosmetic disfigurement. It is often found in combination with dystonia of adjacent body regions2.\n\nThe overall prevalence of primary dystonia is estimated as 164.3 per million3. The prevalence of oromandibular dystonia is estimated to be around 68.9 per million4,5. Lingual dystonia is a rare focal dystonia, with a prevalence of 4%6. Furthermore, isolated oromandibular dystonia is rarely reported and recorded in the literature. Isolated lingual dystonia can be considered a variant of oromandibular dystonia, which are focal dystonia, affecting the muscles of the lower face. This dystonia can lead to repetitive and sustained contractions of the affected muscles. Dystonia has a large number of causes. Isolated lingual dystonia of idiopathic nature is rarely reported2. Here we report a case of isolated lingual dystonia of idiopathic etiology.\n\n\nCase report\n\nA 35-year-old woman presented to our neurology Out Patient Department (OPD) with an 11 month history of abnormal tongue movement and protrusion, which was aggravated on talking. The patient also complained of difficulty in swallowing. Her past medical history was insignificant for any comorbidities, psychiatric illness, endocrine, metabolic and neurological diseases. Her family history was also negative for any neurological diseases. The patient had previously consulted several doctors, none of whom were able to diagnose her condition. She had also visited a psychiatrist, who had put her on Fluoxetine (40mg one tablet daily) and Alprazolam (0.5mg at night) for 3 months continuously. These further aggravated her symptoms and had to be discontinued (2 months before presentation to the OPD). No dystonic postures or movements were noted in any other part of her face, neck, or any other muscle group in her body. The patient had difficulty in eating, drinking, whistling, and singing.\n\nOn examination by a speech therapist, there was difficulty in the movement of the patient’s tongue, and she could not perform simple tongue movement during several tasks. During the examination she was asked to repeat words and sentences, read short texts, converse in automatic speech, sing, and perform vowel and fricative phoneme prolongation. The tongue movement disorder was identified in all circumstances of speech and in all phonemes, except vowel and sound prolongations. Tongue protrusion occurred more often in alveodental and alveolar phonemes and less frequently in palatal and velar phonemes. Slower speech and low voice intensity improved tongue protrusion.\n\nAside from these abnormalities, during general physical and detailed neurological examination, the patient’s motor, sensory and cerebellar functions were all normal. Fundoscopic examination was unremarkable and no abnormality was detected on cranial nerve examination.\n\nSeveral investigations were done to rule out any obvious cause of secondary dystonia. A detailed past medical history had ruled out post traumatic, post drug induced or post infectious causes. All the patient’s baseline lab workup, including complete blood count, renal function tests, serum electrolytes, liver function tests, creatine phosphate kinase and thyroid profile were normal. Peripheral smear was done to rule out neuroacanthocytosis. Normal serum ceruloplasmin ruled out Wilson’s disease. Cerebral spinal fluid workup was unremarkable, and an electroencephalogram revealed a normal rhythm. An MRI brain with contrast was done on 1.5 tesla, which showed no apparent abnormality. After ruling out all likely causes, the patient was reevaluated by a psychologist and a psychiatrist after 2 weeks of her initial presentation to the OPD. Psychiatric testing using mental state examination and assessment for psychiatric symptoms revealed a normal psychiatric state and personality. The worsening of the patient’s dystonia with a past trial of antidepressant drugs, along with her current psychiatric assessment ruled out stress induced dystonia.\n\nThe patient responded well to treatment with Trihexiphenidyl 1mg twice daily (which was changed to 2mg twice daily on her follow-up visit after 1 month), and sensory tricks using chewing gum and pressing on her neck while trying to speak and during swallowing. A follow-up at 1 month with the speech therapist showed objective improvement in her condition; she was able to repeat words and small sentences more easily, her swallowing had drastically improved, and she was able to communicate effectively while using sensory tricks. The patient is still under treatment and is showing good prognosis.\n\n\nDiscussion\n\nThe etiology of dystonia is multifactorial. Drugs like antipsychotics, neuroleptics, dopamine agonists and antagonists, antiepileptic medication and calcium channel blockers are notorious for causing dystonia. Neurological disorders, including perinatal brain injury, Huntington’s Disease, Cerebral Palsy, Tourette’s syndrome, arteriovenous malformations, ischemia, autoimmune and paraneoplastic encephalitis and some tumors of the central nervous system can also lead to dystonia. Infectious diseases of the central nervous system, meningitis, encephalitis, HIV infection, tuberculosis and syphilis have been known to cause dystonia as well. Similarly, toxins, heavy metal poisoning, and genetic diseases, like Wilson’s disease all have the potential to cause dystonia1. In some cases, a psychogenic cause is usually found7.\n\nIn our case, even after an extensive and detailed workup along with multiple psychiatric assessments and speech assessment, no specific cause for the dystonia could be identified. Another case report featuring a case of idiopathic lingual dystonia stated that the dystonia was only speech induced8, whereas in our case dystonia and abnormal movements of the tongue were present at rest and also caused impairment in swallowing. Another case of isolated lingual dystonia attributed electrical injuries as the initiating cause8.\n\nDrugs like levodopa (750mg per day) and trihexyphenidyl (up to 10mg per day) have been reported to alleviate symptoms of dystonia. Anticholinergics are also sometimes prescribed for different dystonias9. In our patient, we preferred trihexyphenidyl over levodopa because of a safer side effect profile and cost effectiveness; our patient’s symptoms were only suppressed by trihexyphenidyl and sensory tricks. Botox injections has shown promising results in the treatment of oro-buccal-lingual dystonia, but its use in isolated lingual dystonia is unknown and requires further study6, as its use can cause dysphagia and lead to choking. Genetic testing could not be performed in our case to rule out any genetic causes or abnormalities as genetic testing was unavailable in our setup and in other medical laboratories nearby.\n\nIsolated lingual dystonia is a rare form of tongue specific movement disorder that warrants further investigations and has many differential diagnoses, which should be considered before a patient is labeled as a case of idiopathic dystonia.\n\n\nConsent\n\nWritten informed consent for the publication of this case report was obtained from the patient.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.",
"appendix": "References\n\nAlbanese A, Bhatia K, Bressman SB, et al.: Phenomenology and classification of dystonia: A consensus update. Mov Disord. 2013; 28(7): 863–73. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGn S, Nag A: Management of Oromandibular Dystonia: A Case Report and Literature Update. Case Rep Dent. 2017; 2017: 3514393. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSteeves TD, Day L, Dykeman J, et al.: The prevalence of primary dystonia: A systematic review and meta-analysis. Mov Disord. 2012; 27(14): 1789–96. PubMed Abstract | Publisher Full Text\n\nEpidemiological Study of Dystonia in Europe (ESDE) Collaborative Group: A prevalence study of primary dystonia in eight European countries. J Neurol. 2000; 247(10): 787–92. PubMed Abstract | Publisher Full Text\n\nDuffy PO, Butler AG, Hawthorne MR, et al.: The epidemiology of the primary dystonias in the north of England. Adv Neurol. 1998; 78: 121–5. PubMed Abstract\n\nEsper CD, Freeman A, Factor SA: Lingual protrusion dystonia: Frequency, etiology and botulinum toxin therapy. Parkinsonism Related Disorders. 2010; 16(7): 438–41. PubMed Abstract | Publisher Full Text\n\nBaik JS, Park JH, Kim JY: Primary lingual dystonia induced by speaking. Mov Disord. 2004; 19(10): 1251–2. PubMed Abstract | Publisher Full Text\n\nOndo W: Lingual dystonia following electrical injury. Mov Disord. 1997; 12(2): 253. PubMed Abstract | Publisher Full Text\n\nJankovic J: Medical treatment of dystonia. Mov Disord. 2013; 28(7): 1001–12. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "63779",
"date": "22 May 2020",
"name": "Syed Shahmeer Raza",
"expertise": [
"Reviewer Expertise Cerebral Oximetry",
"Clinical Surgical Research and Infectious Diseases"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper is very well written. It describes not only the care decisions but also the extensive literature review that was carried out to write the case report.\n\nThe data available and from the Discussion, I can conclude that the authors have put in a fine effort to draft a presentable case\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes",
"responses": []
},
{
"id": "66171",
"date": "06 Aug 2020",
"name": "Stephen Tisch",
"expertise": [
"Reviewer Expertise Movement disorders",
"dystonia",
"deep brain stimulation"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis report of isolated lingual dystonia could be improved. The description of the phenomenology can be improved, although overall the description of the case is convincing for isolated lingual dystonia. Lingual dystonia is not extremely rare and seldom reported as the authors state. A recent paper Yoshida et al. 2017, highlighted occupational risk factors in 95 patients with isolated, task-specific, speech induced, lingual dystonia.1 The lack of genetic testing is a weakness. There was a lack of references in some sections.\n\nIs the background of the case’s history and progression described in sufficient detail? Partly\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? Partly",
"responses": []
},
{
"id": "73169",
"date": "27 Oct 2020",
"name": "Angelo F Gigante",
"expertise": [
"Reviewer Expertise movement disorders",
"dystonia"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe report is well written, a quite exhaustive diagnostic workup was made. However, the authors should add more information on:\npossible episodes or triggers experienced before symptoms onset,\n\ntiming pattern of disease onset (subacute or acute?)\n\npossible fluctuating clinical course (any remission?) before the first visit at Rehman Medical Institute.\n\nIs the background of the case’s history and progression described in sufficient detail? Partly\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes",
"responses": []
},
{
"id": "68916",
"date": "16 Nov 2020",
"name": "Steven J Frucht",
"expertise": [
"Reviewer Expertise dystonia",
"movement disorders"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is.a nicely written single case report of lingual dystonia responsive to Artane. If there were accompanying videos, they would substantially increase the impact of the report. Nonetheless, I think that the paper is well written, the subject is of interest, and it is worth publishing in this format. I approve.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-314
|
https://f1000research.com/articles/9-1257/v1
|
19 Oct 20
|
{
"type": "Method Article",
"title": "The importance of software citation",
"authors": [
"Daniel S. Katz",
"Neil P. Chue Hong",
"Tim Clark",
"August Muench",
"Shelley Stall",
"Daina Bouquin",
"Matthew Cannon",
"Scott Edmunds",
"Telli Faez",
"Patricia Feeney",
"Martin Fenner",
"Michael Friedman",
"Gerry Grenier",
"Melissa Harrison",
"Joerg Heber",
"Adam Leary",
"Catriona MacCallum",
"Hollydawn Murray",
"Erika Pastrana",
"Katherine Perry",
"Douglas Schuster",
"Martina Stockhause",
"Jake Yeston",
"Neil P. Chue Hong",
"Tim Clark",
"August Muench",
"Shelley Stall",
"Daina Bouquin",
"Matthew Cannon",
"Scott Edmunds",
"Telli Faez",
"Patricia Feeney",
"Martin Fenner",
"Michael Friedman",
"Gerry Grenier",
"Melissa Harrison",
"Joerg Heber",
"Adam Leary",
"Catriona MacCallum",
"Hollydawn Murray",
"Erika Pastrana",
"Katherine Perry",
"Douglas Schuster",
"Martina Stockhause",
"Jake Yeston"
],
"abstract": "Software is as integral as a research paper, monograph, or dataset in terms of facilitating the full understanding and dissemination of research. This article provides broadly applicable guidance on software citation for the communities and institutions publishing academic journals and conference proceedings. We expect those communities and institutions to produce versions of this document with software examples and citation styles that are appropriate for their intended audience. This article (and those community-specific versions) are aimed at authors citing software, including software developed by the authors or by others. We also include brief instructions on how software can be made citable, directing readers to more comprehensive guidance published elsewhere. The guidance presented in this article helps to support proper attribution and credit, reproducibility, collaboration and reuse, and encourages building on the work of others to further research.",
"keywords": [
"Software citation",
"publishing",
"scholarly communication",
"guidelines",
"bibliometrics"
],
"content": "\n\nSoftware is as integral as a research paper, monograph, or dataset in terms of facilitating the full understanding and dissemination of research. Books and journal articles have long benefited from an infrastructure that makes them easy to cite, a key element in the process of research and academic discourse in all disciplines. We believe that software (including computational code, scripts, models, notebooks and libraries) should be cited in the same way that other sources of information, such as articles and books, are cited.\n\nCiting software helps further research and provides the means for other researchers to access software in order to:\n\nsupport proper attribution and credit (similar to that of papers, data, etc.);\n\nenable peer-review, validation, and reproducibility of findings;\n\nsupport collaboration and reuse; and\n\nencourage building on the work of others.\n\nSoftware citation elevates software to the level of a first-class object in the digital scholarly ecosystem, consistent with its immense actual present-day significance.\n\nFORCE11 has been developing guidance for software citation. The Software Citation Principles (Smith et al., 2016) were written to encourage broad adoption of a consistent policy for software citation across disciplines and venues. The Software Citation Checklist for Authors (Chue Hong et al., 2019a) and Software Citation Checklist for Developers (Chue Hong et al., 2019b) provide more practical information for those seeking to improve their practice.\n\n\nSoftware citation essentials\n\nThis article is aimed at authors citing software. This includes software developed by others, as well as software developed by any or all of the authors. Making software citable is a critical developer-led step, which is briefly detailed in the next subsection, \"Making Software Citable\".\n\nThe use of persistent identifiers (PIDs) and core descriptive metadata are essential elements of software citation. This is because they are the mechanism used to index and track citations. We recognise that the challenges associated with software deposit and publication vary across disciplines, and we encourage research communities to develop citation systems that work well for them. We also recognise that the citation style formats used vary between disciplines and journals. Independent of the style of any citation, we recommend certain essential metadata elements should always be captured.\n\nThere are multiple use cases for citing software. We recommend citing the specific version used (and the authors and publication date for that version) if you used it directly in the research described in your publication (e.g., the Methods section). We recommend citing the software concept (project) if you are referencing the software elsewhere in your paper.\n\nOur recommended format for software citation is to ensure the following information is provided as part of the reference:\n\nCreator(s): the authors or project that developed the software.\n\nTitle: the name of the software.\n\nPublication venue: the publication venue of the software, preferentially, an archive or repository that provides persistent identifiers.\n\nDate: the date the software was published.\n\nIdentifier: a resolvable pointer to the software, preferentially, a PID that resolves to a landing page containing descriptive metadata about the software, similar to how a Digital Object Identifier (DOI) for a paper that points to a page about the paper rather than directly to a representation of the paper, such as the PDF. DOIs are preferable, and other examples of PIDs include Handles, RRIDs, ASCL IDs, swMath IDs, Software Heritage IDs, ARKs, etc. If there is no PID for the software, a URL to where the software exists may be the best identifier available.\n\nIt may also be desirable, and depending upon the publisher, may be required, to include information about two optional properties (as appropriate):\n\nVersion: the identifier for the version of the software being referenced. If the version is unidentified or unknown, the date of access should be used.\n\nType: some citation styles (e.g., APA), require a bracketed description of the citation (e.g., Computer software) to be included.\n\nIf a published article exists that describes the software, it should be cited as an additional reference.\n\n\nMaking software citable\n\nAuthors should consult the Software Citation Checklist for Developers (Chue Hong et al., 2019b) for information on how to obtain a PID or choose a software license for software they have developed. That document contains a set of steps that developers can take to ensure that they are following good practices. We strongly recommend that journals provide such information to their authors, either by referring to that document, or using text from it or similar text. Example guidance would include instructing authors to version their software, choose a license for their software, perhaps by linking to the information at choosealicense.org, record metadata about the software as part of the repository, deposit their software in a preservation repository that provides a PID, and advertise the recommended citation in the repository. In particular, guidance should explicitly mention that Creative Commons licenses (including CC-BY) must not be used for software, and an open source license should be used.\n\n\nSoftware citation examples\n\nThe following examples show how software can be cited in one common citation style, APA. The general format for downloaded software, from Section 10.10 of (2020) Publication Manual of the American Psychological Association (Seventh Edition) is:\n\nDeveloper, A. A., Developer, B. B., & Developer, C. C. (yyyy)1. Title of the software: Subtitle (Version #.#)2 [Computer software]3. Publisher4, https://URL5\n\nIf no version number or version string exists, we (the FORCE11 Software Citation Implementation Working Group) modify this to:\n\nDeveloper, A. A., Developer, B. B., & Developer, C. C. (yyyy). Title of the software: Subtitle [Computer software]. Archive Name. Retrieved Month dd, yyyy, from https://URL\n\nThe following are examples of software citations.\n\nIdeal citations to the specific version of the software, where all recommended information is present (the first demonstrates a large author list; the second demonstrates a project team as the author):\n\nCoon, E., Berndt, M., Jan, A., Svyatsky, D., Atchley, A., Kikinzon, E., Harp, D., Manzini, G., Shelef, E., Lipnikov, K., Garimella, R., Xu, C., Moulton, D., Karra, S., Painter, S., Jafarov, E., & Molins, S. (2020, March 25). Advanced Terrestrial Simulator (ATS) v0.88 (Version 0.88) [Computer software]. Zenodo. https://doi.org/10.5281/zenodo.3727209\n\nLab For Exosphere And Near Space Environment Studies. (2019, March 20). lenses-lab/LYAO_RT-2018JA026426: Original Release (Version 1.0.0) [Computer software]. Zenodo. https://doi.org/10.5281/zenodo.2598836\n\nCitation referencing software that is preserved in a software archive (e.g. Software Heritage)6:\n\nDelebecque, F., Gomez, C., Goursat, M., Nikoukhah, R., Steer, S., & Chancelier, J.-P. (1994). Scilab (Version 1.1) [Computer software]. Software Heritage, swh:1:dir:1ba0b67b5d0c8f10961d878d91ae9d6e499d746a;origin=https://hal.archives-ouvertes.fr/hal-02090402\n\nDi Cosmo, R. & Danelutto, M. (2020). The Parmap library: Core mapping routine (Version 1.1.1) [Computer software]. Software Heritage, swh:1:cnt:43a6b232768017b03da934ba22d9cc3f2726a6c5;lines=192-228;origin=https://github.com/rdicosmo/parmap\n\nA citation for software that does not have a PID but does have a version and identifier (URL), where authorship is assigned to the project as a whole:\n\nDataverse Project (2020). Dataverse (Version 4.20) [Computer software] https://github.com/IQSS/dataverse/releases/tag/v4.20\n\nA citation for software where there is no version identified and where the publishing date is unknown:\n\nThomas, J. & Daujotas, G.7 (n.d.). is-thirteen [Computer software]. GitHub. Retrieved June 17, 2020 from https://github.com/jezen/is-thirteen\n\nA citation for a software concept (all versions):\n\nBLAS team (n.d.), BLAS (Basic Linear Algebra Subprograms) [Computer software]. Netlib. http://www.netlib.org/blas/\n\nA citation for software where little information is available, perhaps where only the executable program is available. For commercial software, a link to information about availability for purchase is helpful, as shown in the example below.\n\nIBM Corp. (2017). IBM SPSS Statistics for Windows (Version 25.0) [Computer software]. IBM Corp. https://www.ibm.com/products/spss-statistics\n\nTwo examples of how the citations above would be referenced in the text of a paper according to APA style8, the first in the methodology section and the second in a related work section:\n\nWe used version 0.88 of Advanced Terrestrial Simulator (Coon et al., 2019) and version 25.0 of IBM SPSS Statistics for Windows (IBM Corp., 2017) to carry out the analysis of the data in this paper.\n\nIn the field of bibliometrics, a different approach is taken by BLAS (BLAS team, n.d).\n\n\nUsage notes\n\nThis document provides generic guidance about software citation for the communities and institutions publishing academic journals and conference proceedings. We expect those communities and institutions to produce different versions of this document with software examples and citation styles that are appropriate for their intended audience. We request that those documents refer back to (or cite) this one. This document can be cited (in APA 7th Ed. style) as:\n\nKatz, D. S., Chue Hong, N. P., Clark T., Muench, A., Stall, S., Bouquin, D., Cannon, M., Edmunds, S., Faez, T., Farmer, R., Feeney, P., Fenner, M., Friedman, M., Grenier, G., Harrison, M., Heber, J., Leary, A., MacCallum, C., Murray, H., … Yeston, J. (2020) The importance of software citation. F1000 Research. https://doi.org/10.12688/f1000research.26932.1\n\nHardware is important, but we have initially chosen not to overload software citations with hardware requirements directly. This might be better done through linkage between DOIs.\n\n\nData availability\n\nNo data is associated with the article.",
"appendix": "Acknowledgements\n\nThis article is based in part on data citation guidance published by DataCite (Datacite), and on related publications from FORCE11 working groups (Cousijn et al., 2018; Fenner et al., 2019). It was initially drafted by Neil Chue Hong, and further developed by Daniel S. Katz, Neil Chue Hong, Tim Clark, August Muench, and Shelley Stall, along with many participants in the FORCE11 Software Citation Implementation Working Group’s Journals Task Force. We also acknowledge useful advice from Kevin Swanson, Taylor & Francis.\n\n\nFootnotes\n\n1The year is required, or “n.d.” if not identifiable.\n\n2The version is optional but preferred. Note that the version may be a token/string that is not a semantic version (https://semver.org/) and that must be exactly preserved, such as a commit hash (e.g., a149dbc00fe8b0e8260f7c2d39c77692683e7fa4), a semi-numeric tagged release (e.g., v0.4-alpha01), or date string (e.g., 2020-02-20).\n\n3APA style includes additional information that is helpful for software citation (e.g. it requires the [Computer software] bracketed description). Although this is not part of our guidance above, we recommend following APA style and including these elements. Other styles may not use this extra information.\n\n4If the software is downloaded or if the developer is the same as the publisher, the publisher name is omitted.\n\n5In APA style, the URL is used for both URLs and DOIs or other PIDs, e.g., a DOI is expressed as https://doi.org/DOI.\n\n6This example is analogous to citing the preserved version of a webpage on archive.org, rather than the webpage directly.\n\n7The README for the is-thirteen software says “A helpful tool by Jezen Thomas with helpful help from Gytis Daujotas and many fine folk.”; therefore our citation tries to take the developers intentions around authorship into account.\n\n8American Psychological Association. (2020). Publication manual of the American Psychological Association (7th ed.). American Psychological Association. https://doi.org/10.1037/0000165-000\n\n\nReferences\n\nChue Hong NP, Allen A, Gonzalez-Beltran A, et al.: Software Citation Checklist for Authors (Version 0.9.0). Zenodo. 2019a. Publisher Full Text\n\nChue Hong NP, Allen A, Gonzalez-Beltran A, et al.: Software Citation Checklist for Developers (Version 0.9.0). Zenodo. 2019b. Publisher Full Text\n\nCousijn H, Kenall A, Ganley E, et al.: A data citation roadmap for scientific publishers. Sci Data. 2018; 5: 180259. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDataCite: DataCite - Cite Your Data. Reference Source\n\nFenner M, Crosas M, Grethe JS, et al.: A data citation roadmap for scholarly data repositories. Sci Data. 2019; 6(1): 28. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmith AM, Katz DS, Niemeyer KE, et al.: Software Citation Principles. PeerJ Computer Science. 2016; 2: e86. Publisher Full Text"
}
|
[
{
"id": "73368",
"date": "17 Nov 2020",
"name": "Ludo Waltman",
"expertise": [
"Reviewer Expertise Scientometrics",
"quantitative science studies",
"open science"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a very useful contribution. I have some minor comments.\nDiscussions about software citation and data citation are closely related. I would therefore find it helpful to read something about the way in which the guidance on software citation provided in this document relates to standards for data citation. It seems important that standards for software citation and data citation are consistent as much as possible.\nThe title of the contribution (“The importance of software citation”) suggests that the contribution focuses on arguing for the importance of software citation. However, as explained in the abstract, the focus in fact is on providing “broadly applicable guidance on software citation”. My suggestion therefore is to revise the title. An alternative title for instance could be “How to cite software?”.\n“We recommend citing the specific version used (and the authors and publication date for that version) if you used it directly in the research described in your publication (e.g., the Methods section). We recommend citing the software concept (project) if you are referencing the software elsewhere in your paper.”: I don’t fully understand the distinction that is made in these two sentences. The authors seem to have in mind a distinction between citing software because it is used directly in a research project and citing software for other reasons. I would like to know more about what other reasons for citing software the authors have in mind and why they believe citations should be made in different ways in the two situations they distinguish.\n“If a published article exists that describes the software, it should be cited as an additional reference.”: The motivation for this recommendation is not clear to me. The authors seem to give special treatment to published articles, by which I assume they have in mind articles published in scholarly journals. I find this questionable. Suppose we have two pieces of software. Software A is documented in a two-page article published in a scholarly journal. Software B is documented in a comprehensive report made available in GitHub. Why should the article documenting software A be cited, while the report documenting software B does not need to be cited? Note that the article documenting software A probably cannot be updated, and the article is therefore likely to provide an outdated description of the software. The report documenting software B can be updated and therefore is likely to offer an up-to-date description of the software.\n“Hardware is important, but we have initially chosen not to overload software citations with hardware requirements directly. This might be better done through linkage between DOIs.”: I don’t understand these two sentences. Some additional explanation would be helpful.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "6170",
"date": "11 Dec 2020",
"name": "Teresa Gomez-Diaz",
"role": "Reader Comment",
"response": "In order to contribute to this interesting scientific discussion, we would like to point out some other aspects that could be considered. This is an interesting work, both the article and the report, as it contributes to a sounder installation of best practices to be adopted by scientific communities in order to reference and cite research outputs other than articles. The section \"Software citation essentials\" mentions: The use of persistent identifiers (PIDs) and core descriptive metadata are essential elements of software citation. This is because they are the mechanism used to index and track citations. We would like to ask for further explanation and references in order to better understand this important mechanism. Date: the date the software was published. This date point is an interesting issue, as much of the research software used and produced in the scientific communities has not gone through a thorough publication process involving some review procedure. They have been disseminated in a web page, a forge or a deposit like Zenodo or GitHub, usually associating a date to the disseminated version of the software. On the other hand, recent journals in the scientific publishing world have been created to publish software papers, which establishes a publication date. The authors could provide further explanation on the kind of date of publication that is considered in their recommendations. Furthermore, the following recent work: Gomez-Diaz T. and Recio T. On the evaluation of research software: the CDUR procedure [version 2; peer review: 2 approved]. F1000Research 2019, 8:1353 (https://doi.org/10.12688/f1000research.19994.2) studies referencing and citation issues in the context of research software (see the section 2.5 of \"On the evaluation of research software: the CDUR procedure\") and it could be interesting for the authors of the document we are commenting here to compare both approaches. These comments have been prepared with the second author of this publication (T. Recio)."
},
{
"c_id": "6245",
"date": "05 Jan 2021",
"name": "Daniel S. Katz",
"role": "Author Response",
"response": "Thank you for your comments. We have just submitted a revised version that adds some additional description to explain the item about the software's publication date, as you requested. Regarding your second point, suggesting that discuss the recent CDUR work, we believe this paper has a fairly narrow focus, and that a future expanded or follow-on paper would be a better place to compare with that work, as well as much other related work."
},
{
"c_id": "6246",
"date": "05 Jan 2021",
"name": "Daniel S. Katz",
"role": "Author Response",
"response": "Thank you very much for your careful reading and useful comments and suggestions. We have just submitted a revised version of the paper, which has the following changes made in response: Discussions about software citation and data citation are closely related. I would therefore find it helpful to read something about the way in which the guidance on software citation provided in this document relates to standards for data citation. It seems important that standards for software citation and data citation are consistent as much as possible. We've added a sentence at the end of this paragraph to recognize the connection to work on data citation, and to point readers to references for more information. The title of the contribution (“The importance of software citation”) suggests that the contribution focuses on arguing for the importance of software citation. However, as explained in the abstract, the focus in fact is on providing “broadly applicable guidance on software citation”. My suggestion therefore is to revise the title. An alternative title for instance could be “How to cite software?”. We agree with this comment, and have changed the title in response. “We recommend citing the specific version used (and the authors and publication date for that version) if you used it directly in the research described in your publication (e.g., the Methods section). We recommend citing the software concept (project) if you are referencing the software elsewhere in your paper.”: I don’t fully understand the distinction that is made in these two sentences. The authors seem to have in mind a distinction between citing software because it is used directly in a research project and citing software for other reasons. I would like to know more about what other reasons for citing software the authors have in mind and why they believe citations should be made in different ways in the two situations they distinguish. We agree that this was not clear as written, and have rewritten these sentences. “If a published article exists that describes the software, it should be cited as an additional reference.”: The motivation for this recommendation is not clear to me. The authors seem to give special treatment to published articles, by which I assume they have in mind articles published in scholarly journals. I find this questionable. Suppose we have two pieces of software. Software A is documented in a two-page article published in a scholarly journal. Software B is documented in a comprehensive report made available in GitHub. Why should the article documenting software A be cited, while the report documenting software B does not need to be cited? Note that the article documenting software A probably cannot be updated, and the article is therefore likely to provide an outdated description of the software. The report documenting software B can be updated and therefore is likely to offer an up-to-date description of the software. We have removed \"published\", as this was not an important part of the point we were trying to make, and have adjusted the text to make our point more clearly. “Hardware is important, but we have initially chosen not to overload software citations with hardware requirements directly. This might be better done through linkage between DOIs.”: I don’t understand these two sentences. Some additional explanation would be helpful This text made sense in a much earlier version of the paper, but now we agree that this point was confusing as written and also feel it is not important to the article, so we have removed it."
}
]
},
{
"id": "75320",
"date": "07 Dec 2020",
"name": "Gianmaria Silvello",
"expertise": [
"Reviewer Expertise Databases",
"Data citation",
"Information retrieval and Digital Libraries"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper presents an overview of software citation detailing state of the art and providing some indications about how software should be cited in different contexts. There is no innovative method presented, but rather this is a set of community-driven guidelines that can be really useful as a starting point to provide adequate software citations. I am not sure this paper is a good fit as a method article for this journal, but it is a good fit for the journal. To me, it can be something in-between a method article and an opinion paper. What can be considered missing from this paper are considerations or references to transitive citations, which are central for software citation. Nevertheless, this topic may be out of scope for a contribution like this one. Anyway, the paper is well-written, and to me, it can be published as-is provided that it is clear that no major innovative contribution is described or new insight is presented.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "6247",
"date": "05 Jan 2021",
"name": "Daniel S. Katz",
"role": "Author Response",
"response": "Thank you for your review and suggestions. In our newly submitted revision, we considered your point What can be considered missing from this paper are considerations or references to transitive citations, which are central for software citation. Nevertheless, this topic may be out of scope for a contribution like this one. but we feel this is indeed out of scope for this paper."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1257
|
https://f1000research.com/articles/10-18/v1
|
12 Jan 21
|
{
"type": "Research Article",
"title": "Optimization of catheter placement for convection-enhanced delivery to brain tumors",
"authors": [
"Lisa H. Antoine",
"Roy P. Koomullil",
"Timothy M. Wick",
"Arie Nakhmani",
"Roy P. Koomullil",
"Timothy M. Wick",
"Arie Nakhmani"
],
"abstract": "Background: Recent trends suggest that physicians will diagnose thousands of children in the United States with a brain or central nervous system tumor in 2020. Malignant brain tumors are difficult to treat, with low life expectancy rates in children and adults. Convection-enhanced delivery (CED) shows promise for the treatment of brain tumors, yet remains in clinical trials despite being developed more than 20 years ago. To advance CED to standard of care status and help improve survival rates, this study group developed a quantitative computer simulation model to determine and optimize therapy distribution in brain tumors based on the catheter infusion locations for CED. Methods: The simulations resulted in the identification of four infusion reference locations, which were used to conduct an optimization study to identify the optimal locations for CED. Patient-specific T1-weighted images and diffusion-weighted images provided information regarding tumor shape and size and the approximate rate at which therapy distributes at spatial locations within the tumor. Using the images, the researchers in this study developed a model which allowed the calculation of therapy distribution within the tumor while considering its permeability, porosity, and interstitial fluid pressure characteristics. We divided the tumor into regions and calculated distribution for four infusion locations per region. Using the location from each region with the highest volume distribution allowed our study group to conduct the response surface optimization. Results: Twelve optimal locations emerged from the optimization with volume percentage distributions ranging from 7.92% to 9.09%, compared to 2.87% to 6.32% coverage for non-optimal locations. This optimization method improved distribution from 27.80% to 45.95%, which may improve therapeutic value. Conclusions: Catheter placement appears to influence volume therapy distribution percentages. The selection of the highest percentages per region may provide optimal therapy for the entire tumor region.",
"keywords": [
"brain tumor",
"convection-enhanced delivery",
"optimization"
],
"content": "Introduction\n\nSurgery, radiation therapy, and chemotherapy are the standard of care for brain tumors1. Surgery or tumor resection removes all or part of the tumor; however, this method is invasive and risky. Radiation therapy seeks to remove the tumor using high-energy x-rays, while chemotherapy attacks the tumor using drugs taken either orally or intravenously. Radiation therapy and chemotherapy may cause side effects, which may include hair loss and decreased mental functionality1.\n\nAlthough therapy delivery methods, such as Ommaya reservoirs and convection-enhanced delivery (CED), are not standards of care, these methods are available in clinical trials, appear to be less invasive, and could offer effective treatment2–5. Ommaya reservoirs use a catheter-reservoir system implanted under the scalp for therapy delivery, while CED uses a catheter-syringe-pump system for therapy delivery. In this study, researchers used a response surface optimization (RSO) technique for the selection of the ideal catheter placement locations, which may improve the effectiveness of CED.\n\nIn general, optimization techniques improve speed and accuracy during the search for the ideal solution for a given problem. Study groups used an array of these techniques to address the challenges associated with brain tumor treatments6–10. Li et al. developed an optimization method using a genetic algorithm to determine catheter positions that would yield optimal therapy concentrations in the tumor and white matter, while limiting concentrations in the cortex6. The study demonstrated that a genetic algorithm can determine optimal catheter positions and suggests that it may be sufficient to determine the optimal locations for our study. Vidotto et al. used a neural network based algorithm to improve the accuracy and speed of axon diameter distribution within the brain7.\n\nZhang et al. used RSO to determine the optimal extrusion die design, which resulted in a more accurate extrudate shape and lower extrusion die production costs11. Using a response surface methodology Sultana et al. predicted that jute fiber can improve compressive and tensile concrete strength12. This group validated the predicted results against experimental results and concluded that the variation was only approximately 5%. Akhbar et al. used RSO to optimize rotational speed and feed to prevent heat damage during bone-drilling surgical procedures13. Albe et al. examined the effect of pH and salt concentration on the extraction of sunflower protein using the response surface to maximize the protein14. Silva et al. used RSO to achieve a mean recovery of 91% and overall precision of 5% when verified using fortified samples of pharmaceutical residues15. Each of these groups used RSO and improved the accuracy of their respective designs, which suggests that this technique will optimize the infusion locations in our study within 5% precision.\n\nIn this study RSO allowed an efficient determination of the optimal location for therapy distribution and includes design of experiments (DOE), response surface, and optimization. DOE uses input and output parameters as the major components and analyzes the input-output relationship generating design points that lie within boundaries in proximity of the baseline point. The response surface results from an algorithm that does not simulate a complete solution, but instead performs a quick analysis of the DOE sampling percentages using cross-validation to predict surface percentages. The fitness function associates DOE percentages to surface percentages, where alignment of the percentages on a diagonal line suggests that the response surface fits the DOE percentages. Using results from DOE and response surface, optimization determines the ideal locations for the maximum therapy distribution.\n\n\nMethods\n\nTo conduct response surface optimization, the researchers derived the tumor geometry with x, y, z lengths of 12.86 mm, 30.4 mm, and 37.43 mm, respectively, and volume of 5031.35 mm3 from patient specific T1-weighted images (T1W)16. The anonymized T1W (approved for use by the University of Alabama Institutional Review Board as not human-subjects research) includes 135 dicom files for one subject were obtained from The University of Alabama at Birmingham Department of Radiology. Diffusion tensors inside the tumor resulted from the diffusion-weighted images (DWI) of the patient using the DSI Studio 2019_10 version17. Viscous resistance of the tumor is the transformation of the diffusion tensor as discussed by the Støverud group18. Because the tensors may not align to the center of the computational elements, it was necessary to interpolate these values using trilinear interpolation.\n\nWe expect the suspension of the oncolytic herpes simplex virus to be uniform in a fluid and use a transport equation to model fluid flow inside the tumor19. Velocity components in the transport equation are a function of the therapy infusion rate and thereby, calculated by solving the Navier-Stokes equation19 with an assumption of laminar flow inside the tumor. The virus density and viscosity information were unavailable. Therefore, we analyzed the flow using density and viscosity values for water of 1×10-6 kg/mm3 and 1.003×10-6 kg/mm-s, and a porosity value of 0.6 and pressure of 266.65 Pa from previous studies20,21.\n\nThe researchers selected six input parameters, which included x, y, and z coordinates along the tumor boundaries and fixed values for flow rate (0.1 mL/hr), minimum dose (1×10-19 mL/mm3), and radius (1.5 mm). The radius of 1.5 mm ensures that the infusion location is larger than the catheter radius (0.5 mm). After the selection and analysis of 16 locations (x, y, z coordinates) within the tumor, four baseline locations were available for the response surface optimization. The baselines represent the locations that produced the highest therapy distribution percentages in the tumor.\n\nDOE generated 28 sampling points using the input parameters of the baseline point as the reference point. Default lower and upper bounds for the coordinates reflect the baseline point as the midpoint. ANSYS (free student version download available) calculated therapy percentages for sampling points using default parameters – flow rate, infusion location radius, and minimum dose19. Based on the DOE, the response surface determines minimum and maximum percentages. The goodness of fit compared the predicted response surface to the observed surface for the design points and determined the surface quality and discrepancies between observed and predicted percentages.\n\nNonlinear Programming by Quadratic Lagrangian (NLPQL)19 provided a local optimization solution, which maximizes the therapy distribution percentages. Baseline point specification was necessary to define the exploration region and identify three candidate points per region. Number of evaluations varied by region and included several iterations. Tolerance for checking convergence was 0.0001.\n\n\nResults\n\nVolume distribution percentages from region one design points ranged from 3.01 % to 9.02%, region two ranged from 4.19% to 9.09%, region three ranged from 2.51% to 8.12%, and region four ranged from 3.17% to 8.18%.\n\nResponse surfaces in region one suggest that maximum therapy distribution percentage is approximately 9% (Figure 1). Location combinations of y and z (Figure 1c) appear to produce a larger region with percentages that are greater than 8% as opposed to combinations of locations x and y (Figure 1a) and locations x and z (Figure 1b). Maximum therapy distribution for region one occurred at location -13.28 mm, 23.80 mm, 39.77 mm and was approximately 9.02%.\n\n(a) x axis represents input parameters along the x direction, y axis represents input parameters along the y direction; (b) z axis represents input parameters along the z direction; (c) y and z input combinations appear to include the highest number of locations with therapy distribution percentages greater than 8%.\n\nResponse surfaces in region two suggest that the therapy distribution percentage is approximately 9% (Figure 2). Location combinations of y and z (Figure 2c) appear to produce a larger region with percentages that are greater than 8% when compared to combinations of locations x and y (Figure 2a) and x and z (Figure 2b). Maximum therapy distribution for region two occurred at location -11.04 mm, 15.17 mm, 26.68 mm and was approximately 9.09%.\n\n(a) x axis represents input parameters along the x direction, y axis represents input parameters along the y direction; (b) z axis represents input parameters along the z direction; (c) y and z input combinations appear to include the highest number of locations with therapy distribution percentages greater than 8%.\n\nResponse surfaces in region three suggest that maximum therapy distribution percentage is approximately 8% (Figure 3). Location combinations of y and z (Figure 3c) appear to produce a larger region with percentages that are greater than 7% as opposed to combinations of locations x and y (Figure 3a) and locations x and z (Figure 3b). Maximum therapy distribution for region three occurred at location -9.17 mm, 6.81 mm, 31.20 mm and was approximately 8.12%.\n\n(a) x axis represents input parameters along the x direction, y axis represents input parameters along the y direction; (b) z axis represents input parameters along the z direction; (c) y and z input combinations appear to include the highest number of locations with therapy distribution percentages greater than 7%.\n\nResponse surfaces in region four suggest that maximum therapy distribution percentage is approximately 8% (Figure 4). Location combinations of x and z (Figure 4b) appear to produce a larger region with percentages that are greater than 7% as opposed to x and y (Figure 4a) or y and z (Figure 4c). Maximum therapy distribution for region four occurred at location -12.89 mm, 17.93 mm, 48.80 mm and was approximately 8.27%.\n\n(a) x axis represents input parameters along the x direction, y axis represents input parameters along the y direction; (b) z axis represents input parameters along the z direction, x and z input combinations appear to include the highest number of locations with therapy distribution percentages greater than 7%; (c) y and z input combinations appear to include high numbers of locations with percentages greater than 7%.\n\nDuring optimization the goal was to maximize the volume distribution percentage to a value greater than or equal to 8% based upon the observed percentages from the design points. The number of evaluations varied among the regions with region one and two requiring 22 and 47 evaluations, respectively. Regions three and four required 39 and 59 evaluations, respectively.\n\nOptimization resulted in twelve optimal locations – three locations for each region. Maximum therapy distribution percentages throughout the regions ranged from 8.12% to 9.09% (Table 1). The location of the largest therapy distribution of 9.02% in region one was -13.28 mm, 23.80 mm, 39.77 mm, while the location in region two with a distribution of 9.09% was -11.04 mm, 15.17 mm, 26.68 mm. In regions three and four the largest distributions were 8.12% at location -9.17 mm, 6.81 mm, 31.20 mm and 8.27% at -12.89 mm, 17.94 mm, 48.80 mm, respectively. Optimal locations clustered together (Figure 5) and clustered closely to the baseline locations for region three.\n\n\nDiscussion\n\nMore than 25 years ago Bobo et al. introduced CED to deliver therapy to treat brain tumors; however, CED currently remains in clinical trials and appears to be mostly tested for safety and toxicity22,23. Shi et al. reviewed CED clinical treatment methods from 1997 through 2018 and found that irrespective of therapeutic agents, survival rates during phase III clinical trials did not improve24. Median survival times for the Carpenter et al., Weaver et al., Sampson et al., Hau et al., and Desjardins et al. trials were 7, 5.2, 7.5, 11, and 12.5 months, respectively25–29. These survival times may correlate to the difficulty in reaching all tumor cells and in evaluating the movement of therapy within the tumor30. Shi et al. suggested that catheter placement and corresponding therapy distribution may be the missing link.\n\nIn an effort to promote CED to standard of care, the current study aimed to quantitatively identify the optimal catheter locations for therapy delivery to brain tumors. The optimization technique requires one baseline location from each region with the highest therapy distribution. This group randomly selected 16 locations to determine the four baselines.\n\nResponse surfaces revealed 2D regions that show maximum therapy distribution. Both the predicted percentages from the response surface and observed percentages from the design of experiments aligned closely, which suggest that the response surface fits the design points. NLPQL approximated derivatives using the central differences scheme to refine the results and produce three locations per region with the maximum therapy percentages.\n\nThe mean percentage of the locations was 8.82% for region one with location -13.28 mm, 23.80 mm, 39.77 mm, yielding the highest percentage at 9.02%, while the baseline location (-12.07 mm, 26.45 mm, 44.19 mm) yielded therapy distribution of 6.26%, which varies from the mean distribution percentage by 29.05%. The mean Euclidean distance between the baseline and the three locations is 2.05 mm.\n\nThe mean percentage was 9.05% for region two with location -11.04 mm, 15.17 mm, 26.68 mm yielding the highest percentage at 9.09%, which contrasts with the baseline location (-10.03 mm, 13.79 mm, 24.25 mm) and its yield of 6.32%. Although the Euclidean distance between the two locations is 2.97 mm, the baseline provides 30.47% less therapy distribution than this location, which results from the optimization.\n\nLocation (-9.17 mm, 6.81 mm, 31.20 mm) for region three returned a distribution percentage of 8.12%, which varies by 34.73% from the baseline location (-8.73 mm, 6.48 mm, 29.71 mm) at 5.30%. The Euclidean distance between the locations is 1.59 mm. Location (-12.89 mm, 17.94 mm, 48.80 mm) for region four produced an optimal therapy distribution percentage of 8.27%, while the baseline location (-11.72 mm, 19.76 mm, 54.22 mm) returned therapy distribution of 4.47%, which varies by 45.95% from this location. Euclidean distance between the two locations is 5.84 mm.\n\n\nConclusions\n\nIn summary, CED as a brain tumor treatment method shows promise, but the treatment still remains in clinical trials despite being available for more than 20 years. The missing link for advancement to standard of care for CED may be the selection of catheter placement that yields the highest distribution percentage.\n\nWe developed an optimization method to determine the optimal catheter placement based on diffusion, viscous resistance, porosity, fluid pressure, and tumor geometry. Our results suggest that catheter placement does influence volume distribution percentages and that this optimization method improved distribution from 27.80% to 45.95%. The selection of the highest percentages per region may provide optimal therapy for the tumor.\n\nIn the future, investigators can verify this optimization method using patient-specific T1W and DWI data and therapy-specific density and viscosity information. Experimental validation using animal models is a critical step in determining the effectiveness of this quantitative tool31–39.\n\n\nData availability\n\nHarvard Dataverse: Catheter placement selection for convection-enhanced delivery of therapeutic agents to brain tumors. https://doi.org/10.7910/DVN/H7C6A216.\n\n- The dataset includes T1-weighted and diffusion-weighted images for the study subject.\n\nData are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).",
"appendix": "References\n\nBrain Tumor - Types of Treatment 2020. Accessed June 9, 2020. Reference Source\n\nPhase II Study of Intraventricular Methotrexate in Children With Recurrent or Progressive Malignant Brain Tumors. 2020. Accessed June 9, 2020. Reference Source\n\nMethotrexate and Etoposide Infusions Into the Fourth Ventricle in Children With Recurrent Posterior Fossa Brain Tumors. Accessed June 9, 2020. Reference Source\n\nConvection-Enhanced Delivery (CED) of MDNA55 in Adults With Recurrent or Progressive Glioblastoma. Accessed June 9, 2020. Reference Source\n\nA Phase 1 Study of M032, a Genetically Engineered HSV-1 Expressing IL-12, in Patients with Recurrent/Progressive Glioblastoma Multiforme, Anaplastic Astrocytoma, or Gliosarcoma. Accessed June 9, 2020. Reference Source\n\nLi D, Ivanchenko O, Sindhwani N, et al.: Optimal catheter placement for chemotherapy. 20th European symposium on computer aided process engineering. Comput Aided Chem Eng. 2010; 28: 223–228. Publisher Full Text\n\nVidotto M, De Momi E, Gazzara M, et al.: FCNN-based axon segmentation for convection-enhanced delivery optimization. Int J Comput Assist Radiol Surg. 2019; 14(3): 493–499. PubMed Abstract | Publisher Full Text\n\nHajishamsaei M, Pishevar A, Bavi O, et al.: A novel in silico platform for a fully automatic personalized brain tumor growth. Magn Reson Imaging. 2020; 68: 121–126. PubMed Abstract | Publisher Full Text\n\nGirard A, Saint-Jalmes H, Chaboub N, et al.: Optimization of time frame binning for FDOPA uptake quantification in glioma. PLoS One. 2020; 15(4): e0232141. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZade AE, Haghighi SS, Soltani M: Reinforcement learning for optimal scheduling of Glioblastoma treatment with Temozolomide. Comput Methods Programs Biomed. 2020; 193: 105443. PubMed Abstract | Publisher Full Text\n\nZhang G, Huang X, Li S, et al.: Optimized design method for profile extrusion die based on NURBS modeling. Fibers Polym. 2019; 20: 1733–1741. Publisher Full Text\n\nSultana N, Hossain SMZ, Alam MS, et al.: An experimental investigation and modeling approach of response surface methodology coupled with crow search algorithm for optimizing the properties of jute fiber reinforced concrete. Constr Build Mater. 2020; 243: 118216. Publisher Full Text\n\nAkhbar MFA, Yusoff AR: Fast & Injurious: Reducing thermal osteonecrosis regions in the drilling of human bone with multi-objective optimization. Measurement. 2020; 152: 107385. Publisher Full Text\n\nAlbe Slabi S, Mathe C, Basselin M, et al.: Multi-objective optimization of solid/liquid extraction of total sunflower proteins from cold press meal. Food Chem. 2020; 317: 126423. PubMed Abstract | Publisher Full Text\n\nSilva JM, Azcarate FJ, Knobel G, et al.: Multiple response optimization of a QuEChERS extraction and HPLC analysis of diclazuril, nicarbazin and lasalocid in chicken liver. Food Chem. 2020; 311: 126014. PubMed Abstract | Publisher Full Text\n\nAntoine L: \"Catheter placement selection for convection-enhanced delivery of therapeutic agents to brain tumors\". Harvard Dataverse V1, 2020. http://www.doi.org/10.7910/DVN/H7C6A2\n\nYeh FC, Wedeen VJ, Tseng WYI: Generalized q-Sampling imaging. IEEE Trans Med Imaging. 2010; 29(9): 1626–1635. PubMed Abstract | Publisher Full Text\n\nStoverud KH, Darcis M, Helmig R, et al.: Modeling Concentration Distribution and Deformation During Convection-Enhanced Drug Delivery into Brain Tissue. Transp Porous Media. 2012; 92: 119–143. Publisher Full Text\n\nANSYS® [computer program] Version 19.1.\n\nJain R: Transport in Molecules in the Tumor Interstitium: A Review. Cancer Res. 1987; 47(12): 3039–3051. PubMed Abstract\n\nBoucher Y, Salehi H, Witwer B, et al.: Interstitial fluid pressure in intracranial tumours in patients and in rodents. Br J Cancer. 1997; 75(6): 829–836. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBobo R, Laske D, Akbasak A, et al.: Convection-enhanced delivery of macromolecules in the brain. Proc Natl Acad Sci U S A. 1994; 91(6): 2076–2080. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBernstock JD, Wright Z, Bag AK, et al.: Stereotactic Placement of Intratumoral Catheters for Continuous Infusion Delivery of Herpes Simplex Virus -1 G207 in Pediatric Malignant Supratentorial Brain Tumors. World Neurosurg. 2019; 122: E1592–E1598. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShi M, Sanche L: Convection-Enhanced Delivery in Malignant Gliomas: A Review of Toxicity and Efficacy. J Oncol. 2019; 2019: 1–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCarpentier A, Laigle-Donadey F, Zohar S, et al.: Phase 1 trial of a CpG oligodeoxynucleotide for patients with recurrent gliobastoma. Neuro Oncol. 2006; 8(1): 60–66. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWeaver M, Laske D: Transferrin receptor ligand-targeted toxin conjugate (Tf-CRM107) therapy of malignant gliomas. J Neurooncol. 2003; 65(1): 3–13. PubMed Abstract | Publisher Full Text\n\nSampson JH, Akabani G, Friedman AH, et al.: Comparison of intratumoral bolus injection and convection-enhanced delivery of radiolabeled antitenascin monoclonal antibodies. Neurosurg Focus. 2006; 20(4): E14. PubMed Abstract | Publisher Full Text\n\nHau P, Jachimczak P, Schlingensiepen R, et al.: Inhibition of TGF-beta2 with AP 12009 in recurrent malignant gliomas: from preclinical to phase I/II studies. Oligonucleotides. 2007; 17(2): 201–212. PubMed Abstract | Publisher Full Text\n\nDesjardins A, Gromeier M, Herndon JE, et al.: Recurrent glioblastoma treated with recombinant poliovirus. N Engl J Med. 2018; 379(2): 150–161. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWarren K: Beyond the blood-brain barrier: the importance of central nervous system (CNS) pharmacokinetics for the treatment of CNS tumors, including diffuse intrinsic pontine glioma. Front Oncol. 2018; 8: 239. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBeffinger M, Schellhammer L, Pantelyushin S, et al.: Delivery of Antibodies into the Murine Brain via Convection-enhanced Delivery. J Vis Exp. 2019; (149): 1–7. PubMed Abstract | Publisher Full Text\n\nKanemitsu T, Kawabata S, Fukumura M, et al.: Folate receptor-targeted novel boron compound for boron neutron capture therapy on F98 glioma-bearing rats. Radiat Environ Biophys. 2019; 58(1): 59–67. PubMed Abstract | Publisher Full Text\n\nAllen J, Wang J, Zolotarskaya OY, et al.: PEAMOtecan, a novel chronotherapeutic polymeric drug for brain cancer. J Control Release. 2020; 321: 36–48. PubMed Abstract | Publisher Full Text\n\nEnríquez Pérez J, Kopecky J, Visse E, et al.: Convection-enhanced delivery of temozolomide and whole cell tumor immunizations in GL261 and KR158 experimental mouse gliomas. BMC Cancer. 2020; 20(1): 7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStephen ZR, Chiarelli PA, Revia RA, et al.: Time-Resolved MRI Assessment of Convection-Enhanced Delivery by Targeted and Nontargeted Nanoparticles in a Human Glioblastoma Mouse Model. Cancer Res. 2019; 79(18): 4776–4786. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTosi U, Kommidi H, Bellat V, et al.: Real-Time, in Vivo Correlation of Molecular Structure with Drug Distribution in the Brain Striatum Following Convection Enhanced Delivery. ACS Chem Neurosci. 2019; 10(5): 2287–2298. PubMed Abstract | Publisher Full Text\n\nPang HH, Chen PY, Wei KC, et al.: Convection-Enhanced Delivery of a Virus-Like Nanotherapeutic Agent with Dual-Modal Imaging for Besiegement and Eradication of Brain Tumors. Theranostics. 2019; 9(6): 1752–1763. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFinbloom JA, Aanei IL, Bernard JM, et al.: Evaluation of Three Morphologically Distinct Virus-Like Particles as Nanocarriers for Convection-Enhanced Drug Delivery to Glioblastoma. Nanomaterials (Basel). 2018; 8(12): 1007. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKhateeb K, Griggs D, Sabes P, et al.: Convection Enhanced Delivery of Optogenetic Adeno-associated Viral Vector to the Cortex of Rhesus Macaque Under Guidance of Online MRI Images. J Vis Exp. 2019; (147): 1–8. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "86990",
"date": "18 Jun 2021",
"name": "Ryuta Saito",
"expertise": [
"Reviewer Expertise Brain tumor surgery",
"CNS drug delivery."
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this paper, the authors developed a quantitative computer simulation model to optimize the catheter infusion locations for convection-enhanced delivery (CED). The selection of catheter placement is one of the most important points to consider for successful therapeutic distribution.\nUnfortunately, this study just contains mathematical calculations with lots of assumptions without any verification. However, the mathematical methods the authors demonstrated in this manuscript give some suggestions for optimal catheter locations.\nI think solving the following issues will improve this study:\nThe authors selected suspension of the oncolytic HSV as an infusate. Density and viscosity vary among infusates. Does the optimal placement differ when the density and viscosity of infusate change?\n\nHow about changing the flow rate? The calculation is done at a constant rate of 0.1 mL/hr. A higher rate such as 0.2 mL/hr is also used in clinical trials. Does it give any difference in optimal catheter placement?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "93999",
"date": "24 Sep 2021",
"name": "Chengyue Wu",
"expertise": [
"Reviewer Expertise MRI acquisition and analysis",
"medical image processing",
"computational fluid dynamics",
"image-based mathematical modeling for cancer."
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSummary and broad comments:\nIn this study, the authors investigated a response surface-based technique to assist optimization of catheter placement of convection-enhanced delivery (CED). As the main strength, this study presented a surrogate scheme to generate a continuous surface of the therapy delivery outcome (measured by therapy distribution percentage within the tumor) from sparsely sampled candidates of catheter placement. If validated, this technique does have the potential to address the challenge of intensive time consumption of simulating all candidates to place CED catheter(s). Thus, I do appreciate the authors’ effort in this investigation and have a positive point of view on the potential significance of this study.\nHowever, the current presentation of the aim, methods, and detailed implementation is not clear in the paper. It is hard to follow how the experiments (for both MRI data collection and in silico simulation) are performed and why they are designed as such. Many essential details of the mathematical justification and numerical implementation are missing, so the reported technique is not replicable in the present. I would suggest the authors rethink the presentation and the design, as well as provide more descriptions and, if applicable, more in silico validations for revision. Please see below for more specific suggestions.\nSpecific comments:\n\nIntroduction\nThe introduction described optimization techniques that have been used for CED catheter placement and the applications of RSO which was used for this study. However, it did not cover much information about the actual challenges in the field of optimization of catheter placement for CED. For example, a failed planning of the CED catheter could lead to a significant backflow along the catheter, a leakage into CSF, or a blockage of the catheter tube. The choice of CED catheter placement is a complex decision-making process and time-consuming because it needs to consider multiple factors including the distance of catheter tips to CSF volumes, the effect of catheter trajectory on white matter connectivity, and the heterogeneity of brain tissue. There has been commercial software assisting this process, and there are many studies aiming to improve the related techniques. The introduction would benefit from a brief review on the unmet needs or challenges in the field of computer-assisted optimization of CED catheter placement, and how this study might contribute to overcoming any of these challenges.\n\nPlease provide a reference to the response surface optimization technique. And, if “DOE” refers to a developed method or scheme, please also provide a reference.\n\nThis study investigated the oncolytic herpes simplex for the therapy, which could be considered large particles. However, small molecules and nanoparticles would have very different fluid dynamics. Was this factor considered when simulating the distribution of therapy?\nMethods\nTumor parameters: How many patients were included in this study – I assume one? If so, basic information about this patient should be described. For example, what kind of brain cancer they had and did this patient actually go through the CED administration?\n\nTumor parameters: I would refer to the imaging technique to reconstruct diffusion tensors as “Diffusion Tensor Imaging (DTI)”, instead of “DWI” since the standard DWI does not provide a measurement of tensors. Also, please provide the basic information about the imaging protocol (e.g. imaging scanner, acquisition sequence, and parameters).\n\nTherapy parameters: The actual model used for simulating CED in this study is unclear. What are the boundary conditions? How the infusion source term is defined? What numerical methods, solvers, and/or software were used? Was the simulation conducted within the tumor or within the whole brain?\n\nIt is mentioned in the “Tumor parameters” section that the map of viscous resistance is calculated from DTI, while in the section of “Therapy parameters”, it indicated that the flow was analyzed using the viscosity value for water (also, the unit of “1.003 * 10-6 kg/mm-s” is confusing. I assume the authors mean “kg/mm⋅s”). Which assignment was actually used for the experiments?\n\nOptimization: radius = 1.5 mm for what? The point source for infusion?\n\nOptimization: How were the 16 candidates for catheter placement chosen? And what do you mean by “four baseline locations”? It is indicated in the manuscript that “The baselines represent the locations that produced the highest therapy distribution percentages in the tumor”. How did you know this before performing the simulation? If it is a result of the simulation, it should not be mentioned in the Methods section to avoid confusion. If it is prior information, please indicate how it is determined. And is there any relationship between the “16 locations” and “28 DOE-generated sampling points”?\n\nOptimization: If I understand correctly, the response surface is generated using sampling points obtained from DOE by some sort of interpolation. So what algorithm was used to generate this surface? Are there any theories indicating how many sampling points are sufficient to guarantee a unique (or at least robust) solution of the response surface? Are there any assumptions on the geometry of the target response surface for the solution to be realistic, and how these assumptions/conditions are justified for the application in this study? If there are related theories and the assumptions can be justified mathematically, please provide the references or derivations. If the theoretical justification is not available, I would suggest doing a denser sampling in the input parameter space (for example, 500 candidates of the catheter tip location), and then compare the response surfaces generated by dense sampling with that by the sparse sampling (as presented in the current study) to evaluate whether the sparse sampling is sufficient to represent the geometry of real response surface. These could be included in the main text or as a supplemental.\n\nOptimization: How are the tumor regions defined? Why four and are they physiologically different?\n\nOptimization: Why are only locations on the tumor boundaries considered as candidates? In real practice, the tip of CED would be inserted into the tumor (usually the center of the tumor, if surgically accessible), instead of putting on the tumor burden.\nDiscussion\nIt is claimed that the predicted percentages from the response surface and observed percentages from the DOE aligned closely. But the actual value of the goodness of fit is not given. This claim needs to be supported by a quantitative report in the Results section.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-18
|
https://f1000research.com/articles/10-17/v1
|
12 Jan 21
|
{
"type": "Research Article",
"title": "Enhancing nutrition specific interventions through public health policies and public-private partnerships in the Eastern Mediterranean Region: a desk review",
"authors": [
"Arabia Mohammad Ali",
"Hassan Salah",
"Mataria Awad",
"Hammerich Asmus",
"Ayoub Al-Jawaldeh",
"Hassan Salah",
"Mataria Awad",
"Hammerich Asmus",
"Ayoub Al-Jawaldeh"
],
"abstract": "Background: Public private partnerships (PPPs) in public health have been widely promoted as an effective tool for accelerating progress toward achieving the United Nation’s Sustainable Development Goals (SDGs), including SDG2 “to eliminate hunger”. Partnership with the private sector was found to be very instrumental in improving the nutritional status of poor and food-insecure people and promoting healthy lifestyles. In countries of the Eastern Mediterranean Region (EMR), PPPs for nutrition actions have been mainly driven by the United Nations (UN) and international development agencies to support low and middle-income countries in achieving better health outcomes. Despite the increased prominence of engaging the private sector in public health actions in the EMR, evidence on the role of the private sector in the design/implementation of these actions is still not documented. The objective of this study is to assess the role and contribution of the private sector in the design/implementation of nutrition-specific interventions addressing the double burden of malnutrition in countries from the EMR and identify the key factors for successful PPP implementation. Methods: The study design was descriptive using secondary data obtained from digital internet sources, including World Health Organization (WHO) databases, grey literature, and websites of the UN and international development agencies. Results: The desk review revealed that the private sector has played a sizable role in the implementation of food fortification initiatives and in the implementation of nutritional policies promoting healthy diets. The experience of the EMR shows the significant impact of government commitment, and the availability of national policies and systems for monitoring and enforcement on the sustainability and effectiveness of PPP-specific nutrition interventions. Conclusion: The study emphasizes the key actions recommended for governments to enhance the application of PPPs as a tool to accelerate the EMR’s progress toward achieving nutrition targets under SDG2 by 2030.",
"keywords": [
"Role of the private sector",
"Public-Private Partnerships",
"Specific Nutrition Interventions",
"Food Fortification",
"Nutrition EMR",
"Double Burden of Malnutrition."
],
"content": "Introduction\n\nHealth care systems across the world, even the most developed among them, are grappling with serious challenges. Population growth and aging, disruption of economic development due to conflict, increased poverty and economic hardship, and the emergence of new communicable diseases like COVID-19 are most likely to disrupt the public sector performance, leaving many health care systems vulnerable. These realities are posing an unprecedented challenge for governments to meet the United Nation’s Sustainable Development Goals (SDGs) by 2030, and more specifically SDG2 “to eliminate hunger”.\n\nBased on the Global Nutrition Report 2020, “Not one country is on course to meet all ten of the 2025 global nutrition targets and just 8 of 194 countries are on track to meet four targets”1. The gravity of the nutrition challenges demand the involvement of non-state actors including the private sector to complement government’s efforts in addressing the double burden of malnutrition around the world. According to the United Nations (UN) Food and Agriculture Organization (FAO), “The fight against hunger can only be won in partnership with governments and other non-state actors, among which the private sector plays a fundamental role”2.\n\nWith the adoption of the 2030 Agenda for Sustainable Development and namely SDG17 (Partnerships for the Goals), public private partnerships (PPPs) in nutrition actions have been widely promoted and encouraged by many UN development agencies who continue to play a pivotal role in assisting member states in identifying and building networks with the business sector.\n\nIn 2014, the World Health Organization (WHO) established the Global Coordination Mechanism on the Prevention and Control of Non-Communicable Diseases (GCM/NCD) to connect member states with various stakeholders, including the private sector3. At the global level, several UN development agencies have also partnered with food production industries to enhance the scope of several nutrition interventions. Examples of such initiatives include the partnership between the WHO and the International Food and Beverage Alliance (IFBA) to eliminate industrially produced trans-fats in food products4, and the World Food Program (WFP) partnership with several private firms to address child hunger and build nutrition education programs around the world5. Partnership with the private sector was also found to be very instrumental in promoting healthy lifestyle and developing initiatives to prevent childhood obesity in the United States and Europe6. Other examples of well-established PPPs in nutrition actions include food fortification solutions, such as the Food Fortification Initiative (FFI) originally called the Flour Fortification Initiative7.\n\nThe UN has taken further steps to support governments in establishing strong PPPs in nutrition interventions by establishing not-for-profit international agencies like the Scaling Up Nutrition (SUN) movement, the SUN Business Network (SBN), and the Global Alliance for Improved Nutrition (GAIN). The SUN Business Network work aimed at pushing the nutrition agenda to the forefront by establishing alliance and partnership with key stakeholders, including the private sector8. These agencies continue to work closely with several governments worldwide, including Eastern Mediterranean Region (EMR) countries in developing collective actions to address malnutrition.\n\nDespite the increased prominence of PPPs as an effective tool for enhancing governments’ nutritional programs, the added value of engaging the private sector in nutrition actions is still limited. The literature indicates that sustainability of PPPs in nutrition actions is highly associated with the private sector competitive advantage/profit making9. Similar to many countries worldwide, the governments of the EMR opted to enhance the quality and efficiency of their nutrition actions by mobilizing the resources of the private sector at the national level. According to the available literature, the private sector involvement in the region has been limited to the implementation of food fortification initiatives (FFI), and in interventions supporting healthy diets10. Despite the engagement of the private sector in nutrition actions, the double burden of malnutrition continues to escalate in the region11. To date, there is little evidence documenting the impact and effectiveness of PPPs in nutrition actions12. This is specifically true for countries in the EMR, where the role of the private sector in the implementation of nutrition actions is still not documented.\n\nConsidering the above, the focus of this paper is to assess the role and contribution of the private sector in the design and implementation of specific nutrition interventions addressing the double burden of malnutrition in the EMR, and identifying the factors influencing the sustainability and effectiveness of PPPs in these countries.\n\n\nMethods\n\nThe study design was descriptive using secondary data obtained from digital internet sources as the primary data collection method, i.e. a desk review.\n\nThe data on the role/contribution of the private sector, and the challenges encountered in the implementation of specific nutrition interventions across EMR countries was collated through a comprehensive literature review from various sources, including WHO Global database on the implementation of nutrition actions (GINA), grey literature, specific reports on nutrition actions published by implementing partners (WHO, UNICEF, WFP, FAO) and other UN and International development agencies.\n\nThe first step of the data collection process aimed at developing a comprehensive list of specific nutrition policies/actions that have been implemented across EMR countries. A team of three reviewers conducted the database search using GINA database (A–Z country list) to develop a comprehensive list of specific nutrition actions for each of the 22 EMR countries. The countries were distributed among the reviewers as per the WHO classification for EMR countries (Table 1).\n\nSource: Strategy on nutrition for the Eastern Mediterranean Region 2020–203013.\n\nThe information gathered from GINA, included name of country, name/title of the nutrition initiative, background information on the initiative, and the implementing partners.\n\nThe second step of the data collection process included a grey literature and Google web search for articles and reports published between 2010 and 2020 in English. The search also included websites of UN and international development agencies to locate detailed information on the actions that were not included in GINA or google search. For the Google search, reviewers used the following search terms “specific title/name of the nutrition program/ action with the country name”.\n\nThe reviewers used the inclusion criteria (defined below) to screen the full text independently and as a team of three to determine the eligibility of the nutrition action/ program.\n\nThe initial list was then narrowed down to include only specific nutrition actions/interventions that were implemented in partnership with at least one private for-profit organization.\n\nA short data extraction table (Microsoft Excel 2013) was used to extract the information from the internet sources to allow for data analysis and documentation of findings. The main variables that were extracted from the literature included country name, name of the action/intervention, program objectives, implementing partners, role and responsibilities of the partners, reported challenges, and reported health outcomes/evaluation if available.\n\nIn order to find and select the nutrition actions relevant to the objective of this study, the following inclusion criteria was developed to inform the data collection process:\n\n1. Nutrition specific interventions: For the sake of this paper, the assessment of the private sector role was limited to nutrition-specific actions that were recognized by the World Bank as the most cost effective and high return nutrition interventions (Table 2), in addition to interventions targeting the increased prevalence of obesity in the EMR.\n\n2. Public-private partnerships: For the specific purpose of this study, we used Reich’s definition of PPPs: “Public-Private Partnerships involve at least one private for-profit organization with at least one not-for-profit organization, who provide a joint sharing of efforts and of benefits and are committed to the creation of social value (improved nutrition and health), especially for disadvantaged populations.”15.\n\nSource: Scaling up Nutrition: A Framework for Action, April 201114.\n\n\nResults\n\nBased on the desk review, it was evident that countries of the EMR have invested in developing various national nutritional policies, nutrition strategies and action plans to accelerate their progress toward achieving SDG2. Most of these countries have worked with various stakeholders, including UN development agencies, national and international non-governmental organizations (NGOs), other sectors, and the private sector in the implementation of their national nutrition strategies. Unfortunately, the role of the private sector in the implementation of the nutrition policies/actions in the region is not adequately covered in the literature. Based on the relevant literature, the involvement of the private sector in actions addressing nutrition actions varied across interventions and countries and ranged from being high to non-existent in certain actions as described below.\n\nFollowing the WHO guidelines, all EMR countries have integrated the promotion of infant and young children feeding practices (IYCF) at the primary care level as part of the antenatal and postnatal packages of care. Initiatives for promoting exclusive breastfeeding and complementary feeding counseling for infants and young children were also provided during ante-natal and post-natal visits in most of the countries of the region. The implementation of these initiatives has required the involvement of many stakeholders including the Ministries of Health, international and UN development agencies as well as national NGOs. Other initiatives targeting IYCF in these countries included the implementation of the Baby-Friendly Hospitals) as part of the global efforts to protect and support breastfeeding.\n\nThe role of the private sector in the development and implementation of IYCF actions in advanced nutrition transition countries (ANT)1 was only reported in Bahrain. The IYCF program in Bahrain was developed in collaboration with the International Baby Food Action Network (IBFAN), and the agencies for importing and distributing breast milk substitutes in the country16. In early nutrition transition countries (ENT)2, private sector involvement was only reported in Lebanon. The Lebanese Ministry of Health worked in close collaboration with UN agencies (WHO, UNICEF), international NGOs, Order of Nurses, and the private media to promote exclusive breastfeeding17. In EMR countries with significant under-nutrition/in complex emergency (SUN/CE)3, the involvement of the private sector in the implementation of IYCF actions was reported only in Pakistan and Iraq. In 2019, the government of Pakistan launched a Breast-Feeding Campaign in collaboration with UNICEF, and the private sector18. Through this program, 50 million people were reached by disseminating IYCF messages through theaters, seminars, social media and television spots. The program was funded through corporate social responsibility. A similar breastfeeding promotion campaign was also implemented in Iraq in collaboration with UNICEF and in partnership with the national mobile network provider Zain19.\n\nFollowing WHO guidelines, ANT and ENT countries have also enacted laws/regulations to protect and promote exclusive breastfeeding by implementing the International code for marketing of breast milk substitutes (the code)4. However, none of these countries have succeeded in achieving full implementation of the code regulations. The most common reported challenge in the implementation of the code included the governments’ inability to monitor and enforce the compliance of the relevant stakeholders, including the private sector. Lack of adequate monitoring and enforcement mechanisms of the code was also reported in Bahrain16, and the Kingdom of Saudi Arabia (KSA)20. In Kuwait, over-prescription of infant formula by the private sector physicians has also been identified as one of the main challenges in implementing the code21. A weak monitoring and evaluation framework, accompanied by weak advocacy and implementation mechanisms of the code was also reported in Jordan22, and Lebanon. In Lebanon, some hospitals violated the code by providing pre-lacteal feeding before breastfeeding and offering samples for mothers upon discharge23. The implementation of the code regulations is highly determined to the engagement and compliance of the private sector (food manufacturers and distributors). However, information on the role of the private sector in the implementation of the code was lacking.\n\nWithin EMR countries, programs to provide different micronutrients supplementation (vitamin D, iron, folic acid) to different age groups are mainly led by the government with no involvement from the private sector. Engagement of the private sector in the implementation of vitamin A supplementation programs in EMR countries was only reported in Palestine through government partnership with local producers of fortified biscuits and milk24. In Somalia, the National Micronutrient Deficiency Control Strategy indicated the important role of the private pharmaceutical industry in supplying, promoting and implementing quality controls on the program25. However, available evidence on the involvement of the private sector in this initiative was not available in the literature.\n\nIn both ANT and ENT countries, protocols for screening and management of severe acute malnutrition (SAM) and moderate acute malnutrition (MAM) have been incorporated through food distribution programs and have been integrated at the primary health care level. Yet none of these initiatives has involved the private sector. In SUN/CE countries, the screening and management of SAM and MAM are being led by Ministries of Health in collaboration with UNICEF as the lead agency and other UN development agencies, including the WHO, WFP, and local NGOs on the ground.\n\nThe involvement of the private sector in SAM treatment was reported only in Sudan and Pakistan. In 2019, the Sudan Ministry of Health in partnership with UNICEF collaborated with local ready-to-use therapeutic food (RTUF) manufacturers to strengthen the supply chain of RTUF in three Darfur states and the states of South Kordofan and Khartoum26. This partnership resulted in maintaining an adequate reserve of RTUF in the supply pipeline. In Pakistan, the WFP and UNICEF are working with local manufacturers to increase the local production of RTUF27.\n\nSalt iodization (SI) and wheat flour fortification (WFF) are the two most common food fortification initiatives in the region. According to the literature, the implementation of WFF programs in most of the EMR countries is being led by the Ministries of Health working in partnership with several UN development agencies, international development agencies, donors, and the private sector (mills industry). The same applies to SI initiatives. For the past two decades, WHO and UNICEF have been working in close collaboration with the International Council for Control of Iodine Deficiency Disorders (ICCIDDs), other international agencies and the salt industry to support the governments in the region in developing their national salt iodization programs. In 2008, UNICEF fostered a new partnership with the Global Alliance for Improved Nutrition (GAIN) to further support the SI programs in selected priority countries in urgent need for SI28.\n\nThe majority of the ANT countries have achieved good coverage of WFF and have reported improvement in health outcomes as a result of the food fortification initiatives. Examples of successful food fortification initiatives include Kuwait, whereby one industrial mill in the country produces all of the flour, and 100% coverage of fortification is reported29. In ENT countries, WFF is mandatory in the Islamic Republic of Iran, Jordan and Morocco. Other countries have introduced voluntary measures for flour fortification such as Lebanon. In Egypt, a flour fortification program (with iron and folic acid) has been stalled since the political changes in 2011, despite having undergone a 5-year preparatory phase29. Examples of successful flour fortification initiatives include the ones implemented in the Islamic Republic of Iran and in Jordan, with fortification levels of 100% and 93%, respectively29. The success of the Jordanian WFF program was attributed to the government strength in enforcing the food industry compliance with food regulations and standards30. Other countries including Egypt and Lebanon are experiencing challenges in implementing and sustaining these initiatives. These challenges are the result of poor enforcement mechanisms and coordination by the governments, as well as limited resources29.\n\nIn SUN/CE countries, WFF is mandatory in Djibouti, Iraq, Palestine, and Yemen. Most countries have introduced WFF voluntary legal standards, except for Libya and Somalia where flour fortification is not implemented yet29. The Pakistan partnership with the private miller’s industry for the implementation of the national WFF program, did not yield the expected health benefits/desired national coverage31.\n\nAll ANT countries have also implemented SI initiatives on a national scale, except for the KSA. SI iodization programs in the United Arab Emirates (UAE) and Kuwait are progressing quickly toward achieving full coverage and eliminating iodine deficiency disorders in their countries32.\n\nENT countries have also implemented mandatory SI initiatives on a national scale. Some have policies guiding the process of salt iodization including Egypt, the Islamic Republic of Iran, Lebanon, and Morocco32. According to the literature, the partnership between the government and salt producers for the implementation of the national salt iodization program is both sustainable and effective. The SI in Jordan has significantly contributed to achieving better health outcomes by reducing iodine deficiency disorders in the country33. In Lebanon, a policy brief to inform the national SI program was published through a joint effort of affiliated centers at the American University of Beirut (AUB), international NGOs and donors. The brief noted that, despite the existence of a law for SI, the lack of financial resources and poor enforcement mechanisms are hindering effective implementation34. In Morocco, the government failed to sustain the SI initiative due to increased competition from non-iodized salt producers32. Other challenges imposed due to the activities of the private sector are notable in Egypt, where many salt producers, re-packagers, and unlicensed salt producers continue to threaten the sustainability of salt iodization due to the lack of effective governmental monitoring mechanisms35.\n\nIn SUN/CE countries, SI initiatives are being implemented in Afghanistan, Iraq, Libya, Pakistan, Palestine, Sudan, Syria, and Yemen. According to UNICEF, Afghanistan is on track but still has a long way to go. Countries including Sudan, Pakistan Somalia and Yemen continue to struggle with attaining the minimal required consumption of iodized salt36.\n\nBased on the available literature, it was evident that the private sector plays a significant role in the implementation of food fortification initiatives (FFIs) in most of the EMR countries. However, it was also evident that many of the countries in the region are facing major challenges in the implementation of their FFIs due to the gaps in their regulatory framework, lack of financial resources, and weak monitoring and enforcement mechanisms.\n\nThe literature showed that the private sector has a sizable role in the implementation of nutrition actions promoting healthy diets in both ANT and ENT countries. However, the information on the countries’ experience with PPPs in the domain of obesity prevention and control is not readily available.\n\nAt the policy level, some of the ANT and ENT countries such as Bahrain, Saudi Arabia, Kuwait, Oman, Egypt and Jordan have enacted reformulation policies/legislative measures to reduce salt and trans-fatty acids in food products37. Others have taken nutrition actions to reduce sugar intake, such as Egypt, Jordan, Morocco, and Tunisia. Nutrition labeling legislation was also adopted in countries such as Bahrain, the UAE, Saudi Arabia, Tunisia and Jordan. In Bahrain and the UAE, labeling is mandatory and supported by the country’s national policy38.\n\nMarketing of unhealthy food for children is also being regulated in some of the EMR countries as well. Oman has established legislation to ban the marketing of unhealthy foods and beverages to school age children. Egypt, the Islamic Republic of Iran, Morocco, and Tunisia have introduced a taxation and price policy on unhealthy foods and/or beverages38. Food guidelines and standards were also adopted in the Islamic Republic of Iran, Lebanon, and Morocco. According to best practices, the implementation of these policies requires the involvement of the private sector (food manufacturers). However, data on the role of the private sector in these initiatives was not available.\n\nAt the program level, the majority of ANT and ENT countries have developed multi-sectoral platforms for the prevention and control of non-communicable diseases. These actions involved different partners, including ministries of Health, Education and Sports, alongside different UN agencies, the WHO, national NGOs and the private sector (representatives from the food industry). Qatar has established a large scale multi-sectoral network, including the private sector, media, education, and research centers, as well as the tourism sector to support the implementation of its obesity nutrition actions. The Qatar experience illustrates the role of the tourism sector (hotels and restaurants) in increasing the reach of obesity awareness and the role of the private sector in mobilizing resources38. A similar multi-sectoral program was also implemented in Oman to prohibit the marketing of high-fat, energy dense, and/or micronutrient-poor foods and beverages on school premises38.\n\nThe involvement of the private sector in obesity interventions in both ANT and ENT countries is well reported through a variety of programs. Engagement of the media in the implementation of awareness campaigns on healthy eating practices was reported in Bahrain, Qatar, KSA, Lebanon, the Islamic Republic of Iran, Jordan, Morocco, and Tunisia. Similarly, the UAE has worked closely with the private sector to implement various obesity awareness initiatives, such as the UAE’s national program “2021 Healthy Children Initiative “and the “Health Heroes” smart app39. Bahrain and Saudi Arabia were the only countries to implement awareness campaigns at the workplace to increase awareness about obesity and its related consequences.\n\nPartnership with the food sector (local bakeries) to reduce the amount of salt added to bakery products was reported in Bahrain, Morocco, and Tunisia. Other examples of the food industry involvement include, the “United for Healthier Kids U4HK” initiative led by Nestlé Middle East and implemented in the UAE and Kuwait to help parents establish healthier eating, drinking and lifestyle habits for children aged four to 1239. In Saudi Arabia, the “Al-Haraka Baraka physical activity promotional Initiative” is a collaborative effort between a public university (King Saud University), a non-profit organization (Arab Nutrition Center), and the private sector (Mars Middle East Inc.). This initiative is designed as an educational program targeting school aged children 6-12 years of age40.\n\nThese findings indicate that EMR countries have taken major steps towards developing policies and implementing strategies to prevent and combat obesity in partnership with the private sector. However, more evaluation evidence is needed to determine the impact of these partnership on health outcomes.\n\nAmong SUN/CE, only Pakistan has taken some action to control the increased prevalence of obesity among adolescents and adults. In 2019, the Pakistan Ministry of Health developed a policy paper on adolescent (boys and girls) nutrition in Pakistan41. This project was done in partnership with the World Bank, GAIN, and Safansi. In 2020, the Ministry of Health in Pakistan took serious steps toward eliminating industrially produced trans fatty acids with the support of SUN Business Network (SBN), and GAIN, in partnership with the International Food and Beverage Alliance42. Both programs involve partnership between the government of Pakistan and the private sector. However, both projects are still in the initiation phase and thus evaluation data was not available.\n\n\nConclusion and recommendations\n\nThe desk review revealed that there are serious gaps in the information available about the role and the contribution of the private sector in the implementation of nutrition actions in the EMR. Thus, more in-depth information is needed to be able to determine clear quantitative indicators on the impact and outcome of PPPs in nutrition interventions for the region.\n\nThe findings also showed that there have been mixed experiences in engaging the private sector across nutrition interventions as well as across countries of the EMR. The private sector has played a sizable role in the implementation of food fortification initiatives promoting IYCF, and diet related NCDs actions across many countries in the EMR and was found to be extremely limited in the implementation of micronutrient supplementation programs and the treatment of MAM and SAM.\n\nThe experience of EMR countries with PPPs in nutrition specific interventions also illustrates the potential benefits of the private sector engagement in the implementation of specific nutrition interventions. This is true in food fortification initiatives that have succeeded in improving nutrition health indicators. However, the study also revealed that many EMR countries are still struggling in the implementation of PPPs addressing FFI, the code of marketing of breast milk substitutes and marketing of healthy food. The failure to attain the PPP objectives and priorities in these countries were attributed to the absence of a supporting policy framework, processes to manage conflicts of interest, effective monitoring, and enforcement systems, as well as lack of transparency and accountability at the government level. Considering the above, PPPs have a great potential in supporting EMR countries in reaching SDG2 if they follow adequate principles and rules in the engagement of the private sector.\n\nThe recommendations below emphasize the key actions that governments and international implementing partners in EMR countries can take to ensure the success of PPP implementation in specific nutrition interventions:\n\n1. Establishing national regulatory and statutory frameworks in place to guarantee the commitment, accountability, and transparency of all key partners.\n\n2. Instituting strong monitoring and enforcement systems to monitor the performance of the private sector.\n\n3. Developing clear governance structure by enacting legally binding agreements with clear description of roles, and responsibilities as well as partnership objectives and desired outcomes.\n\n4. Developing a national policy/ strategy to attract and facilitate the private sector engagement in nutrition actions.\n\n5. Setting up a systematic approach to evaluating the impact and outcomes of PPPs in nutrition actions in the EMR.\n\nThe adoption of these recommendations/actions will enhance the application of PPPs as a tool to accelerate the EMR countries progress toward achieving the nutrition targets under SDG2 by 2030.\n\n\nData availability\n\nDryad: Public –Private Partnerships in Nutrition Specific Interventions in EMR Countries, https://doi.org/10.5061/dryad.dv41ns1ww43.\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).\n\n\nNotes\n\n1 Countries in Advanced Nutrition Transition, include Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and United Arab Emirates\n\n2 Countries in Early Nutrition Transition include: Algeria, Egypt, Islamic Republic of Iran, Jordan, Lebanon, Morocco, Palestine, and Tunisia.\n\n3 Countries with Significant Under Nutrition & Countries in Complex Emergencies include: Afghanistan, Djibouti, Iraq, Libya, Pakistan, Palestine (Gaza Strip), Somalia, Sudan, Syrian Arab Republic, and Yemen\n\n4 The International Code of Marketing of Breast-milk Substitutes contributes to the provision of safe and adequate nutrition for infants, by the protection and promotion of breast-feeding, and by ensuring the proper use of breast-milk substitutes (WHO, 1981. https://www.unicef.org/nutrition/files/nutrition_code_english.pdf.)",
"appendix": "References\n\nGlobal Nutrition Report. 2020. Reference Source\n\nFood and Agriculture Organization of the United Nations (FAO): FAO Strategy for partnerships with the private Sector. 2013. Reference Source\n\nWorld Health Organization (WHO): WHO Global Coordination Mechanism on the Prevention and Control of Non-Communicable Diseases. 2016. Reference Source\n\nWorld Health Organization (WHO): WHO Report on Global Trans Fat Elimination. 2019. Reference Source\n\nWorld Food Program (WFP): 6 partners helping us to Change the World. 2017. Reference Source\n\nKraak VI, Story M: A public health perspective on healthy lifestyles and public-private partnerships for global childhood obesity prevention. J Am Diet Assoc. 2010; 110(2): 192–200. PubMed Abstract | Publisher Full Text\n\nFood Fortification Initiative: \"About Us\". 2015. Reference Source\n\nScaling Up Nutrition (SUN): The history of the SUN Movement. 2015. Reference Source\n\nHoddinott JF, Gillespie S, Yosef S: Public-private partnerships and the reduction of undernutrition in developing countries. International Food Policy Research Institute (IFPRI). 2015. Publisher Full Text\n\nWorld Health Organization (WHO): Global nutrition policy review 2016-2017: country progress in creating enabling policy environments for promoting healthy diets and nutrition. 2018. Reference Source\n\nNasreddine L, Ayoub JJ, Al Jawaldeh A: Review of the nutrition situation in the Eastern Mediterranean Region. East Mediterr Health J. 2018; 24(1): 77–91. PubMed Abstract | Publisher Full Text\n\nGillespie S, Haddad L, Mannar V, et al.: The Politics of Reducing Malnutrition: Building Commitment and Accelerating Progress. Lancet. 2013; 382(9891): 552–569. PubMed Abstract | Publisher Full Text\n\nWorld Health Organization (WHO): Strategy on nutrition for the Eastern Mediterranean Region 2020-2030. 2019. Reference Source\n\nSUN: Scaling up Nutrition: A Framework for Action. 2011. Reference Source\n\nReich M: Public-Private Partnerships for Public Health. 2002. Reference Source\n\nInternational Baby Food Action Network (IBFAN): The World Breastfeeding Trends Initiative (WBTi), Assessment Report, Bahrain. 2015. Reference Source\n\nGlobal database on the Implementation of Nutrition Action (GINA): National campaign on breastfeeding. 2016.\n\nUnited Nations Children's Fund (UNICEF). 2019. Reference Source\n\nUnited Nations Children's Fund (UNICEF): UNICEF Annual Report 2017 Iraq. 2017. Reference Source\n\nInternational Baby Food Action Network (IBFAN): Report on The Situation of Infant and Young Child Feeding in Saudi Arabia. 2016. Reference Source\n\nRios R, Riquelme H, Beshlawy S: Prescribing under the Influence: The Business of Breastmilk Substitutes. Soc Sci. 2016; 5(4): 53. Publisher Full Text\n\nInternational Baby Food Action Network(IBFAN): The World Breastfeeding Trends Initiative (WBTi), Assessment Report, Jordan. 2015. Reference Source\n\nInternational Baby Food Action Network(IBFAN): World Breast Feeding Trends Initiative (WBTi), Assessment Report. 2016. Reference Source\n\nFood and Agriculture Organization of the United Nations(FAO): Nutrition Country Profile Palestine. 2005. Reference Source\n\nWorld Health Organization(WHO): Somali National Micronutrient Deficiency Control Strategy (2014 – 2016). 2014. Reference Source\n\nUnited Nations Children's Fund(UNICEF): UNICEF Sudan. 2016. Reference Source\n\nUnited Nations Children's Fund(UNICEF). 2019. Reference Source\n\nUnited Nations Children's Fund (UNICEF) and GAIN: Brighter Futures: Protecting early brain development through salt iodization – The UNICEF-GAIN partnership project. New York: UNICEF. 2018. Reference Source\n\nAl Jawaldeh A, Pena-Rosas J P, McColl K, et al.: Wheat flour fortification in the Eastern Mediterranean Region. 2019. Reference Source\n\nAlwan A: Nutrition in Jordan; Update and Plan of Action. 2006. Reference Source\n\nNHSR&C: National Nutrition Survey 2018. 2018. Reference Source\n\nDoggui R, Al-Jawaldeh H, Al-Jawaldeh A: Trend of Iodine Status in the Eastern Mediterranean Region and Impact of the Universal Salt Iodization Programs: a Narrative Review. Biol Trace Elem Res. 2020; 198(2): 390–402. PubMed Abstract | Publisher Full Text\n\nMassa’d H, Barham R: National survey to assess iodine deficiency disorders (IDD) among Jordanian children—2010. Jordan Ministry of Health Nutrition Department, World Health Organization Non-Communicable Disease Department. Amman, Jordan. 2011.\n\nAkik C, El-Mallah C, Ghattas H, et al.: K2P Policy Brief: Informing Salt Iodization Policies in Lebanon to Ensure Optimal Iodine Nutrition. Knowledge to Policy (K2P) Center, Beirut, Lebanon. 2016. Reference Source\n\nIsmail MB, Hussein I: Sustaining universal salt iodization in Egypt: program successes and challenges. IDD NEWSLETTER. 2018; 46: 12–13. Reference Source\n\nUnited Nations Children's Fund (UNICEF): Sustainable Elimination of Iodine Deficiency. 2008. Reference Source\n\nGlobal database on the Implementation of Nutrition Action (GINA): Obesity and diet-related NCDs - Removal/reduction of trans fatty acids - All population groups. 2009. Reference Source\n\nGlobal database on the Implementation of Nutrition Action (GINA): Promotion of healthy diet and prevention of obesity and diet-related NCDs. Healthy lifestyle Campaign: Our future lies in our Health - Media promotion of healthy nutrition - All population groups. 2016. Reference Source\n\nDrewnowski A, Caballero B, Das JK, et al.: Novel public–private partnerships to address the double burden of malnutrition. Nutr Rev. 2018; 76(11): 805–821. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAl-Hazzaa HM, AlMarzooqi MA: Descriptive analysis of physical activity initiatives for health promotion in Saudi Arabia. Front Public Health. 2018; 6: 329. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTumilowicz A, Badar A, Pasha M: A Policy Paper on Adolescent Nutrition in Pakistan. 2019. Reference Source\n\nScaling Up Nutrition Business Network. (SunBizNet): Eliminating Industrially Produced Trans Fatty Acids in Pakistan. 2020. Reference Source\n\nArabia: Public –Private Partnerships in Nutrition Specific Interventions in EMR Countries, Dryad. Dataset. 2020. http://www.doi.org/10.5061/dryad.dv41ns1ww"
}
|
[
{
"id": "77162",
"date": "28 Jan 2021",
"name": "Majid Mqbel Alkhalaf",
"expertise": [
"Reviewer Expertise Nutrition",
"Dietary and nutrition assessment methods",
"public and clinical nutrition."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study is well written. I have few and minor comments need to be addressed below:\nFor methodology part:\n\nNeed to clarify whether the authors performed double revision for the extracted data.\n\nNeed to attach a table of the collected studies and the extracted data.\n\nAuthors mentioned that the collected studies were conducted between 2010-2020. However, they did not provide any justification for this criterion.\nFor result part:\n\nThe flow of presenting the results was consistent and clear enough. However, there were many missing references need to be cited. For example:\nIn the “Role of the private sector in promoting good nutritional practices in EMR countries” part, for the second paragraph, there was no cited reference after the following sentence: “the involvement of the private sector in the implementation of IYCF actions was reported only in Pakistan and Iraq”\n\nFor the references part:\n\nThe volume and issue numbers were missing in the following references: 9, 29, 34, 40, and 41. In reference 18, “United Nations Children's Fund (UNICEF)” should be bolded.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "78862",
"date": "09 Apr 2021",
"name": "Majid Hajifaraji",
"expertise": [
"Reviewer Expertise Nutrition and food policy",
"dietary and nutrition assessment methods",
"clinical nutrition",
"research methodology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper is well written and is based on rigorous academic standards. The supporting evidence in this paper is strongly reliable and properly validated and the introduction provides the necessary background information.\nHowever, I have a few and minor comments that need to be addressed:\nThe keywords should be amended based on Mesh: The private sector, Public-Private Partnerships, Nutritional support, Intervention, Food, Fortification, Malnutrition, Eastern Mediterranean.\n\nNeeds to add a table of the collected studies and the extracted data.\n\nNo need for the footnotes in page 5, as they have already presented in table 1.\n\nSo, by adding the acronyms into the table you could use them in the following paragraphs.\n\nIn paragraph 3 of page 7: Iran has also enacted reformulation policies to reduce salt in traditional bread by involving the private sectors, and trans-fatty acids in some food products and, sugar (especially in beverages).\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6818",
"date": "21 Jun 2021",
"name": "Arabia Ali",
"role": "Author Response",
"response": "Majid Hajifaraji Thank you for your constructive comments, the dataset (table of collected studies) are already available here: https://datadryad.org/stash/dataset/doi:10.5061/dryad.dv41ns1ww Unfortunately, we couldn’t have access to the evidence that supports your point on the following: In paragraph 3 of page 7: Iran has also enacted reformulation policies to reduce salt in traditional bread by involving the private sectors, and trans-fatty acids in some food products and, sugar (especially in beverages). Thank you."
},
{
"c_id": "6819",
"date": "21 Jun 2021",
"name": "Arabia Ali",
"role": "Author Response",
"response": "Dear Dr. Majid I just want to inform you that a new version was submitted taking in response to your comments. Thank you and best regards."
}
]
}
] | 1
|
https://f1000research.com/articles/10-17
|
https://f1000research.com/articles/10-16/v1
|
12 Jan 21
|
{
"type": "Review",
"title": "Therapeutic implications of statins in heart failure with reduced ejection fraction and heart failure with preserved ejection fraction: a review of current literature",
"authors": [
"Chol Techorueangwiwat",
"Chanavuth Kanitsoraphan",
"Panupong Hansrivijit",
"Chol Techorueangwiwat",
"Chanavuth Kanitsoraphan"
],
"abstract": "Statins are one of the standard treatments to prevent cardiovascular events such as coronary artery disease and heart failure (HF). However, data on the use of statins to improve clinical outcomes in patients with established HF remains controversial. We summarized available clinical studies which investigated the effects of statins on clinical outcomes in patients with HF with reduced ejection fraction (HFrEF) and HF with preserved ejection fraction (HFpEF). Statins possess many pleiotropic effects in addition to lipid-lowering properties that positively affect the pathophysiology of HF. In HFrEF, data from two large randomized placebo-controlled trials did not show benefits of statins on mortality of patients with HFrEF. However, more recent prospective cohort studies and meta-analyses have shown decreased risk of mortality as well as cardiovascular hospitalization with statins treatment. In HFpEF, most prospective and retrospective cohort studies as well as meta analyses have consistently reported positive effects of statins, including reducing mortality and improving other clinical outcomes. Current evidence also suggests better outcomes with lipophilic statins in patients with HF. In summary, statins might be effective in improving survival and other clinical outcomes in patients with HF, especially for patients with HFpEF. Lipophilic statins might also be more beneficial for HF patients. Based on current evidence, statins did not cause harm and should be continued in HF patients who are already taking the medication. Further randomized controlled trials are needed to clarify the benefits of statins in HF patients.",
"keywords": [
"statins",
"heart failure",
"mortality",
"HFrEF",
"HFpEF",
"prevention",
"HMG-CoA inhibitor"
],
"content": "Introduction\n\nStatins are blood-cholesterol-lowering drugs that are used worldwide to prevent cardiovascular diseases. They inhibit the 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) enzyme and thereby reduce the synthesis of cholesterol in the liver1. Statins also upregulate the expression of low-density lipoprotein (LDL) receptors on the cell membrane2. Clinically, statins can reduce LDL by 20–60% and triglyceride by 10-40%, and increase high-density lipoprotein (HDL) by 5%-15%3–5. Statins have a dramatic effect on reducing cardiovascular events and the data from one clinical trial has shown that the risk for major vascular events is reduced by 22% for each 1 mmol/L reduction in LDL6.\n\nThe role of statins in the primary and secondary prevention of coronary artery disease is well established7–9, irrespective of patients’ cholesterol level10. Robust evidence also exists supporting the role of statins in the prevention of new incident heart failure (HF)11. Given their many pleiotropic effects, statins were hypothesized to also be effective in improving outcomes in patients with established HF12. However, the evidence for the use of statins in patients with HF is controversial13–17, so the role of statins in this population remains unclear. Current guidelines have not recommended the routine use of statins in most patients with HF without other indications for their use (e.g. coronary artery disease, hypercholesterolemia), but suggest that statins can be continued for patients who were already on treatment18. More recent evidence is emerging suggesting potential benefits of statins in patients with HF to improve clinical outcomes19–22.\n\nIn this review, we summarized potential mechanisms of statins that might be beneficial for patients with HF and available clinical studies which investigated the effects of statins on clinical outcomes in patients with HF with reduced ejection fraction (HFrEF) and HF with preserved ejection fraction (HFpEF).\n\n\nPotential mechanisms of statins in heart failure\n\nMultiple mechanisms have been described to explain the potential benefits, as well as harms, of statins in HF. Various experimental and clinical studies have described pleiotropic effects of statins independent of their lipid-lowering ability that might improve the outcomes in patients with HF23,24. Statins stabilize atheromatous plaques and possess antiatherogenic properties that help reduce atheroma volume and prevent formation of new atherosclerotic lesions, which in turn reduces ischemic burden and results in less cardiac tissue damage11,25. Statins promote left ventricular healing after myocardial infarction by improving left ventricular remodeling26 and modulating endothelial function27. Statins have also been found to possess anti-inflammatory properties28–30 leading to the reduction of oxidative stress in vascular and myocardial tissues31,32. Other important mechanisms of statins have been reported, including enhancement of endothelial function33, inhibition of the action of angiotensin II34, reduction of sympathetic activities35, delaying apoptosis32, and suppression of arrhythmogenesis36. In addition, statins may exert their pleiotropic effects through epigenetic modifications as they can act as histone deacetylase inhibitors37,38. Despite two large randomized controlled trials (RCTs) exploring the effects of statins in systolic HF showing that statins did not improve all-cause mortality or other major cardiovascular outcomes13,14, some meta-analyses and observational studies have shown the contrary15,20,22,39. This is discussed in detail below for HFrEF and HFpEF, respectively.\n\nThere are two types of statins: hydrophilic (e.g. rosuvastatin, pravastatin) and lipophilic (e.g. atorvastatin, simvastatin). Each type has different properties. Hydrophilic statins employ active transport into hepatocytes, whereas lipophilic statins enter cells by passive diffusion and therefore are more easily absorbed by peripheral tissues, including myocardial cells40–42. Current evidence suggest that the benefits of statins also depend on their lipophilicity, thus the effects in populations with HF should not be expected to be identical across all statin drug class41,43,44.\n\nMultiple mechanisms have also been proposed that raise concerns for the adverse effects of statins in patients with HF. Low serum total cholesterol has been associated with higher mortality in patients with HF for instance45–50; however, it is unclear if low cholesterol is an independent risk factor for poor outcomes or is merely a surrogate marker for cardiac cachexia in patients with more severe disease51. It has also been hypothesized that hepatic and intestinal congestion in patients with HF can impair hepatic cholesterol synthesis and cholesterol absorption, respectively, therefore resulting in lower plasma cholesterol level46,52. In a South Korean cohort study analyzing 2,797 patients with HF, those with lower serum total cholesterol had more markers of severe diseases (e.g. lower blood pressure, lower serum sodium), but low cholesterol itself did not affect the clinical outcomes in a propensity score matched analysis53. Another mechanism through which statins might adversely affect HF is that statins have been shown to reduce serum coenzyme Q10 level, an important enzyme in myocardial bioenergetics54. Studies have reported that low serum coenzyme Q10 level was associated with increased mortality in patients with HF55,56. Nevertheless, two large randomized trials have shown that statin use was not associated with significant worsening of clinical outcomes or adverse events in HF populations13,14.\n\nIn this non-systematic review, we searched Ovid MEDLINE for randomized controlled trials on the use of statins in heart failure patients. The review covers from current clinical evidence to pathophysiology review accrued from bench research. The search terms used are depicted in Table 1.\n\n\nStatins in heart failure with reduced ejection fraction: current evidence\n\nA list of relevant clinical studies on the use of statins in HFpEF patients is presented in Table 2. Of these studies, the best evidence to date comes from two large randomized placebo-controlled trials: CORONA14 and GISSI-HF13. Neither of these trials found benefits of statin therapy in patients with HF. The CORONA trial recruited 5,011 patients with ischemic HFrEF (mean age 73 years and mean left ventricular ejection fraction [LVEF] 31%) and randomized the subjects to either rosuvastatin 10 mg daily or a placebo. After the median follow-up of 32.8 months, there was no significant reduction in the primary endpoints, including the composite of death from cardiovascular causes, nonfatal myocardial infarction, and nonfatal stroke, and no significant difference between groups14. However, a post-hoc analysis of this trial showed that, when comparing outcomes by N-terminal pro-B-type natriuretic peptide (NT-proBNP) tertile, patients in the lowest NT-proBNP tertile (< 103 pmol/L) had a significantly lower risk of the primary outcomes57. In addition, rosuvastatin also reduced the incidence of hospitalization for cardiovascular causes which has also been found by other researchers14,58. Secondly, the GISSI-HF trial randomized 4,574 patients with HF from any cause (mean age 68 years, mean LVEF 33%) to receive either 10 mg daily of rosuvastatin or a placebo. Similar to the result of the CORONA trial, there were no significant differences in the co-primary outcomes of all-cause mortality or in the combined endpoints of death or hospitalization for cardiovascular causes13.\n\nDespite the above results, certain concerns regarding the generalizability of the trial results to the entire HFrEF population have been raised. The study population in both trials had a mean age of 73 years and most of the participants were already in advanced HF stages, namely New York Heart Association classification III and IV14,65. The observation from the CORONA trial that the patients in the lowest NT-proBNP tertile had a significantly lower risk of the primary adverse outcomes suggested that patients with less severe HF might benefit from statin therapy57. Another study also found that the benefits of simvastatin toward reduction in major vascular events was relatively smaller in patients with higher NT-proBNP level66. Moreover, the results of the GISSI-trial might have been impacted by compliance issues, as about one third of the study participants discontinued therapy for various reasons13. Furthermore, the CORONA and GISSI-HF trials also assessed rosuvastatin at a low dose, and therefore the results cannot be inferred to a higher dose or different types of statins. It is worth recalling that rosuvastatin is a hydrophilic statin, which has poor uptake by cardiac muscles12. On the other hand, lipophilic statins (e.g. atorvastatin, simvastatin, pitavastatin), which have better penetration into cardiac muscle cells and therefore possibly better influence the myocardium through pleiotropic effects12,40,67, leading to the improvement in outcomes of patients with HF20,43,44,68,69. In addition, other small randomized studies have shown that statins can improve certain surrogate endpoints in HF (e.g. LVEF, BNP, inflammatory markers), but not major outcomes (e.g. mortality, cardiovascular hospitalization), although it appears that most of the studies might not be adequately powered to assess major clinical outcomes12,30,70–77.\n\nA more recent, albeit observational, nationwide prospective study, propensity score matched 21,864 patients with HFrEF from 86% of all Swedish hospitals. Statins were associated with reduced all-cause mortality (HR 0.81; 95% CI 0.76-0.686; p < 0.001), cardiovascular mortality, cardiovascular hospitalization, combined all-cause mortality, and HF hospitalization22. The results concur with other observational studies that were reported both before and after the publications of CORONA and GISSI-HF trials15,59,62,78–80. In addition, a recent meta-analysis20 including 88,100 patients from 17 studies (2 randomized controlled trials and 15 cohorts) found that statin therapy was associated with reduced all-cause mortality and other secondary endpoints in patients with HF, similarly to other observational studies. The benefit was also consistent in HFrEF patients in the subgroup analysis20. It has been suggested that the reason for the different results between large randomized trials and real-life cohort studies is that the randomized cohorts might not be representative of those in clinical practice as well as the dose and type of statins chosen in randomized trials, among other reasons15,22,25. Therefore, additional randomized trials using statins other than rosuvastatin with more generalized inclusion in HFrEF populations may be warranted.\n\n\nStatins in heart failure with preserved ejection fraction: current evidence\n\nUnlike HFrEF, no large RCTs have been conducted to assess the benefits of statins in patients with HFpEF. However, many studies have suggested the benefits of statins in this population20,21,39,81–86. A list of relevant clinical studies of statins in patients with HFpEF is presented in Table 3. In a preliminary report of 137 consecutive patients with HF with LVEF ≥ 50% where half of the patients received statins, statin therapy was associated with lower mortality (RR 0.22; 95% CI 0.07-0.64; p = 0.006)82. An observational prospective study following 9,140 patients with HFpEF in the Swedish Heart Failure Registry found that statin therapy was associated with decreased 1-year all-cause mortality (HR, 0.80; 95% CI: 0.72-0.89; P < 0.001), as well as decreased cardiovascular and composite all-cause mortality or cardiovascular hospitalization81. Another multicenter prospective observational study in Japan with 3,124 patients with HFpEF also showed a similar decrease in mortality in patients with statins at three years (adjusted HR 0.74; 95% CI; 0.58-0.94; p < 0.001)83. In the United States, a retrospective cohort study of 13,440 patients with HF, of which 7,563 had HFpEF (LVEF ≥ 50%) showed that statin use was associated with decreased mortality (HR 0.73; 95% CI 0.66-0.81; p < 0.001) in patients with HFpEF but not in patients with HFrEF or mid-range ejection fraction (HFmrEF)84. Several meta-analyses have been conducted to assess the effects of statins on mortality in patients with HFpEF and have suggested similar findings of reduced mortality rate20,39,85,86.\n\nThe studies mentioned above had included patients with coronary artery disease. In a most recent multicenter observational prospective cohort study in Japan evaluating 414 HFpEF patients without coronary artery disease, the use of statins was associated with a substantial decrease in 3-year mortality both in the entire and propensity score-matched cohorts21. Moreover, the subgroup analyses also showed consistent benefits of statins across all range of blood cholesterol and HF severity21. This study further suggested that statins might be beneficial for HFpEF patients even in those without coronary artery disease. Interestingly, it has been suggested that the findings where benefits of statins were consistently observed in HFpEF whereas in HFrEF the results were mixed might be due to greater effects of pleiotropic properties of statins on the pathophysiology of HFpEF87,88 and patients’ comorbidities83,88. While these findings appear promising, it is important to emphasize that these observational studies are only hypothesis-generating. Well-designed RCTs are needed to confirm the effects of statin therapy in HFpEF patients.\n\nPossible unique pleiotropic effects of statins in patients with HFpEF have not been clearly defined. Statins may help prevent left ventricular hypertrophy and fibrosis in HFpEF21,89,90. Experimental studies have shown that statins can improve diastolic dysfunction95,96 and attenuated left ventricle stiffness97, both of which are important underlying pathogeneses of HFpEF. Anti-inflammatory and anti-oxidant effects of statins may be beneficial in HFpEF as systemic inflammatory state plays a role in the pathogenesis of HFpEF87. The pleiotropic effects of statins are illustrated in Figure 1. Pathogenesis of HFpEF may also be driven by several comorbidities such as overweight/obesity, diabetes mellitus, renal dysfunction, and hypertension87, as current evidence suggests that statins have been shown to be beneficial in these conditions as well98–100. Lipophilic statins may have a favorable effect on sympathetic activity in HFpEF, as increased sympathetic activity is associated with the pathogenesis of HFpEF101. Further mechanistic studies focusing on the underlying cellular and molecular changes in HFpEF caused by statins are needed. Possible roles of collagen synthesis, matrix metalloproteinases inhibition, Rho kinase 1 gene expression in relation to diastolic dysfunction in HFpEF may be an area of further studies96. The mechanism of stains and regulation of the Rho GTPase cycle is illustrated in Figure 2.\n\nStatins causes inhibition of membrane translocation of small g proteins, such as Rac and Rho. Inhibition of Rac pathway leads to the reduction of NADPH oxidase which results in less cardiac damage including cardiac myocyte hypertrophy and apoptosis. Similarly, inhibition of Rho pathway causes increased endothelial nitric oxide synthase (NOS) production leading to lower subclinical ischemia by reducing interstitial fibrosis, vascular function, and vascularization. Both mechanisms are proposed to prevent heart failure.\n\nWithout statins, HMG-CoA is converted to mevalonate and subsequently geranylgeranyl (GG) pyrophosphate. The GG component is added to a complex of Rho protein and guanine nucleotide dissociation inhibitors (GDI) and therefore activates the Rho kinase (ROCK) pathway. ROCK mediates the downstream effects of Rho and has effect on endothelial cells, white blood cells, smooth muscle cells, and cardiac myocytes. The effect of statins inhibits the conversion of HMG-CoA to GG pyrophosphate which results in downregulation of ROCK and its associated effects on different target cells.\n\n\nOther lipid lowering medications and their role in heart failure\n\nProprotein convertase subtilisin–kexin type 9 inhibitors (PCSK9i) have not been shown to reduce death or hospitalizations from worsening HF in two large RCTs102,103. Icosapent ethyl, a form of highly purified omega-3 eicosapentaenoic acid ethyl ester (EPA), is used to lower triglyceride levels. EPA has been shown to effectively reduce ischemic events and cardiovascular death among patients who have high triglyceride levels and are already on statin therapy104. EPA may also reduce new HF hospitalization if a high level of on-treatment EPA in the blood is achieved105.\n\n\nConclusions and future perspectives\n\nThe current evidence we have for statins in the treatment of established HF is far from satisfactory. Our review shows contradicting evidence for statin therapy in HFrEF. Despite negative results from two large RCTs, many real word data do suggest benefits of statins in HFrEF. Thus, an adequately powered randomized placebo-controlled trial to determine the effects of statins other than rosuvastatin with more generalized inclusion criteria, preferably with long term follow up, is needed to clarify the benefits of statins in HFrEF. On the other hand, the benefits of statins in improving clinical outcomes among HFpEF patients were consistently reported by observational studies both in the United States and international institutions, even in patients without coronary artery disease. Nevertheless, there exists a dire need for new treatment options for patients with HFpEF as the current choice of pharmacological treatments is very limited. A future RCT comparing the treatment outcomes of statin therapy in HFpEF is also needed and would contribute a significant impact in closing the knowledge gap of HF management. Moreover, a direct head-to-head comparison of hydrophilic versus lipophilic statins in HF will also help elucidate the hypothesis that lipophilic statins might be more appropriate for patients with HF. To generate this much needed evidence, artificial intelligence - a rapidly developing field in medicine - has the potential to be incorporated in the design and execution of future RCTs106. Application of artificial intelligence may improve efficacy of RCTs by improving the patient’s selection process, minimizing measurement errors when assessing endpoints, or even providing trials with synthetic control groups107.\n\nMore recently, network medicine is a novel discipline that studies and integrates heterogenous interconnected molecular and genetic data as a network and identifies perturbations in these networks that ultimately causes disease108,109. This concept has been applied in cardiovascular medicine and may also provide the insights into spatio-temporal statin-mediated mechanisms of statins in patients with HF. For example, a group of investigators has conducted a network analysis integrating myocardial infarction drugs, drugs interactors, drug targets, and myocardial infarction disease genes onto the human interactome110. By integrating gene-disease associations, they were able to identify “drug-target-disease modules”, which provide a better understanding of the drug actions and mechanisms110. This approach should also be pursued to elucidate the mechanisms of statins in patients with HF in addition to the previously existing clinical evidence from RCTs.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.",
"appendix": "Author contributions\n\n\n\nC.T. performed the systematic search. C.T. and C.K. extracted the included articles. C.T. and P.H. drafted the manuscript. P.H. constructed the figures. C.T. and C.K. constructed the tables. C.T. and P.H. revised the manuscript prior to publication.\n\n\nReferences\n\nBrown MS, Goldstein JL: Multivalent feedback regulation of HMG CoA reductase, a control mechanism coordinating isoprenoid synthesis and cell growth. J Lipid Res. 1980; 21(5): 505–17. PubMed Abstract\n\nBrown MS, Goldstein JL: Lowering plasma cholesterol by raising LDL receptors. N Engl J Med. 1981; 305(9): 515–7. PubMed Abstract | Publisher Full Text\n\nKazi DS, Penko JM, Bibbins-Domingo K: Statins for Primary Prevention of Cardiovascular Disease: Review of Evidence and Recommendations for Clinical Practice. Med Clin North Am. 2017; 101(4): 689–99. PubMed Abstract | Publisher Full Text\n\nOdden MC, Pletcher MJ, Coxson PG, et al.: Cost-effectiveness and population impact of statins for primary prevention in adults aged 75 years or older in the United States. Ann Intern Med. 2015; 162(8): 533–41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNiazi M, Galehdar N, Jamshidi M, et al.: A Review of the Role of Statins in Heart Failure Treatment. Curr Clin Pharmacol. 2020; 15(1): 30–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCholesterol Treatment Trialists’ (CTT) Collaboration, Baigent C, Blackwell L, et al.: Efficacy and safety of more intensive lowering of LDL cholesterol: a meta-analysis of data from 170,000 participants in 26 randomised trials. Lancet. 2010; 376(9753): 1670–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRandomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet. 1994; 344(8934): 1383–9. PubMed Abstract\n\nLong-Term Intervention with Pravastatin in Ischaemic Disease Study G: Prevention of cardiovascular events and death with pravastatin in patients with coronary heart disease and a broad range of initial cholesterol levels. N Engl J Med. 1998; 339(19): 1349–57. PubMed Abstract | Publisher Full Text\n\nCholesterol Treatment Trialists' (CTT) Collaboration, Fulcher J, O'Connell R, et al.: Efficacy and safety of LDL-lowering therapy among men and women: meta-analysis of individual data from 174,000 participants in 27 randomised trials. Lancet. 2015; 385(9976): 1397–405. PubMed Abstract | Publisher Full Text\n\nSacks FM, Pfeffer MA, Moye LA, et al.: The effect of pravastatin on coronary events after myocardial infarction in patients with average cholesterol levels. Cholesterol and Recurrent Events Trial investigators. N Engl J Med. 1996; 335(14): 1001–9. PubMed Abstract | Publisher Full Text\n\nLee MMY, Sattar N, McMurray JJV, et al.: Statins in the Prevention and Treatment of Heart Failure: a Review of the Evidence. Curr Atheroscler Rep. 2019; 21(10): 41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBonsu KO, Kadirvelu A, Reidpath DD: Statins in heart failure: do we need another trial? Vasc Health Risk Manag. 2013; 9: 303–19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTavazzi L, Maggioni AP, Marchioli R, et al.: Effect of rosuvastatin in patients with chronic heart failure (the GISSI-HF trial): a randomised, double-blind, placebo-controlled trial. Lancet. 2008; 372(9645): 1231–9. PubMed Abstract | Publisher Full Text\n\nKjekshus J, Apetrei E, Barrios V, et al.: Rosuvastatin in older patients with systolic heart failure. N Engl J Med. 2007; 357(22): 2248–61. PubMed Abstract | Publisher Full Text\n\nGastelurrutia P, Lupon J, de Antonio M, et al.: Statins in heart failure: the paradox between large randomized clinical trials and real life. Mayo Clin Proc. 2012; 87(6): 555–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaison P, Desamericq G, Hemery F, et al.: Relationship between recommended chronic heart failure treatments and mortality over 8 years in real-world conditions: a pharmacoepidemiological study. Eur J Clin Pharmacol. 2013; 69(4): 901–8. PubMed Abstract | Publisher Full Text\n\nThambidorai SK, Deshmukh AR, Walters RW, et al.: Impact of statin use on heart failure mortality. Int J Cardiol. 2011; 147(3): 438–43. PubMed Abstract | Publisher Full Text\n\nMach F, Baigent C, Catapano AL, et al.: 2019 ESC/EAS Guidelines for the management of dyslipidaemias: lipid modification to reduce cardiovascular risk: The Task Force for the management of dyslipidaemias of the European Society of Cardiology (ESC) and European Atherosclerosis Society (EAS). European Heart Journal. 2019; 41(1): 111–88. Publisher Full Text\n\nBonsu KO, Owusu IK, Buabeng KO, et al.: Statin Treatment and Clinical Outcomes of Heart Failure Among Africans: An Inverse Probability Treatment Weighted Analysis. J Am Heart Assoc. 2017; 6(4): e004706. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBielecka-Dabrowa A, Bytyci I, Von Haehling S, et al.: Association of statin use and clinical outcomes in heart failure patients: a systematic review and meta-analysis. Lipids Health Dis. 2019; 18(1): 188. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarume K, Takashio S, Nagai T, et al.: Effect of Statins on Mortality in Heart Failure With Preserved Ejection Fraction Without Coronary Artery Disease- Report From the JASPER Study. Circ J. 2019; 83(2): 357–67. PubMed Abstract | Publisher Full Text\n\nAlehagen U, Benson L, Edner M, et al.: Association between use of statins and outcomes in heart failure with reduced ejection fraction: prospective propensity score matched cohort study of 21 864 patients in the Swedish Heart Failure Registry. Circ Heart Fail. 2015; 8(2): 252–60. PubMed Abstract | Publisher Full Text\n\nCorreale M, Abruzzese S, Greco CA, et al.: Pleiotropic effects of statin in therapy in heart failure: a review. Curr Vasc Pharmacol. 2014; 12(6): 873–84. PubMed Abstract | Publisher Full Text\n\nArslan F, Pasterkamp G, de Kleijn DP: Unraveling pleiotropic effects of statins: bit by bit, a slow case with perspective. Circ Res. 2008; 103(4): 334–6. PubMed Abstract | Publisher Full Text\n\nMasoudi FA: Statins for ischemic systolic heart failure. N Engl J Med. 2007; 357(22): 2301–4. PubMed Abstract | Publisher Full Text\n\nBauersachs J, Galuppo P, Fraccarollo D, et al.: Improvement of left ventricular remodeling and function by hydroxymethylglutaryl coenzyme a reductase inhibition with cerivastatin in rats with heart failure after myocardial infarction. Circulation. 2001; 104(9): 982–5. PubMed Abstract | Publisher Full Text\n\nLeenders GJ, Smeets MB, van den Boomen M, et al.: Statins Promote Cardiac Infarct Healing by Modulating Endothelial Barrier Function Revealed by Contrast-Enhanced Magnetic Resonance Imaging. Arterioscler Thromb Vasc Biol. 2018; 38(1): 186–94. PubMed Abstract | Publisher Full Text\n\nAlbert MA, Danielson E, Rifai N, et al.: Effect of statin therapy on C-reactive protein levels: the pravastatin inflammation/CRP evaluation (PRINCE): a randomized trial and cohort study. JAMA. 2001; 286(1): 64–70. PubMed Abstract | Publisher Full Text\n\nRay KK, Cannon CP, Cairns R, et al.: Relationship between uncontrolled risk factors and C-reactive protein levels in patients receiving standard or intensive statin therapy for acute coronary syndromes in the PROVE IT-TIMI 22 trial. J Am Coll Cardiol. 2005; 46(8): 1417–24. PubMed Abstract | Publisher Full Text\n\nSola S, Mir MQS, Lerakis S, et al.: Atorvastatin improves left ventricular systolic function and serum markers of inflammation in nonischemic heart failure. J Am Coll Cardiol. 2006; 47(2): 332–7. PubMed Abstract | Publisher Full Text\n\nBrown JH, Del Re DP, Sussman MA: The Rac and Rho hall of fame: a decade of hypertrophic signaling hits. Circ Res. 2006; 98(6): 730–42. PubMed Abstract | Publisher Full Text\n\nSchleicher E, Friess U: Oxidative stress, AGE, and atherosclerosis. Kidney Int Suppl. 2007; (106): S17–26. PubMed Abstract | Publisher Full Text\n\nGounari P, Tousoulis D, Antoniades C, et al.: Rosuvastatin but not ezetimibe improves endothelial function in patients with heart failure, by mechanisms independent of lipid lowering. Int J Cardiol. 2010; 142(1): 87–91. PubMed Abstract | Publisher Full Text\n\nHsieh CC, Li CY, Hsu CH, et al.: Mitochondrial protection by simvastatin against angiotensin II-mediated heart failure. Br J Pharmacol. 2019; 176(19): 3791–804. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRosenson RS, Tangney CC, Casey LC: Inhibition of proinflammatory cytokine production by pravastatin. Lancet. 1999; 353(9157): 983–4. PubMed Abstract | Publisher Full Text\n\nTousoulis D, Oikonomou E, Siasos G, et al.: Statins in heart failure--With preserved and reduced ejection fraction. An update. Pharmacol Ther. 2014; 141(1): 79–91. PubMed Abstract | Publisher Full Text\n\nAllen SC, Mamotte CDS: Pleiotropic and Adverse Effects of Statins-Do Epigenetics Play a Role? J Pharmacol Exp Ther. 2017; 362(2): 319–26. PubMed Abstract | Publisher Full Text\n\nOesterle A, Laufs U, Liao JK: Pleiotropic Effects of Statins on the Cardiovascular System. Circ Res. 2017; 120(1): 229–43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang JQ, Wu GR, Wang Z, et al.: Long-term clinical outcomes of statin use for chronic heart failure: a meta-analysis of 15 prospective studies. Heart Lung Circ. 2014; 23(2): 105–13. PubMed Abstract | Publisher Full Text\n\nSchachter M: Chemical, pharmacokinetic and pharmacodynamic properties of statins: an update. Fundam Clin Pharmacol. 2005; 19(1): 117–25. PubMed Abstract | Publisher Full Text\n\nFord JS, Tayek JA: Lipophilicity and cardiovascular outcome in patients with CHF. Am Heart J. 2008; 156(2): e7; author reply e9. PubMed Abstract | Publisher Full Text\n\nNezasa KI, Higaki K, Matsumura T, et al.: Liver-specific distribution of rosuvastatin in rats: comparison with pravastatin and simvastatin. Drug Metab Dispos. 2002; 30(11): 1158–63. PubMed Abstract | Publisher Full Text\n\nBonsu KO, Reidpath DD, Kadirvelu A: Lipophilic Statin Versus Rosuvastatin (Hydrophilic) Treatment for Heart Failure: a Meta-Analysis and Adjusted Indirect Comparison of Randomised Trials. Cardiovasc Drugs Ther. 2016; 30(2): 177–88. PubMed Abstract | Publisher Full Text\n\nLipinski MJ, Cauthen CA, Biondi-Zoccai GGL, et al.: Meta-analysis of randomized controlled trials of statins versus placebo in patients with heart failure. Am J Cardiol. 2009; 104(12): 1708–16. PubMed Abstract | Publisher Full Text\n\nCatapano AL, Graham I, De Backer G, et al.: 2016 ESC/EAS Guidelines for the Management of Dyslipidaemias. Eur Heart J. 2016; 37(39): 2999–3058. PubMed Abstract | Publisher Full Text\n\nAnanaba I, Taegtmeyer H: Low serum cholesterol as prognostic indicator in heart failure. J Card Fail. 2012; 18(7): 596. PubMed Abstract | Publisher Full Text\n\nRauchhaus M, Clark AL, Doehner W, et al.: The relationship between cholesterol and survival in patients with chronic heart failure. J Am Coll Cardiol. 2003; 42(11): 1933–40. PubMed Abstract | Publisher Full Text\n\nAfsarmanesh N, Horwich TB, Fonarow GC: Total cholesterol levels and mortality risk in nonischemic systolic heart failure. Am Heart J. 2006; 152(6): 1077–83. PubMed Abstract | Publisher Full Text\n\nHorwich TB, Hernandez AF, Dai D, et al.: Cholesterol levels and in-hospital mortality in patients with acute decompensated heart failure. Am Heart J. 2008; 156(6): 1170–6. PubMed Abstract | Publisher Full Text\n\nWeiss A, Beloosesky Y, Schmilovitz-Weiss H, et al.: Serum total cholesterol: a mortality predictor in elderly hospitalized patients. Clin Nutr. 2013; 32(4): 533–7. PubMed Abstract | Publisher Full Text\n\nAraujo JP, Frioes F, Azevedo A, et al.: Cholesterol--a marker of nutritional status in mild to moderate heart failure. Int J Cardiol. 2008; 129(1): 65–8. PubMed Abstract | Publisher Full Text\n\nChen Y, He XM, Meng H, et al.: Relationship between lipids levels and right ventricular volume overload in congestive heart failure. J Geriatr Cardiol. 2014; 11(3): 192–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYoon CH, Youn TJ, Ahn S, et al.: Low serum total cholesterol level is a surrogate marker, but not a risk factor, for poor outcome in patients hospitalized with acute heart failure: a report from the Korean Heart Failure Registry. J Card Fail. 2012; 18(3): 194–201. PubMed Abstract | Publisher Full Text\n\nKloer HU, Belardinelli R, Ruchong O, et al.: Combining Ubiquinol With a Statin May Benefit Hypercholesterolaemic Patients With Chronic Heart Failure. Heart Lung Circ. 2020; 29(2): 188–95. PubMed Abstract | Publisher Full Text\n\nRosenfeldt F, Hilton D, Pepe S, et al.: Systematic review of effect of coenzyme Q10 in physical exercise, hypertension and heart failure. Biofactors. 2003; 18(1–4): 91–100. PubMed Abstract | Publisher Full Text\n\nOkello E, Jiang X, Mohamed S, et al.: Combined statin/coenzyme Q10 as adjunctive treatment of chronic heart failure. Med Hypotheses. 2009; 73(3): 306–8. PubMed Abstract | Publisher Full Text\n\nCleland JG, McMurray JJ, Kjekshus J, et al.: Plasma concentration of amino-terminal pro-brain natriuretic peptide in chronic heart failure: prediction of cardiovascular events and interaction with the effects of rosuvastatin: a report from CORONA (Controlled Rosuvastatin Multinational Trial in Heart Failure). J Am Coll Cardiol. 2009; 54(20): 1850–9. PubMed Abstract | Publisher Full Text\n\nRogers JK, Jhund PS, Perez AC, et al.: Effect of rosuvastatin on repeat heart failure hospitalizations: the CORONA Trial (Controlled Rosuvastatin Multinational Trial in Heart Failure). JACC Heart Fail. 2014; 2(3): 289–97. PubMed Abstract | Publisher Full Text\n\nHorwich TB, MacLellan WR, Fonarow GC: Statin therapy is associated with improved survival in ischemic and non-ischemic heart failure. J Am Coll Cardiol. 2004; 43(4): 642–8. PubMed Abstract | Publisher Full Text\n\nSola S, Mir MQS, Rajagopalan S, et al.: Statin therapy is associated with improved cardiovascular outcomes and levels of inflammatory markers in patients with heart failure. J Card Fail. 2005; 11(8): 607–12. PubMed Abstract | Publisher Full Text\n\nHong YJ, Jeong MH, Hyun DW, et al.: Prognostic significance of simvastatin therapy in patients with ischemic heart failure who underwent percutaneous coronary intervention for acute myocardial infarction. Am J Cardiol. 2005; 95(5): 619–22. PubMed Abstract | Publisher Full Text\n\nGo AS, Lee WY, Yang J, et al.: Statin therapy and risks for death and hospitalization in chronic heart failure. JAMA. 2006; 296(17): 2105–11. PubMed Abstract | Publisher Full Text\n\nHuan Loh P, Windram JD, Tin L, et al.: The effects of initiation or continuation of statin therapy on cholesterol level and all-cause mortality after the diagnosis of left ventricular systolic dysfunction. Am Heart J. 2007; 153(4): 537–44. PubMed Abstract | Publisher Full Text\n\nColeman CI, Kluger J, Bhavnani S, et al.: Association between statin use and mortality in patients with implantable cardioverter-defibrillators and left ventricular systolic dysfunction. Heart Rhythm. 2008; 5(4): 507–10. PubMed Abstract | Publisher Full Text\n\nFonarow GC: Randomized clinical outcome trials of statins in heart failure. Heart Fail Clin. 2008; 4(2): 225–9. PubMed Abstract | Publisher Full Text\n\nHeart Protection Study Collaborative Group, Emberson JR, Ng LL, et al.: N-terminal Pro-B-type natriuretic peptide, vascular disease risk, and cholesterol reduction among 20,536 patients in the MRC/BHF heart protection study. J Am Coll Cardiol. 2007; 49(3): 311–9. PubMed Abstract | Publisher Full Text\n\nWierzbicki AS, Poston R, Ferro A: The lipid and non-lipid effects of statins. Pharmacol Ther. 2003; 99(1): 95–112. PubMed Abstract | Publisher Full Text\n\nLiu G, Zheng XX, Xu YL, et al.: Effects of lipophilic statins for heart failure: a meta-analysis of 13 randomised controlled trials. Heart Lung Circ. 2014; 23(10): 970–7. PubMed Abstract | Publisher Full Text\n\nZvizdic F, Godinjak A, Durak-Nalbantic A, et al.: Impact of Different Types of Statins on Clinical Outcomes in Patients Hospitalized for Ischemic Heart Failure. Med Arch. 2018; 72(6): 401–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVrtovec B, Okrajsek R, Golicnik A, et al.: Atorvastatin therapy may reduce the incidence of sudden cardiac death in patients with advanced chronic heart failure. J Card Fail. 2008; 14(2): 140–4. PubMed Abstract | Publisher Full Text\n\nWojnicz R, Wilczek K, Nowalany-Kozielska E, et al.: Usefulness of atorvastatin in patients with heart failure due to inflammatory dilated cardiomyopathy and elevated cholesterol levels. Am J Cardiol. 2006; 97(6): 899–904. PubMed Abstract | Publisher Full Text\n\nXie RQ, Cui W, Liu F, et al.: Statin therapy shortens QTc QTcd, and improves cardiac function in patients with chronic heart failure. Int J Cardiol. 2010; 140(2): 255–7. PubMed Abstract | Publisher Full Text\n\nNode K, Fujita M, Kitakaze M, et al.: Short-term statin therapy improves cardiac function and symptoms in patients with idiopathic dilated cardiomyopathy. Circulation. 2003; 108(7): 839–43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTousoulis D, Andreou I, Tentolouris C, et al.: Comparative effects of rosuvastatin and allopurinol on circulating levels of matrix metalloproteinases and tissue inhibitors of metalloproteinases in patients with chronic heart failure. Int J Cardiol. 2010; 145(3): 438–43. PubMed Abstract | Publisher Full Text\n\nTsutamoto T, Sakai H, Ibe K, et al.: Effect of atorvastatin vs. rosuvastatin on cardiac sympathetic nerve activity in non-diabetic patients with dilated cardiomyopathy. Circ J. 2011; 75(9): 2160–6. PubMed Abstract | Publisher Full Text\n\nAndreou I, Tousoulis D, Miliou A, et al.: Effects of rosuvastatin on myeloperoxidase levels in patients with chronic heart failure: a randomized placebo-controlled study. Atherosclerosis. 2010; 210(1): 194–8. PubMed Abstract | Publisher Full Text\n\nHorwich TB, Middlekauff HR, Maclellan WR, et al.: Statins do not significantly affect muscle sympathetic nerve activity in humans with nonischemic heart failure: a double-blind placebo-controlled trial. J Card Fail. 2011; 17(11): 879–86. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKrum H, Latini R, Maggioni AP, et al.: Statins and symptomatic chronic systolic heart failure: a post-hoc analysis of 5010 patients enrolled in Val-HeFT. Int J Cardiol. 2007; 119(1): 48–53. PubMed Abstract | Publisher Full Text\n\nOuzounian M, Tu JV, Austin PC, et al.: Statin therapy and clinical outcomes in heart failure: a propensity-matched analysis. J Card Fail. 2009; 15(3): 241–8. PubMed Abstract | Publisher Full Text\n\nMozaffarian D, Nye R, Levy WC: Statin therapy is associated with lower mortality among patients with severe heart failure. Am J Cardiol. 2004; 93(9): 1124–9. PubMed Abstract | Publisher Full Text\n\nAlehagen U, Benson L, Edner M, et al.: Association Between Use of Statins and Mortality in Patients With Heart Failure and Ejection Fraction of ≥50. Circ Heart Fail. 2015; 8(5): 862–70. PubMed Abstract | Publisher Full Text\n\nFukuta H, Sane DC, Brucks S, et al.: Statin therapy may be associated with lower mortality in patients with diastolic heart failure: a preliminary report. Circulation. 2005; 112(3): 357–63. PubMed Abstract | Publisher Full Text\n\nNochioka K, Sakata Y, Miyata S, et al.: Prognostic impact of statin use in patients with heart failure and preserved ejection fraction. Circ J. 2015; 79(3): 574–82. PubMed Abstract | Publisher Full Text\n\nLee MS, Duan L, Clare R, et al.: Comparison of Effects of Statin Use on Mortality in Patients With Heart Failure and Preserved Versus Reduced Left Ventricular Ejection Fraction. Am J Cardiol. 2018; 122(3): 405–12. PubMed Abstract | Publisher Full Text\n\nFukuta H, Goto T, Wakami K, et al.: The effect of statins on mortality in heart failure with preserved ejection fraction: a meta-analysis of propensity score analyses. Int J Cardiol. 2016; 214: 301–6. PubMed Abstract | Publisher Full Text\n\nLiu G, Zheng XX, Xu YL, et al.: Meta-analysis of the effect of statins on mortality in patients with preserved ejection fraction. Am J Cardiol. 2014; 113(7): 1198–204. PubMed Abstract | Publisher Full Text\n\nPaulus WJ, Tschöpe C: A novel paradigm for heart failure with preserved ejection fraction: comorbidities drive myocardial dysfunction and remodeling through coronary microvascular endothelial inflammation. J Am Coll Cardiol. 2013; 62(4): 263–71. PubMed Abstract | Publisher Full Text\n\nOhte N, Little WC: Statins Beneficial for Heart Failure With Preserved Ejection Fraction But Not Heart Failure With Reduced Ejection Fraction? Circ J. 2015; 79(3): 508–9. PubMed Abstract | Publisher Full Text\n\nAkahori H, Tsujino T, Naito Y, et al.: Atorvastatin ameliorates cardiac fibrosis and improves left ventricular diastolic function in hypertensive diastolic heart failure model rats. J Hypertens. 2014; 32(7): 1534–41; discussion 41. PubMed Abstract | Publisher Full Text\n\nGómez-Garre D, González-Rubio ML, Muñoz-Pacheco P, et al.: Rosuvastatin added to standard heart failure therapy improves cardiac remodelling in heart failure rats with preserved ejection fraction. Eur J Heart Fail. 2010; 12(9): 903–12. PubMed Abstract | Publisher Full Text\n\nRoik M, Starczewska MH, Huczek Z, et al.: Statin therapy and mortality among patients hospitalized with heart failure and preserved left ventricular function--a preliminary report. Acta Cardiol. 2008; 63(6): 683–92. PubMed Abstract | Publisher Full Text\n\nShah R, Wang Y, Foody JM: Effect of statins, angiotensin-converting enzyme inhibitors, and beta blockers on survival in patients >or=65 years of age with heart failure and preserved left ventricular systolic function. Am J Cardiol. 2008; 101(2): 217–22. PubMed Abstract | Publisher Full Text\n\nTehrani F, Morrissey R, Phan A, et al.: Statin therapy in patients with diastolic heart failure. Clin Cardiol. 2010; 33(4): E1–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTsujimoto T, Kajio H: Favorable effects of statins in the treatment of heart failure with preserved ejection fraction in patients without ischemic heart disease. Int J Cardiol. 2018; 255: 111–7. PubMed Abstract | Publisher Full Text\n\nMannheim D, Herrmann J, Bonetti PO, et al.: Simvastatin preserves diastolic function in experimental hypercholesterolemia independently of its lipid lowering effect. Atherosclerosis. 2011; 216(2): 283–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXu Z, Okamoto H, Akino M, et al.: Pravastatin attenuates left ventricular remodeling and diastolic dysfunction in angiotensin II-induced hypertensive mice. J Cardiovasc Pharmacol. 2008; 51(1): 62–70. PubMed Abstract | Publisher Full Text\n\nChang SA, Kim YJ, Lee HW, et al.: Effect of rosuvastatin on cardiac remodeling, function, and progression to heart failure in hypertensive heart with established left ventricular hypertrophy. Hypertension. 2009; 54(3): 591–7. PubMed Abstract | Publisher Full Text\n\nCollins R, Armitage J, Parish S, et al.: MRC/BHF Heart Protection Study of cholesterol-lowering with simvastatin in 5963 people with diabetes: a randomised placebo-controlled trial. Lancet. 2003; 361(9374): 2005–16. PubMed Abstract | Publisher Full Text\n\nFerrier KE, Muhlmann MH, Baguet JP, et al.: Intensive cholesterol reduction lowers blood pressure and large artery stiffness in isolated systolic hypertension. J Am Coll Cardiol. 2002; 39(6): 1020–5. PubMed Abstract | Publisher Full Text\n\nTonelli M, Moyé L, Sacks FM, et al.: Pravastatin for secondary prevention of cardiovascular events in persons with mild chronic renal insufficiency. Ann Intern Med. 2003; 138(2): 98–104. PubMed Abstract | Publisher Full Text\n\nTokuhisa H, Murai H, Okabe Y, et al.: Differential effects of lipophilic and hydrophilic statins on muscle sympathetic nerve activity in heart failure with preserved left ventricular ejection fraction. Auton Neurosci. 2018; 213: 8–14. PubMed Abstract | Publisher Full Text\n\nSchwartz GG, Steg PG, Szarek M, et al.: Alirocumab and Cardiovascular Outcomes after Acute Coronary Syndrome. N Engl J Med. 2018; 379(22): 2097–107. PubMed Abstract | Publisher Full Text\n\nSabatine MS, Giugliano RP, Keech AC, et al.: Evolocumab and Clinical Outcomes in Patients with Cardiovascular Disease. N Engl J Med. 2017; 376(18): 1713–22. PubMed Abstract | Publisher Full Text\n\nBhatt DL, Steg PG, Miller M, et al.: Cardiovascular Risk Reduction with Icosapent Ethyl for Hypertriglyceridemia. N Engl J Med. 2018; 380(1): 11–22. PubMed Abstract | Publisher Full Text\n\nBhatt DL MM, Steg PG, et al.: Achieved eicosapentaenoic acid levels strongly predict cardiovascular benefit in REDUCE-IT. 2020:Presented at the 2020 American College of Cardiology/World Congress of Cardiology. Abstract 20-LB-20501-ACC.\n\nKrittanawong C, Johnson KW, Tang WW: How artificial intelligence could redefine clinical trials in cardiovascular medicine: lessons learned from oncology. Per Med. 2019; 16(2): 83–8. PubMed Abstract | Publisher Full Text\n\nLee CS, Lee AY: How Artificial Intelligence Can Transform Randomized Controlled Trials. Transl Vis Sci Technol. 2020; 9(2): 9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee LY, Pandey AK, Maron BA, et al.: Network Medicine In Cardiovascular Research. Cardiovasc Res. 2020; cvaa321. PubMed Abstract | Publisher Full Text\n\nBenincasa G, Marfella R, Della Mura N, et al.: Strengths and Opportunities of Network Medicine in Cardiovascular Diseases. Circ J. 2020; 84(2): 144–52. PubMed Abstract | Publisher Full Text\n\nWang RS, Loscalzo J: Illuminating drug action by network integration of disease genes: a case study of myocardial infarction. Mol Biosyst. 2016; 12(5): 1653–66. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "87223",
"date": "14 Jun 2021",
"name": "Claudio Bilato",
"expertise": [
"Reviewer Expertise Clinical Cardiology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a comprehensive review on the role of statins in heart failure both with reduced and preserved ejection fraction. The paper is well-written, informative and easy to read. The authors reported the most important studies and the meta-analysis on the topic in a very exhaustive manner.\nAlthough some data are not analyzed in-depth the manuscript deserves indexing because it is very educational and useful.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": [
{
"c_id": "6797",
"date": "14 Jun 2021",
"name": "Panupong Hansrivijit",
"role": "Author Response",
"response": "Thank you for your time and comment."
}
]
},
{
"id": "164582",
"date": "02 Mar 2023",
"name": "Rui Valdiviesso",
"expertise": [
"Reviewer Expertise Heart failure",
"frailty",
"sarcopenia",
"nutritional and functional status in heart failure",
"muscle strength",
"medication affecting muscle function in heart failure."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript presents a much needed review of the existing literature regarding the effects of statins on clinical outcomes of HF patients with preserved or reduced ejection fraction. The article is clear, informative, and correctly organised. It cites the most relevant clinical studies in the field, such as the CORONA and GISSI-HF, while giving relevant context on the strengths and limitations of each study. The important and many times overlooked behaviour of lipophilic vs. hydrophilic statins was also included in this review.\nAlthough the discussion could go slightly deeper, particularly regarding other outcomes that could add to the utility of statin therapy in HF patients, the conclusions seem correct and in adequacy with the findings of the reviewed evidence.\n\nOther impression that I feel obliged to add to this review is that these findings seem to be in line with my own impressions from both clinical practice with HF patients and studies I conducted or participated in.\n\nThe article is also very informative and written in a pedagogical fashion. Figures 1 and 2 were tastefully added and are useful illustrations to the implied mechanisms.\n\nIn face of my review, I recommend the indexing of this manuscript.\n\nMy congratulations to the authors.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-16
|
https://f1000research.com/articles/10-15/v1
|
11 Jan 21
|
{
"type": "Research Article",
"title": "Enhancement of EPC migration by high-dose lisinopril is superior compared to captopril and ramipril",
"authors": [
"Yudi Her Oktaviono",
"Hanang Anugrawan Ahmad",
"Makhyan Jibril Al Farabi",
"Parama Gandi",
"Caesar Lagaliggo Givani",
"Indah Sari Purna Lumeno",
"Yusuf Azmi",
"Hanang Anugrawan Ahmad",
"Makhyan Jibril Al Farabi",
"Parama Gandi",
"Caesar Lagaliggo Givani",
"Indah Sari Purna Lumeno",
"Yusuf Azmi"
],
"abstract": "Background Angiotensin-converting enzyme (ACE) inhibitors have been shown to promote endothelial progenitor cell (EPC) function. However, the efficacies of different ACE inhibitors in improving the migratory capabilities of ECPs in coronary artery disease (CAD) patients is unclear. This study compared the effectiveness of captopril, lisinopril, and ramipril toward the migration capability of impaired EPCs from CAD patients.\n\nMethods We isolated peripheral blood mononuclear cells (PBMCs), separated EPCs from PBMCs, and divided them into an untreated group (control) and treated groups of captopril, lisinopril, and ramipril at doses of 1mM, 10mM, and 100mM. EPC migration was evaluated using the Boyden chamber assay. Analysis of variance (ANOVA) was performed using SPSS 25.0.\n\nResults This study showed that treatment with captopril, lisinopril, and ramipril starting at the lowest dose (1 mM) increased EPC migration (65,250 ± 6,750 cells; 60,750± 5,030 cells; and 49,500 ± 8,400 cells, respectively) compared to control (43,714 ± 7,216 cells). Increased migration of EPCs was observed by increasing the treatment dose to 10 mM with captopril, lisinopril, and ramipril (90,000 ± 16,837 cells; 79,071 ± 2,043 cells; and 64,285 ± 11,824 cells, respectively). The highest EPC migration was shown for lisinopril 100 mM (150,750 ± 16,380 cells), compared to captopril and ramipril at the same dose (105,750 ± 8112 cells and 86,625 ± 5,845 cells, respectively).\n\nConclusions Captopril, ramipril, and lisinopril were shown to increase EPC migration in a dose-dependent manner. Low-dose (1 mM) and medium-dose (10 mM) captopril had a larger effect on ECP migration than lisinopril and ramipril. Meanwhile, high-dose lisinopril (100mM) had the highest migration effect, suggesting it may be preferable for promoting EPC migration in CAD patients.",
"keywords": [
"ACE Inhibitors",
"Coronary artery disease",
"Endothelial progenitor cells",
"Migration"
],
"content": "Introduction\n\nEndothelial dysfunction and impaired endothelial regeneration are thought to play an important role in the pathogenesis of arteriosclerosis in coronary arterial disease (CAD)1. Endothelial regeneration is not only fulfilled by resident endothelial cells but also repaired by endothelial progenitor cells (ECPs) originating from the bone marrow2. ECPs are premature circulating cells, a specific subtype of hematopoietic stem cells that differentiate into endothelial cells in situ and promote neovascularization3,4. Several studies have shown that in patients with CAD, there is a significant decrease in the number and migratory function of circulating EPCs, which leads to impaired neovascularization of ischemic tissue5,6. Low EPC counts can predict severe endothelial dysfunction, cardiovascular events, and deaths from cardiovascular causes7,8. It is suggested that intracellular damage and impaired redox balance in EPCs due to oxidative stress are the predisposes of imbalance in vascular pathology9,10.\n\nAngiotensin-converting enzyme (ACE) inhibitors are widely used in cardiovascular disease and have been shown to be associated with beneficial effects on EPCs in several in vitro and clinical studies11–14. An animal study on mice with increased left ventricular pressure showed that ramipril increases the number and improves EPC migration12. A clinical study in hypertensive patients showed that enalapril and zofenopril reduce EPC levels and prevent vascular damage and carotid intima-media thickening13. In a small clinical trial, administration of ramipril for four weeks in patients with stable CAD augments and increases the functional activity of EPCs, including migration, adhesion, and in vitro capacity of vasculogenesis15.\n\nHowever, no studies have investigated the role of ACE inhibitors of captopril and lisinopril in relation to the EPCs. In addition, the comparison between different types of ACE inhibitors toward the impaired migration function of EPCs remains to be investigated. We aimed to evaluate the effects of captopril, lisinopril, and ramipril on EPCs migration from CAD patients.\n\n\nMethods\n\nOur study protocol was approved by the Institutional Ethics Committee of Dr. Soetomo General Hospital (945/KEPK/II/2019). Informed consent for peripheral blood sampling procedures and participation in research studies was obtained from all patients before the blood was drawn. We have omitted all data that could reveal the identity of the patients.\n\nIn the present study, we used peripheral blood samples from the same participants and performed similar methodology to that described in our previous study16. From June 2018 to December 2018, we studied a total of eight patients with stable CAD who underwent coronary angiography. Only patients with the left main coronary artery stenosis of more than 50% or stenosis in other coronary arteries more than 70% were recruited. To prevent the effects of myocardial ischemia on ECP kinetics, we excluded patients with a history of new-onset acute myocardial infarction. In addition, patients with anemia, diabetes, a history of percutaneous coronary intervention, or coronary artery bypass grafting were not included in the study. Physical examination was performed to determine body mass index (BMI) and to assess the vital signs. We also examined the lipid profile and performed echocardiography to assess left ventricular function. The characteristics of the study population are summarized in Table 1.\n\nBMI, body mass index; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SD, standard deviation.\n\nWe collected 40 ml peripheral blood samples from the median cubital vein following WHO guidelines on drawing blood17. From freshly drawn heparinized blood, we isolated peripheral blood mononuclear cells (PBMCs) using Ficoll Histopaque 1077 (Sigma-Aldrich, USA). Briefly, peripheral blood was diluted 1:1 with phosphate buffer saline (PBS) + 2% fetal bovine serum (FBS) to a total volume of 30–35 ml. It was then carefully layered into 20 ml of Ficoll Histopaque 1077 (Sigma-Aldrich, USA) in a 50 ml conical tube. Subsequently, the tube was put into a centrifuge at 300xg for 30 minutes. The PBMC layer was obtained in the form of a buffy coat layer. Using a sterile plastic pipette, the PBMCs were carefully taken and put into another 50 ml conical tube. Furthermore, PBMC was added with PBS + 2% FBS in a ratio of 1: 1, then stirred until homogeneous and centrifuged at 300xg for 7 minutes. This step was repeated with the supernatant removed, 15 ml of PBS + 2% FBS was added to the precipitate formed at the bottom of the tube and centrifuged at 300xg for 7 minutes. Finally, the supernatant was removed, and the sediment was dissolved with a basal medium. Cells were concentrated up to 5×106 cells/ml.\n\nTo separate EPCs from PBMCs, we used standard protocols18. Briefly, PBMCs isolated from blood samples in a concentration of 5×106 cells/mL were collected in Stemline II Hematopoietic Stem Cell Expansion Medium (Sigma-Aldrich, USA) supplemented with endothelial basal medium (EBM) containing 40 ng/ml of vascular endothelial growth factor (VEGF) and 15% FBS. Then PBMCs were seeded in the six-well plate with fibronectin coating. The cultures were maintained with a humidified atmosphere at 37°C and 5% carbon dioxide. Forty-eight hours after seeding, we separated the medium liquid containing the non-adherent cells from the adherent cells attached to the bottom of the plate. All the medium liquid containing the non-adherent cells was collected into one tube, centrifuged with a spin at 300xg for 7 minutes, and the supernatant was discarded. The precipitate formed was dissolved with basal medium and supplement with a concentration of 1×106 cells/ml. We confirmed the cells as EPCs through immunofluorescence tests using fluorescein isothiocyanate (FITC) mouse anti-human CD34 monoclonal antibody (cat. no. 343604; Biolegend, USA).\n\nIsolated EPCs were divided into three treatment groups and one negative control group. The treatment groups were divided into 1) captopril 1 μM, 10 μM, and 100 μM; 2) lisinopril 1 μM, 10 μM, and 100 μM; and 3) ramipril 1 μM, 10 μM, and 100 μM. The cultures were maintained with a humidified atmosphere at 37°C and 5% carbon dioxide for 48 hours. Each data point represented the mean value of quadruplicate cultures.\n\nWe used the Boyden chamber migration assay to measure ECP migration. Briefly, migration chambers with 8μm pore-size filters were placed in 24-well plates. Using 1 mmol/L EDTA in PBS, isolated EPCs were detached and then centrifuged at 400xg for ten minutes. EPCs were seeded in the upper chamber (5×105/ml in serum-free medium), and the lower compartment of the Boyden chamber was filled with endothelial basal medium. After 24 hours of incubation at 37°C, we scraped off non-migratory cells on the upper chamber with cotton swabs. The migration chamber was put into a new basal medium and added with 500μL of trypsin + EDTA 0.5% solution. After 10 minutes of incubation, we verified using a light microscope to ensure more than 90% of adhering cells were released from the lower surface of the migration chambers. For quantification, the cells were harvested and stained with trypan blue/Giemsa. Migrated ECPs were counted using an TC20 automated cell counter (Bio-Rad, USA).\n\nContinuous data were presented as mean ± SD. Multiple experimental group analysis of total migrated EPCs was performed using analysis of variance (ANOVA). A p-value of less than 0.05 was considered statistically significant. All statistical analysis was completed using SPSS version 25.0 for Windows.\n\n\nResults\n\nCD34 is a positive marker for EPCs, and CD34 expression was found in the early to mature culture of EPCs. CD34 expression was characterized by the presence of green luminescence using a fluorescence microscope, indicating the presence of EPCs, as shown in Figure 119. The migration capability of EPCs was evaluated by calculating the number of cells that moved from the upper chamber to the membrane facing the lower chamber with Giemsa staining (Figure 2).\n\nLight-inverted microscope view of endothelial progenitor cells under 48 h-treatment of (a) 100 mM captopril, (b) 100 mM lisinopril, (c) 100 mM ramipril, (d) negative control (medium only), and (e) positive control (100 ng/mL VEGF). White bar represents 100µM.\n\nThe number of EPC migrations in the captopril-treated group at different doses (65,250 ± 6,750 cells at 1 mM; 90,000 ± 16,837 cells at 10mM; and 105,750 ± 8112 cells at 100 mM) was significantly higher than the control group (43,714 ± 7,216 cells) (p < 0.05) (Figure 3). The number of EPC migrations in the lisinopril-treated group at different doses (60,750 ± 5,030 cells at 1 mM; 79,071 ± 2,043 cells at 10mM; and 150,750 ± 16,380 cells at 100 mM) was significantly higher than the control group (43,714 ± 7,216 cells) (p < 0.05) (Figure 4). The number of EPC migrations in the ramipril-treated group at different doses (49,500 ± 8,400 cells at 1 mM; 64,285 ± 11,824 cells at 10mM; and 86,625 ± 5,845 cells at 100 mM) was significantly higher than the control group (43,714 ± 7,216 cells) (p < 0.05) (Figure 5).\n\nTotal migrated cells are expressed as mean ± SD (n = 4). Different annotations (a,b,c,d) denounce significant difference in ANOVA test (p<0.05).\n\nTotal migrated cells are expressed as mean ± SD (n = 4). Different annotations (a,b,c,d) denounce significant difference in ANOVA test (p<0.05).\n\nTotal migrated cells are expressed as mean ± SD (n = 4). Different annotations (a,b,c,d) denounce significant difference in ANOVA test (p<0.05).\n\nThe increase in the migration of EPCs was consistent with the increase in the dose of ACE inhibitor. Captopril at doses of 1 mM and 10 mM had a higher migration effect than lisinopril and ramipril at the same doses (p < 0.05). Meanwhile, lisinopril at a dose of 100mM had the highest migration effect (p < 0.05) (Figure 6).\n\nTotal migrated cells are expressed as mean ± SD (n = 4). Different annotations (a,b,c,d) denounce significant difference in ANOVA test (p<0.05).\n\n\nDiscussion\n\nIn this present study, we demonstrated that captopril, lisinopril, and ramipril therapy in EPC cultures from CAD patients was associated with improved migration of EPCs. This study showed that ACE inhibitor treatment increases EPCs migration in a dose-dependent manner. At the doses of 1 mM and 10 mM, there was no significant difference in EPCs migration between captopril and lisinopril. However, both of them exceeded the results of ramipril at the same dose. Meanwhile, lisinopril at the dose of 100 mM had a superior outcome compared to captopril and ramipril at the same dose.\n\nCirculating EPCs are derived from hematopoietic stem cells produced in the bone marrow, which can repair endothelial dysfunction through endogenous mechanisms. In patients with CAD, the number and migration capacity of ECPs are decreased, and thus they are unable to maintain adequate endothelial stability6,20–22. During ischemic conditions, EPCs are known to play an essential role in reendothelization and neovascularization. Animal and clinical studies have shown that EPCs contribute up to 25% of newly formed vascular endothelial cells after ischemic conditions23,24.\n\nSeveral pharmacological agents have reported the beneficial effects on EPCs, such as HMG-CoA reductase inhibitors/statin25,26, one of which has been demonstrated by our previous study16, peroxisome proliferator-activated receptor (PPAR) agonists27, dihydropyridine calcium channel blocker28, and angiotensin II receptor antagonists (ARB)29. Antioxidative agents with anti-inflammatory properties, such as ginsenoside, salvianolic acids, berberine, Ginkgo biloba, resveratrol, and puerarin, also have been found to increase the number or functional activity of EPCs30. ACE inhibitors, which are widely used in cardiovascular therapy, such as for hypertension and congestive heart failure, may have a potential role in restoring the role of EPCs in repair, healing, and neovascularization11,31. Several studies have demonstrated the role of ACE inhibitors in increasing the number and function of EPCs in patients with hypertension and stable CAD13,15. Each of the ACE inhibitors has a different chemical functional group, which may explain the varying effects of different ACE inhibitor types in several studies, either in vitro or in vivo. Sulfhydryl-containing ACE inhibitors are known to be the most effective compared to other types of ACE inhibitors31–34. Captopril has one sulfhydryl group, and zofenopril has two sulfhydryl groups to coordinate the zinc ion of the active side, whereas lisinopril, ramipril, and enalapril do not have sulfhydryl groups35–38. Sulfhydryl-containing ACE inhibitors can reduce oxidative stress and stimulate nitric oxide (NO) activity in human endothelial cells39 and patients with primary hypertension40. In vitro studies have shown that zofenopril is more effective compared to enalapril in preventing foam cell formation and thereby slowing atherosclerosis. In addition, zofenopril can also reduce reactive oxygen species and increase NO production in the endothelium37,41–44\n\nThe finding that ACE inhibition therapy augmented the number of circulating EPCs in patients with CAD, and also enhanced EPCs functional activity, may provide a novel strategy to improve neovascularization and reendothelialization after ischemia, thereby providing a therapeutic concept to improve EPC numbers and functions in patients with CAD.\n\n\nConclusion\n\nCaptopril, ramipril, and lisinopril were shown to increase EPC migration in a dose-dependent manner. Low-dose (1 mM) and medium-dose (10 mM) captopril had a larger effect on ECP migration than lisinopril and ramipril. Meanwhile, high-dose lisinopril (100mM) had the highest migration effect, suggesting it may be preferable for promoting EPC migration in CAD patients.",
"appendix": "Data availability\n\nFigshare: Dataset for Enhancement of EPC migration by high-dose lisinopril is superior compared to captopril and ramipril. https://doi.org/10.6084/m9.figshare.13130303.v219\n\nThis project contains the following underlying data:\n\n- Transwell_Migration_Assay_Dataset.xlsx\n\n- Image Repository.zip (original, unedited microscopy images in JPG format)\n\n- Clinical and demographic data of study population.docx\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).\n\n\nReferences\n\nRoss R: Atherosclerosis--an inflammatory disease. N Engl J Med. 1999; 340(2): 115–126. PubMed Abstract | Publisher Full Text\n\nEndtmann C, Ebrahimian T, Czech T, et al.: Angiotensin II impairs endothelial progenitor cell number and function in vitro and in vivo: implications for vascular regeneration. Hypertension. 2011; 58(3): 394–403. PubMed Abstract | Publisher Full Text\n\nAsahara T, Masuda H, Takahashi T, et al.: Bone marrow origin of endothelial progenitor cells responsible for postnatal vasculogenesis in physiological and pathological neovascularization. Circ Res. 1999; 85(3): 221–228. PubMed Abstract | Publisher Full Text\n\nHu Y, Foteinos G, Xiao Q, et al.: Rapid endothelial turnover in atherosclerosis-prone areas coincides with stem cell repair in apoe-deficient mice. Atherosclerosis. 2008; 199(2): 467. Publisher Full Text\n\nOktaviono YH, Sargowo D, Widodo MA, et al.: Proliferation of Peripheral Blood-derived Endothelial Progenitor Cells from Stable Angina Subjects. Indones Biomed J. 2014; 6(2): 91. Publisher Full Text\n\nVasa M, Fichtlscherer S, Aicher A, et al.: Number and migratory activity of circulating endothelial progenitor cells inversely correlate with risk factors for coronary artery disease. Circ Res. 2001; 89(1): E1–7. PubMed Abstract | Publisher Full Text\n\nWerner N, Wassmann S, Ahlers P, et al.: Endothelial progenitor cells correlate with endothelial function in patients with coronary artery disease. Basic Res Cardiol. 2007; 102(6): 565–571. PubMed Abstract | Publisher Full Text\n\nWerner N, Kosiol S, Schiegl T, et al.: Circulating endothelial progenitor cells and cardiovascular outcomes. N Engl J Med. 2005; 353(10): 999–1007. PubMed Abstract | Publisher Full Text\n\nLoomans CJM, De Koning EJP, Staal FJT, et al.: Endothelial progenitor cell dysfunction in type 1 diabetes: another consequence of oxidative stress? Antioxid Redox Signal. 2005; 7(11–12): 1468–1475. PubMed Abstract | Publisher Full Text\n\nRehman J, Li J, Orschell CM, et al.: Peripheral blood \"endothelial progenitor cells\" are derived from monocyte/macrophages and secrete angiogenic growth factors. Circulation. 2003; 107(8): 1164–1169. PubMed Abstract | Publisher Full Text\n\nYou D, Cochain C, Loinard C, et al.: Combination of the angiotensin-converting enzyme inhibitor perindopril and the diuretic indapamide activate postnatal vasculogenesis in spontaneously hypertensive rats. J Pharmacol Exp Ther. 2008; 325(3): 766–773. PubMed Abstract | Publisher Full Text\n\nMüller P, Kazakov A, Jagoda P, et al.: ACE inhibition promotes upregulation of endothelial progenitor cells and neoangiogenesis in cardiac pressure overload. Cardiovasc Res. 2009; 83(1): 106–114. PubMed Abstract | Publisher Full Text\n\nCacciatore F, Bruzzese G, Vitale DF, et al.: Effects of ACE inhibition on circulating endothelial progenitor cells, vascular damage, and oxidative stress in hypertensive patients. Eur J Clin Pharmacol. 2011; 67(9): 877–883. PubMed Abstract | Publisher Full Text\n\nWang CH, Verma S, Hsieh IC, et al.: Enalapril increases ischemia-induced endothelial progenitor cell mobilization through manipulation of the CD26 system. J Mol Cell Cardiol. 2006; 41(1): 34–43. PubMed Abstract | Publisher Full Text\n\nMin TQ, Zhu CJ, Xiang WX, et al.: Improvement in endothelial progenitor cells from peripheral blood by ramipril therapy in patients with stable coronary artery disease. Cardiovasc Drugs Ther. 2004; 18(3): 203–209. PubMed Abstract | Publisher Full Text\n\nOktaviono YH, Al Farabi MJ, Meuthia F, et al.: Rosuvastatin is Superior Compared to Simvastatin and Atorvastatin to Induce Endothelial Progenitor Cells Migration. J Clin Diagnostic Res. 2019; 13(5): OC05–OC08. Publisher Full Text\n\nWorld Health Organization (WHO): WHO guidelines on drawing blood: best practices in phlebotomy. World Health Organization. 2010. Reference Source\n\nBueno-Betí C, Novella S, Lázaro-Franco M, et al.: An affordable method to obtain cultured endothelial cells from peripheral blood. J Cell Mol Med. 2013; 17(11): 1475–1483. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAl Farabi MJ: Dataset for Enhancement of EPC migration by high-dose lisinopril is superior compared to captopril and ramipril. figshare. Dataset. 2020. http://www.doi.org/10.6084/m9.figshare.13130303.v2\n\nFadini GP, Coracina A, Baesso I, et al.: Peripheral blood CD34+KDR+ endothelial progenitor cells are determinants of subclinical atherosclerosis in a middle-aged general population. Stroke. 2006; 37(9): 2277–2282. PubMed Abstract | Publisher Full Text\n\nLiguori A, Fiorito C, Balestrieri ML, et al.: Functional impairment of hematopoietic progenitor cells in patients with coronary heart disease. Eur J Haematol. 2008; 80(3): 258–264. PubMed Abstract | Publisher Full Text\n\nBriguori C, Testa U, Riccioni R, et al.: Correlations between progression of coronary artery disease and circulating endothelial progenitor cells. FASEB J. 2010; 24(6): 1981–1988. PubMed Abstract | Publisher Full Text\n\nMurayama T, Tepper OM, Silver M, et al.: Determination of bone marrow-derived endothelial progenitor cell significance in angiogenic growth factor-induced neovascularization in vivo. Exp Hematol. 2002; 30(8): 967–972. PubMed Abstract | Publisher Full Text\n\nSuzuki T, Nishida M, Futami S, et al.: Neoendothelialization after peripheral blood stem cell transplantation in humans: a case report of a Tokaimura nuclear accident victim. Cardiovasc Res. 2003; 58(2): 487–492. PubMed Abstract | Publisher Full Text\n\nLlevadot J, Murasawa S, Kureishi Y, et al.: HMG-CoA reductase inhibitor mobilizes bone marrow--derived endothelial progenitor cells. J Clin Invest. 2001; 108(3): 399–405. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLandmesser U, Engberding N, Bahlmann FH, et al.: Statin-induced improvement of endothelial progenitor cell mobilization, myocardial neovascularization, left ventricular function, and survival after experimental myocardial infarction requires endothelial nitric oxide synthase. Circulation. 2004; 110(14): 1933–1939. PubMed Abstract | Publisher Full Text\n\nGensch C, Clever YP, Werner C, et al.: The PPAR-gamma agonist pioglitazone increases neoangiogenesis and prevents apoptosis of endothelial progenitor cells. Atherosclerosis. 2007; 192(1): 67–74. PubMed Abstract | Publisher Full Text\n\nPassacquale G, Desideri G, Croce G, et al.: Nifedipine improves the migratory ability of circulating endothelial progenitor cells depending on manganese superoxide dismutase upregulation. J Hypertens. 2008; 26(4): 737–746. PubMed Abstract | Publisher Full Text\n\nBahlmann FH, de Groot K, Mueller O, et al.: Stimulation of endothelial progenitor cells: a new putative therapeutic effect of angiotensin II receptor antagonists. Hypertension. 2005; 45(4): 526–529. PubMed Abstract | Publisher Full Text\n\nLin CP, Lin FY, Huang PH, et al.: Endothelial progenitor cell dysfunction in cardiovascular diseases: role of reactive oxygen species and inflammation. Biomed Res Int. 2013; 2013: 845037. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNapoli C, Cicala C, D’Armiento FP, et al.: Beneficial effects of ACE-inhibition with zofenopril on plaque formation and low-density lipoprotein oxidation in watanabe heritable hyperlipidemic rabbits. Gen Pharmacol. 1999; 33(6): 467–477. PubMed Abstract | Publisher Full Text\n\nCandido R, Jandeleit-Dahm KA, Cao Z, et al.: Prevention of accelerated atherosclerosis by angiotensin-converting enzyme inhibition in diabetic apolipoprotein E-deficient mice. Circulation. 2002; 106(2): 246–253. PubMed Abstract | Publisher Full Text\n\nSun YP, Zhu BQ, Browne AE, et al.: Comparative effects of ACE inhibitors and an angiotensin receptor blocker on atherosclerosis and vascular function. J Cardiovasc Pharmacol Ther. 2001; 6(2): 175–181. PubMed Abstract | Publisher Full Text\n\nChobanian AV, Haudenschild CC, Nickerson C, et al.: Antiatherogenic effect of captopril in the Watanabe heritable hyperlipidemic rabbit. Hypertension. 1990; 15(3): 327–331. PubMed Abstract | Publisher Full Text\n\nUnger T: The role of the renin-angiotensin system in the development of cardiovascular disease. Am J Cardiol. 2002; 89(2A): 3A–9A; discussion 10A. PubMed Abstract | Publisher Full Text\n\nKowala MC, Grove RI, Aberg G: Inhibitors of angiotensin converting enzyme decrease early atherosclerosis in hyperlipidemic hamsters. Fosinopril reduces plasma cholesterol and captopril inhibits macrophage-foam cell accumulation independently of blood pressure and plasma lipids. Atherosclerosis. 1994; 108(1): 61–72. PubMed Abstract | Publisher Full Text\n\nBuikema H, Monnink SH, Tio RA, et al.: Comparison of zofenopril and lisinopril to study the role of the sulfhydryl-group in improvement of endothelial dysfunction with ACE-inhibitors in experimental heart failure. Br J Pharmacol. 2000; 130(8): 1999–2007. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPredel HG, Düsing R, Bäcker A, et al.: Combined treatment of severe essential hypertension with the new angiotensin converting enzyme inhibitor ramipril. Am J Cardiol. 1987; 59(10): 143D–148D. PubMed Abstract | Publisher Full Text\n\nJacoby DS, Rader DJ: Renin-angiotensin system and atherothrombotic disease: from genes to treatment. Arch Intern Med. 2003; 163(10): 1155–64. PubMed Abstract | Publisher Full Text\n\nNapoli C, Bruzzese G, Ignarro LJ, et al.: Long-term treatment with sulfhydryl angiotensin-converting enzyme inhibition reduces carotid intima-media thickening and improves the nitric oxide/oxidative stress pathways in newly diagnosed patients with mild to moderate primary hypertension. Am Heart J. 2008; 156(6): 1154.e1–8. PubMed Abstract | Publisher Full Text\n\nCominacini L, Pasini A, Garbin U, et al.: Zofenopril inhibits the expression of adhesion molecules on endothelial cells by reducing reactive oxygen species. Am J Hypertens. 2002; 15(10 Pt 1): 891–895. PubMed Abstract | Publisher Full Text\n\nEvangelista S, Manzini S: Antioxidant and cardioprotective properties of the sulphydryl angiotensin-converting enzyme inhibitor zofenopril. J Int Med Res. 2005; 33(1): 42–54. PubMed Abstract | Publisher Full Text\n\nde Nigris F, D’Armiento FP, Somma P, et al.: Chronic treatment with sulfhydryl angiotensin-converting enzyme inhibitors reduce susceptibility of plasma LDL to in vitro oxidation, formation of oxidation-specific epitopes in the arterial wall, and atherogenesis in apolipoprotein E knockout mice. Int J Cardiol. 2001; 81(2–3): 107–115; discusssion 115-6. PubMed Abstract | Publisher Full Text\n\nScribner AW, Loscalzo J, Napoli C: The effect of angiotensin-converting enzyme inhibition on endothelial function and oxidant stress. Eur J Pharmacol. 2003; 482(1–3): 95–99. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "84488",
"date": "17 May 2021",
"name": "Mehdi Hassanpour",
"expertise": [
"Reviewer Expertise Clinical Biochemistry"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript entitled “Enhancement of EPC migration by high-dose lisinopril is superior compared to captopril and ramipril” seems an interesting topic in the field of cardiac and vessel cell biology. However, there are some points that need revision before indexing.\nReviewer comments:\nThe topic is interesting and the aim of the study is commendable, but this article has some typos errors, abbreviations, and grammatical issues and needs editing.\n\nThe EPC should be in complete form in the title of the article.\n\nAs a result of EPC confirmation through immunofluorescence tests, what percentage of cultured cells were CD34 positive?\n\nThe underlying molecular mechanism and involved signaling pathways of Lisinopril on EPC migration were neglected and should be discussed.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "96122",
"date": "13 Oct 2021",
"name": "Dyana Sarvasti",
"expertise": [
"Reviewer Expertise Cardiovascular prevention and rehabilitation."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study explores the role of ACE inhibitors (captopril, lisinopril, and ramipril) in improving the migratory capabilities of EPCs in coronary artery disease (CAD). Overall, the research published by this author is very interesting. However, several things need to be added to this research report so that readers can understand it better. For example, in the abstract, it is necessary to add the characteristics of the patients included in this study so that the reader can understand the study's outline.\n\nIn the research method, it is also necessary to explain whether or not the patients included in this study are already on ACE inhibitor therapy. Was the number of patients included in the study (eight patients) sufficient to conclude? What about the effect of other therapies (not ACE inhibitors) that the patient had taken before being included in the study? How to “clean up” the various confounders in the study? Some of these questions may need to be added and explained in this research report.\nIn addition, there are some inconsistencies in the writing of abbreviations and mistypes that need to be corrected. Here are some points related to this:\n“However, the efficacies of different ACE inhibitors in improving the migratory capabilities of ECPs in coronary artery disease (CAD) patients is unclear.” The singular verb “is” does not appear to agree with the plural subject \"patients\" or \"the efficacies\".\n\n“Keywords: ACE Inhibitors, Coronary artery disease, Endothelial progenitor cells, Migration.” It should be in lowercase.\n\nThe description of the abbreviation CAD differs between the description in the abstract and the introduction.\n\nThe author is inconsistent in writing the abbreviation for endothelial progenitor cells. Sometimes it says “EPC”, but in other parts, it says “ECP”.\n\n“It is suggested that intracellular damage and impaired redox balance in EPCs due to oxidative stress are the predisposes of imbalance in vascular pathology.” The phrase “are the predisposes of” maybe wordy. Consider changing to “predisposes to”.\n\n“Forty-eight hours after seeding, we separated the medium liquid ontaining the non-adherent cells from…….”. It should be “containing”.\n\n“The precipitate formed was dissolved with basal medium and supplement with……”. It should be “supplemented”.\n\nThe author is inconsistent in writing the unit dose of the drug. In the abstract, it says “mM”, but in other parts, it says “μM”.\n\n“Migrated EPCs were counted using an TC20 automated cell counter (Bio-Rad, USA)”. It should be “a”.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-15
|
https://f1000research.com/articles/10-12/v1
|
11 Jan 21
|
{
"type": "Research Article",
"title": "Development of a Lebanese food exchange system based on frequently consumed Eastern Mediterranean traditional dishes and Arabic sweets",
"authors": [
"Maha Hoteit",
"Edwina Zoghbi",
"Alissar Rady",
"Iman Shankiti",
"Ayoub Al-Jawaldeh",
"Maha Hoteit",
"Edwina Zoghbi",
"Alissar Rady",
"Iman Shankiti"
],
"abstract": "Background: The important role of Mediterranean diet was elucidated in various clinical and epidemiological studies underlying its impact on reducing the burden of non-communicable diseases in Mediterranean and non-Mediterranean populations. Objective: The aim of this study was to convert the recipes of the Lebanese traditional dishes into meal planning exchange lists whose items are expressed in grams and adjusted to Lebanese household measures (cups and spoons) that could be used by healthcare professionals. Methodology: Thirty traditional Lebanese dishes were collected in which the carbohydrate, fat and protein were analyzed using Association of Official Analytical Chemists procedures then followed by a calculation of exchange lists of foods per serving using Wheeler method. Results: The variations in macronutrients and fiber content were found among the Lebanese dishes. Carbohydrate was lowest (1.1g/100g) and protein was highest (29.7g/100g) in Shawarma Dajaj whereas fat content ranged between 0.5 and 22.4 g/100 g in the dishes. For each dish and according to each serving size, carbohydrate, milk (whole milk, reduced fat or skim), fat and protein (lean meat, medium fat meat and high fat meat) exchanges were calculated. Conclusion: This study provides healthcare professionals, dietitians and consumers the chance to proficiently plan traditional-type dishes, ensuring prominent dietetic and medical nutritional therapy practices and patient's self-control.",
"keywords": [
"Exchange list",
"carbohydrate",
"protein",
"fat",
"traditional dishes",
"Lebanon"
],
"content": "Background\n\nThe Eastern Mediterranean Region (EMR) is currently facing rapid social and economic changes, urbanization, and advances in technology along with a shift in the prevalence of non-communicable diseases (NCDs) due to the obesogenic environment, nutrition transition and modification of lifestyle patterns characterized by changes in food intake and a reduction in physical activity attitudes and practices1. Lebanon is an Arab country with a population of over six million2 and is considered as part of the EMR. Few data have discussed food attitudes and practices in Lebanon; however, recent evidence published in 2019 from two surveys enrolled previously between 1997 and 2008/2009, showed a shift in the Lebanese diet with regards to an expansion in energy intake from 1728 ± 24 kcal in 1997 to 1877 ± 15 kcal in 2008/2009 and dietary fat from 34.63 ± 0.32% in 1997 to 36.97 ± 0.21% in 2008/2009, coupled with decreases in carbohydrate (CHO) (48.97 ± 0.23% in 2008/2009 compared to 51.32 ± 0.36% in 1997), in fruit consumption (4.72 ± 0.15% vs. 7.36 ± 0.22% in 1997) and a decrease in micronutrient dietary density along with a decrease in the consumption of milk (1.09 ± 0.08% vs. 1.53 ± 0.11% in 2008/2009 and 1997, respectively). No changes were reported with regards to protein and fiber intake3. According to the World Health Organization (WHO) statistics, the worldwide probability of premature death due to NCDs accounts for 70% of the 41 million deaths each year. In Lebanon in 2016, 91% (10,334 individuals) of all deaths were due to NCDs4. Furthermore, among the 10,334 premature deaths from NCDs, 27.2% of deaths were due to cancer. Around 31% of the population was obese, 13% had diabetes type II, 20% had raised blood pressure and 36% had a sedentary lifestyle4.\n\nThe implementation of food composition tables and development of exchange lists is drawing attention at a national and international level due to the recommendations and guidelines5 published by the public and private sectors with the purpose of implementing programs aiming to ameliorate medical nutritional therapeutics for widely distributed NCDs. Food composition tables and exchange lists also have an impact on marketing, on the trade of products and on consumer safety and health6. Many factors influence meal planning for a healthy diet, of which food choices, personal preferences, ethnic behaviors and tradition may increase the responsibility and commitment of health care professionals and dietitians in promoting improved nutrition at an individual and community level, through adherence to food composition tables and exchange lists7.\n\nAt the Middle Eastern level, there is a lack in regional and national exchange lists that incorporate traditional foods. The food exchange list8, which helps in translating recipes into food serving sizes and energy intake, was designed to assist individuals in an easily operated and easily understood way in improving their healthy eating habits and in adhering to a healthy diet plan. In the food exchange list, foods available in the same category can be used reciprocally without any change in the quantities of macronutrients and energy yielded by a dish. According to numerous sources, there is a need to add cultural relevance to food exchange lists to improve NCD management9, taking into consideration ethnic variations and traditions which may have enormous influence on individual and community health10. Thus, the key success factor for improving health care professionals, dietitians and individuals’ adherence to healthy food choices is to design and implement culturally accepted exchange lists that add traditional dishes into their meal planning7. Currently, Lebanese dietitians are using the American Dietetic Association exchange list to design meal plans. However, this exchange list is limited by a gap in traditional meals consumed in Lebanon. In spite of the availability of a few national food composition tables that involve a limited number of dishes common in Lebanese cuisine11,12, ingredients and preparation methods differed substantially across time and these are outdated. The aim of this study is to enable Lebanese healthcare professionals and dietitians to develop, design, and implement practical culture-based meal plans that include traditional cooking.\n\n\nMethods\n\nThe definition of ‘composite dishes’ is “dishes consumed at main meals (i.e. lunch or dinner), whose preparation involves culinary skills and contains ingredients from at least three of five main food groups: meat/poultry/fish and eggs; dairy products; fruits and vegetables; starchy foods including legumes; added sweets and fats”13. The list of Lebanese composite dishes frequently consumed by Lebanese citizens was retrieved from a study performed in 2005 on a representative sample of 799 Lebanese adults13, and in line with a study conducted in 2009 where the objective was to compare the consumption of traditional dishes between Lebanon and France14. As for Arabic sweets, a broad selection of almost all types frequently consumed by Lebanese people was compiled. The Lebanese diet includes a range of foods with often complex recipes, and it is rarely possible to analyze all of the types of dishes. In such cases, laboratory analysis of the traditional dishes and a calculation of some nutrients should be achieved. The names of the dishes and Arabic sweets most eaten by Lebanese citizens and chosen for this study are shown in Table 1.\n\nA sample of 500 g of each dish or sweet was collected and used for analysis. According to Greenfield et al., this size is a convenient sample to avoid errors during analysis15. Our research group collected 500 g of 30 types of traditional dishes from a central kitchen in Beirut and 35 types of Arabic sweets from popular sweet retails in the same area. These popular kitchen and sweet retailers, which serve traditional meals and Arabic sweets were found on the internet and chosen based on the following criteria: 1) their specialty in cooking home-made dishes and serving Arabic sweets; 2) their popularity; and 3) the popular kitchen’s involvement in supporting women, as part of social entrepreneurship initiatives that are aimed at empowering women. Regarding traditional meals, five samples for each dish were collected previously from various regions of Lebanon and tested to eliminate any discrepancies16. The laboratory analyses of the 150 samples showed, after being tested using Chi-square, no significant differences in all of the variables tested (p=0.4, data not shown) between different governorates in Lebanon. IBM SPSS Statistics 26.0 was the software used for analysis of the results.\n\nAfter the receipt of the food samples, 500 g of each composite dish was mashed and then analyzed in the laboratory. The remaining samples were kept frozen at -18°C in tight containers for further analysis. According to the standard Association of Official Agricultural Chemists17 procedures, components of the dishes such as ash, moisture, crude protein and crude fat were analyzed. Using thermal drying methods with a Fisher Isotemp vacuum oven, moisture was identified. The food samples were heated to 105°C and the measurement of moisture content was based on the loss of weight of the sample. According to the International Dairy Federation Standard, crude protein was calculated through the multiplication of 6.38 by total Kjeldahl nitrogen (AOAC 991.20-23)17. Total fatty acid was analyzed by extraction with petroleum ether using the Soxhlet extraction system. The Roese-Gotlieb method was used in the investigation of the fat content (AOAC 945.48, 933.05 & 963.15, 2019)17. To obtain ash, using an Isotemp muffle furnace, an oxidation of all organic matter in a weighed sample was achieved by incineration in a muffle furnace at 550°C overnight; then the weight of the remaining ash was measured. CHO was calculated by subtracting the sum of the percentages of the measured weights of fat, protein, moisture, and ash from the total weight (100g). Energy was expressed in kilocalories (kcal). Using Atwater calorie conversion factors, calorie values were calculated based on the total grams of protein, fat, and CHO, as 4, 4, and 9 kcal/g respectively18.\n\nThe macronutrient exchanges were determined based on the laboratory values provided from the analysis of 100 g of each dish. Wheeler and collaborators (Wheeler et al. 1996)19 described a round-off method which was used to yield exchange numbers. The macronutrient exchanges were calculated as follows:\n\nCHO exchange. The dish was not considered as one serving if it contained 1–5 g carbohydrate. If the food portion contained 6–10 g CHO, the dish was considered as half a serving and if it contained 11–20 g CHO, it was considered as one serving.\n\nProtein exchange. If the amount of protein ranged between 0–3 g in the meat and meat substitute dishes, it was not counted as a serving. If it contained 4–10 g protein, it was considered as one serving.\n\nFat exchange. If the amount of fat in food portions ranged between 0–2 g, it was not considered as a serving. However, if the dish contained 3g of fat, it was counted as half a serving and if it contained between 4 and 7 g fat, it was counted as one serving. Moreover, the amount of the meal (in grams) that yielded one CHO, one protein and one fat exchange was obtained by the calculation of CHO grams, protein grams, and fat grams by 15, 7, and 5, respectively.\n\n\nResults and discussion\n\nThe results of the analysis of 100g from each dish and Arabic sweet are presented in Table 2.\n\nThe daily recommended amount is 275 g/d for CHO (55%), 50g/d for protein (10%) and 78g/d for fat (35%).\n\nThe amount of moisture ranged from 31.3 % in Falafel to 91.5% in Baba Ghanouj. The protein content was lowest in Baba Ghanouj (1.1 %) and highest in Shawarma Dajaj (29.7%). Hindbeh bi zeit, which is fried chicory with onions, contained the top increased fat content (22.4%) of the analyzed dishes and was greater than the 6.7 % shown previously in 1970 in Lebanon11. Falafel, a fried dish, has the highest energy value (339 kcal/100 g). On the other hand, almost all the stews such as Yakhnet Bamia, Yakhnet Fassoulia and Yakhnet Mouloukhia have medium levels of energy ranging between 110 and 140 kcal per 100 g (Table 2). On the other hand, although the amount of energy in the dishes was identical, the dissimilarity in protein, CHO or fatty acid content had nutritional implications on health, since a high intake of CHO or fats is associated with a high-risk factor for non-communicable diseases5.\n\nThe nutrient goal represents the average intake that is compatible with the maintenance of good health in individuals20. According to the US Food and Drug Administration (FDA) definition, the daily value (DV) is described as the “reference values for reporting nutrients on the nutrition labels”. The percentage (%) of DV assists the consumer in recognizing how the serving of food, and its content in nutrients, fits into their daily diet. As per FDA regulations, the expression “high,” “rich in,” or “excellent source of” nutrients are used if the food has ≥20% of the daily value per reference amount. The terms “good source,” “contains,” or “provides” are used if the food yields 5–19% of the recommended dietary intake (RDI) per reference amount of the nutrient. Foods that carry <5% of the RDI from the nutrient per reference amount are considered to have low amounts. In our study, the contribution of each dish (per 100g) to the overall amount of CHO, protein and fat needed per day was calculated. The calculations are presented in Table 2.\n\nPellet et al., in 1970, showed high total fatty acid content in Lahm bi ajin (39.4%) which was higher than the reported value of 8% in our study11. There is limited available research on the composition of Lebanese traditional composite dishes, thus the results provided in this study were compared with data from other countries in the EMR11,21–26, mainly the amount of total fatty acid content in these foods. The amount of total fatty acid in the foods consumed in Lebanon, Bahrain, Kuwait, Jordan and Saudi Arabia are extremely important given the elevated prevalence of non-communicable diseases in the countries (Table 3)26. Compared to our findings, increased fatty acid content was observed in Falafel that was also reported in many other Arab countries (14.3% in Saudi Arabia to 18.4% in Jordan)23,24. The fatty acid content in Baba Ghanouj was double the level described in Jordanian Baba Ghanouj (9.4% and 5.4%, respectively)23 and triple the level reported previously in 1970 in the Lebanese Baba Ghanouj. Furthermore, the Kuwaiti Baba Ghanouj’s fatty acid content was moderately lower than that described in our study (8.7 and 9.4, respectively)22. Total fatty acid levels in Batata Mehchi ranged from 5.6% in Lebanon at 1970 to 5.9 % in Saudi Arabia11,24. Musiager et al., in 199821, found double the amount of fatty acid levels in Bourgul bi banadoura, Chichbarak and yakhnet Bamia when compared to our study (Table 3). The high fatty acid content of Fatayer Sabanikh was shown in almost all other studies enrolled in Arab countries; the content found in our findings was higher than all values reported in all countries (Table 3). In our study, Fattoush contained lower fatty acid content compared to the same dish of Kuwait origin (2.94% and 2.17%, respectively)22. In contrast, our results contradicted the values reported previously in Lebanon and in Jordan (6.3% and 8.6% respectively)11,23. Since Foul Moudamas is frequently consumed with added olive oil in Lebanon, thus, the availability of total fatty acids is high in this dish. The findings of our study show that the amount of fatty acid in Foul Moudamas was similar to those prepared in all the Arab countries except for Jordan23. The values of fat in the remaining dishes are shown in Table 2. Since the protein and CHO content of the meals studied were not explored in all the Arab country-based studies, it is impossible to compare these variables to our findings.\n\nNA: Non available; *Pellet, 1970; ^Musaiger, 1998;@Dashti, 2001; Ω Bawadi, 2009; €Alfares, 2018; ¥Musaiger, 2011.\n\nThe macronutrient exchanges yielded from the analysis of 100 g of each of the 30 dishes are shown in Table 4. In addition, the serving size of each dish which would provide one exchange of each macronutrient was calculated (Table 4). At least one exchange of starch was found in almost all dishes except Baba Ghanouj, Fattat homos, Fattoush, Hindbe bil zet, Loubia bil zet, Shawarma Dajaj, Shawarma Lahma, Yakhnet Mouloukhiya and Tabboula. The bean’s stew (Yakhnet Fassoulia) contained the highest amount of CHO exchanges (1.5 exchanges). The highest numbers of fat exchanges are found in Mousaka Batinjan, Fatayer sabanikh and Hindbe bil zet (Table 4). Less than 10 g of protein per portion size was determined in all dishes except for Kafta wa Batata and Shawarma Dajaj (Table 4).\n\nWM: whole milk; MFM: medium fat meat; LM: lean meat; HFM: high fat meat; RFM: reduced fat meat; Tbsp: Table spoon; 1 Tbsp: 15 g.\n\nThe amount of moisture ranged between 0.7% in Halawa light to 55.4% in Mohallabiya. The highest amount of protein was observed in Foustoukia (19.2%) and the lowest amount was in Moushabak (2.1%). Barazik, which is a sesame cookie cooked with butter, contains predominantly more than 40% of fat and had the highest energetic content (553 kcal/100 g). on the other hand, the least energy-dense foods were puddings (Riz bi halib and Mohallabiya).\n\nAs stated before, there is a gap in the research field on the composition of Lebanese traditional composite dishes and Arabic sweets, thus our findings were compared with data from other countries in the EMR11,21–26, mainly the amount of total fatty acid content in these sweets. Compared to our findings, the fatty acid content of Baklava in Lebanon did not differ to that in Jordan; however, it was lowest than the value reported in Bahrain. Lebanese Barazik, Ghourayba, Katayef Kashta, Maakaron and Moushabbak and Maamoul fostok had the highest fatty acid content compared to other EM countries (Table 3).\n\nThe macronutrient exchanges yielded from the analysis of 100 g of each of the 35 types of Arabic sweets are shown in Table 4. In addition, the serving size of each dish which would provide one exchange of each macronutrient was calculated (Table 4). At least one exchange of starch was found in all Arabic sweets. Barazik and Halawa contained the highest amount of fat exchanges (2 exchanges per serving). In addition, less than 5 g of protein per portion size was determined in all Arabic sweets (Table 4). These exchange lists for traditional dishes and Arabic sweets will assist healthcare professionals and dietitians in organizing culturally appropriate planning of healthy food, especially for those with non-communicable diseases. Furthermore, for effective medical nutritional therapy, these exchange lists may assist in monitoring food portions for these traditional dishes. In addition, the chemical composition of traditional foods is highly necessary in order to investigate the dietary consumption of populations25 and explore the impact of healthy food consumption on disease prevention26.\n\nThe principal limitation of this study is that the dishes and Arabic sweets analyzed were commercially prepared and the dishes’ ingredients were not reported, only recipes. Table 1 provides the ingredient quantities available from Lebanese cookbooks; however, all the relevant findings in our study were analyzed in an accredited laboratory. In addition, there are differences in ingredient proportions and cooking methods among various countries in the Arabian Middle Eastern and Gulf regions22,27.\n\nDespite these limitations, this study provides healthcare professionals and consumers with an updated food composition table and a new exchange list of Lebanese traditional dishes and Arabic sweets consumed in Eastern Mediterranean countries by providing the laboratory composition of 30 frequently consumed traditional foods and 35 frequently consumed Arabic sweets. This can help improving diet quality, the achieving weight loss and implementing self-control in obese or overweight individuals and/or individuals with diabetes.\n\n\nConclusions\n\nTo conclude, the Lebanese food exchange lists for the 30 frequently consumed Middle Eastern traditional dishes and the 35 mostly consumed Arabic sweets are now available28. This guide is a good source of information about the macronutrient content of traditional dishes and Arabic sweets cooked in Lebanon. It is important to consumers, dietitians, and researchers and it offers accessible, user-friendly, practical models and uses household measures that allow consumers, dietitians and healthcare professionals to develop meal plans with healthier selections. Jordan, Syria and Palestine also can get the maximum benefit from this work because of the similarity in their traditional dishes. Finally, Lebanese cuisine offers a wide variety of recipes rich in micronutrients which could prevent the rise in NCDs. Thus, data on micronutrients in traditional dishes and Arabic sweets would be of greater importance in halting the rise of diet-related NCDs in the EMR. Efforts like this will provide a solid framework for the implementation of nutrition policies and practices in the region.\n\n\nData availability\n\nOSF: Development of a Lebanese food exchange system based on frequently consumed Eastern Mediterranean traditional dishes and Arabic sweets\n\nhttps://doi.org/10.17605/OSF.IO/QKFN828\n\nThis project contains the following underlying data:\n\nData-F1000Research-exchange list-traditional dishes.xlsx\n\nExcel-Arabic sweets Exchange-F1000research.xlsx\n\nData are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).",
"appendix": "Acknowledgements\n\nThe authors would like to thank Dr. Joseph Matta and the laboratory technician Mr. Halim El Bayeh at the Industrial Research Institute and the Research assistants Ms. Nadia Hallak and Mrs. Iman Kheir at the Lebanese University. The authors confirm the approval of the acknowledged persons to be acknowledged.\n\n\nReferences\n\nMehio Sibai A, Nasreddine L, Mokdad AH, et al.: Nutrition transition and cardiovascular disease risk factors in Middle East and North Africa countries: reviewing the evidence. Ann Nutr Metab. 2010; 57(3–4): 193–203. PubMed Abstract | Publisher Full Text\n\nWorldometer: Lebanon population (1950-2020). Updated on October 5, 2020. Accessed October 5, 2020. Reference Source\n\nNasreddine L, Ayoub JJ, Hachem F, et al.: Differences in Dietary Intakes among Lebanese Adults over a Decade: Results from Two National Surveys 1997-2008/2009. Nutrients. 2019; 11(8): 1738. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorld Health Organization: Noncommunicable Diseases (NCD) Country Profiles. 2018. Reference Source\n\nWorld Health Organization: Healthy Diet. 2020. Updated 29 April 2020. Accessed October 5, 2020. Reference Source\n\nPetrescu D, Vermeir I, Petrescu-Mag R: Consumer Understanding of Food Quality, Healthiness, and Environmental Impact: A Cross-National Perspective. Int J Environ Res Public Health. 2019; 17(1): 169. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDucrot P, Mejean C, Aroumougame V, et al.: Meal planning is associated with food variety, diet quality and body weight status in a large sample of French adults. Int J Behav Nutr Phys Act. 2017; 14(1): 12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNational Institute of Health: Food exchange list. Updated 2020. Accessed October 5, 2020. Reference Source\n\nKhan M, Kalsoom S, Khan A: Food Exchange List and Dietary Management of Non-Communicable Diseases in Cultural Perspective. Pak J Med Sci. 2017; 33(5): 1273–1278. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEzzati M, Riboli E: Behavioral and dietary risk factors for noncommunicable diseases. N Engl J Med. 2013; 369(10): 954–64. PubMed Abstract | Publisher Full Text\n\nPellett P, Shadarevian S: Food composition tables for use in the Middle East. Beirut: American University of Beirut. 2nd edition. 1970. Reference Source\n\nMusiager AO, Miladi SS: Establishing Food Composition Data for the Near East. Bahrain, Bahrain Center for studies and research. 1998; 126. Reference Source\n\nBatal M, Hunter E: Traditional Lebanese recipes based on wild plants: an answer to diet simplification? Food Nutr Bull. 2007; 28(2 Suppl): S303–11. PubMed Abstract | Publisher Full Text\n\nIssa C, Salameh P, Batal M, et al.: The nutrient profile of traditional Lebanese composite dishes: comparison with composite dishes consumed in France. Int J Food Sci Nutr. 2009; 60 Suppl 4: 285–95. PubMed Abstract | Publisher Full Text\n\nGreenfield H, Southgate DAT: Food Composition Data Production Management and Use. Rome, Food and Agriculture Organization of the United Nations. 2nd edition. 2003. Reference Source\n\nHoteit M, Zoghbi E, Al-Iskandarani M, et al.: Nutritional value of the Middle Eastern diet: analysis of total sugar, salt, and iron in Lebanese traditional dishes [version 1; peer review: 2 approved]. F1000Res. 2020; 9: 1254. Publisher Full Text\n\nAssociation of Official Analytical Chemists: Official Methods of Analysis. Washington. 21st edition. 2019. Reference Source\n\nAtwater W, Woods C: The chemical composition of American food materials. Washington: Government Printing Office. 1896; bulletin no.28. 162. Reference Source\n\nWheeler M, Franz M, Barrier P: Helpful hints: Using the 1995 Exchange system for meal planning. Diabetes spectrum. 1995; 8: 325–326.\n\nFood and Drug Administration: How to understand and use the nutrition facts label. Updated March 11, 2020. Accessed June 10, 2020. Reference Source\n\nMusaiger A, Abuirmeileh N: Food consumption patterns of adults in the United Arab Emirates. J R Soc Promot Health. 1998; 118(3): 146–150. PubMed Abstract | Publisher Full Text\n\nDashti B, Al-Awadi F, Khalafawi MS, et al.: Nutrient contents of some traditional Kuwaiti dishes: proximate composition, and phytate content. Food Chem. 2001; 74(2): 169–175. Publisher Full Text\n\nBawadi H, Al-Sahawneh S: Developing a meal-planning exchange list for traditional dishes in jordan. J Am Diet Assoc. 2008; 108(5): 840–6. PubMed Abstract | Publisher Full Text\n\nAl Faris N: Nutritional Evaluation of Selected Traditional Foods Commonly Consumed in Saudi Arabia. J Food Nutr Res. 2018; 5(3): 168–175. Publisher Full Text\n\nBailey R, Dodd K, Gahche J, et al.: Best Practices for Dietary Supplement Assessment and Estimation of Total Usual Nutrient Intakes in Population-Level Research and Monitoring. J Nutr. 2019; 149(2): 181–197. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNoncommunicable diseases progresses monitor 2020. Geneva: World Health Organization. 2020. Reference Source\n\nMusaiger A: Food composition tables for kingdom of Bahrain. INFOODS regional database center. Bahrain: Arab Center for Nutrition. 1st edition. 2011. Reference Source\n\nHoteit M, zoghbi E, Rady A, et al.: Development of a Lebanese food exchange system based on frequently consumed Eastern Mediterranean traditional dishes and Arabic sweets. 2020. http://www.doi.org/10.17605/OSF.IO/QKFN8"
}
|
[
{
"id": "77080",
"date": "03 Mar 2021",
"name": "Yonna Sacre",
"expertise": [
"Reviewer Expertise Nutrition",
"epidemiology",
"public health",
"obesity ."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAfter reviewing the paper, I find the topic very original and useful since this is a need in our community to have a reference on which dietitians can rely to provide nutritional recommendations for patients in the clinical setting.\nOn the other side, I have a few suggestions and questions for clarification:\nWhat are the 150 samples, since it was about the 65 items selected?\n\nI recommend adding the quantities of the ingredients used in the recipes of the meals selected in table 1, noting that the method of preparation and quantities can be different from one region to another.\n\nI recommend adding a more recent reference to be based on in order to explain how the 30 items were selected.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "78747",
"date": "03 Mar 2021",
"name": "Majid Mqbel Alkhalaf",
"expertise": [
"Reviewer Expertise Nutrition",
"dietary and nutrition assessment methods",
"public and clinical nutrition"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study is well written. I have few and minor comments need to be addressed below:\nAbstract:\nThe last part of the results seems to be part of the methodology. The results should be well written.\nIntroduction:\nThe study aims need to be unified. The article aim is not matching with the abstract aim (the abstract aim is better).\nMethodology:\nThe authors need to clarify why they calculated the carbs rather than analyze them in laboratory as protein and fats.\n\nThe authors also need to clarify how the food portions have been selected.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-12
|
https://f1000research.com/articles/9-1154/v1
|
18 Sep 20
|
{
"type": "Opinion Article",
"title": "Option-based guarantees to accelerate urgent, high-risk vaccines: a new market-shaping approach",
"authors": [
"David Manheim",
"Derek Foster",
"Derek Foster"
],
"abstract": "Accelerating the availability of COVID-19 vaccines is critical to preventing further waves and mitigating the impact on society. However, preparations for large-scale manufacturing, such as building production facilities, are typically delayed until a vaccine is proven safe and effective. This makes sense from a commercial perspective, but incurs great costs in terms of lives lost and damage to the economy. Several policy options are available to reduce this delay, all of which involve incentives or subsidies to invest in production facilities. We review existing approaches, then propose a novel alternative using “option-based guarantees” in which the government commits to paying a proportion of the manufacturer’s preparation costs should the product turn out not to be viable. Counterintuitively, this “payment for failure” is appropriate because in the case of success, a company makes a profit from the product itself, and does not need additional money from the government. While other approaches have critical roles, we argue that option-based guarantees are the most promising approach to ensuring a rapid vaccine for COVID-19. Compared to the alternative approaches, they reduce both costs to the government and risk to the companies, while maintaining an incentive to produce a high-quality product quickly and at scale.",
"keywords": [
"Vaccines",
"Advance market commitments",
"Health economics",
"Financing mechanisms",
"Public Policy"
],
"content": "Introduction\n\nThe urgent need for a COVID-19 vaccine is widely recognized, with governments setting ambitious targets for timelines, testing, and approval. Promising candidates have been identified, but most will optimistically take at least several months to complete Phase 3 studies – perhaps less if human challenge trials are permitted1, and perhaps much longer if the best-case assumptions do not apply.\n\nA recent op-ed by Nobel Prize winner Michael Kremer and colleagues notes that “we need multiple shots on goal” because a 90% chance of finding a successful vaccine requires funding 15–20 candidates. While the top candidates are already pushing forward with manufacturing well in advance of clarity about efficacy, these vaccines may not prove effective, or safe enough to be approved by regulators. In this case, there are economic issues that may prove critical barriers to much needed vaccines. For example, companies with a smaller chance of being first to market have little incentive to invest in very expensive production facilities before their product achieves regulatory approval2. In the case that the top candidates fail, scaling up production of alternative vaccine candidates is likely to add months to the overall timeline, costing thousands of lives and tens or hundreds of billions of dollars of lost GDP – far more than the cost of preparing to manufacture products that do not end up reaching the market.\n\nPerhaps more concerning, the first successful vaccine is not necessarily the best option.\n\nVaccine efficacy can vary widely. It is very likely that at least one of the front-runner candidates will be approved, but still be significantly less effective than, or have issues avoided by, other approaches. The FDA has recently said that it will approve vaccines which are only 50% effective, far short of the level needed for herd immunity, much less near full protection3. If early vaccines are in fact only partially effective, or those vaccines are found to have safety issues that present major drawbacks despite approval, investment in alternatives may become a priority. In this fairly likely case, the delay in investment could lead to delays in scaling up production of potentially safer or more effective candidates.\n\nA more diverse portfolio of options reduces the risk that a small group of vaccines are all insufficient, or that none are safe or effective enough to be approved. Even if this risk were low, the costs of failure would be high enough to warrant additional efforts. A critical task for governments is therefore to reduce delays due to risk-avoidance by firms so that multiple vaccines are developed and manufactured, thereby increasing the probability of finding a viable candidate and accelerating its availability. Before introducing solutions, however, it is worth reviewing part of the current landscape on which these solutions will operate.\n\nThere is a variety of mechanisms by which biomedical research, including vaccines, can be funded. This starts with the funding of basic research by governments, universities, and foundations, and continues through to mechanisms we discuss below for supplementing extant measures. In between, there is translational research, corporate research and development, and production costs. The funding for these stages comes from capital markets or private investments, or from nonprofits, nongovernmental organizations, and governments.\n\nDespite these funding sources, there seem to be market failures, where socially beneficial products like vaccines are not produced because the firm-level incentives are insufficient to overcome the firm-level costs and risks.\n\nIt is useful to distinguish between two market failures that are implicated here. The first is a buyer/seller mismatch, where a market would exist for a product, but producers are unaware, or are concerned that the market will not buy sufficient quantity to justify investment. The second is a risk tolerance mismatch, where the socially optimal level of risk-taking is higher than the level for individual firms, so firms will not take risks in developing something that may not be purchased. This mismatch is particularly acute for smaller firms that cannot afford to absorb the loss from failures, but it also applies to larger firms, which may have high opportunity costs of capital.\n\n\nPotential solutions\n\nFor a pressing need like COVID-19 vaccines, a solution to the market failure provided by government or another organization must address multiple economic problems simultaneously. First, it needs to create incentives for companies and investors to take on high-risk projects, many of which have individually low probabilities of success. Second, it needs to create a motivation to scale up production capacity before the success of different approaches is known, to ensure timely availability. Finally, it should do these two things in ways that do not incentivise throwing money at projects that are too unlikely to succeed.\n\nThese different issues apply differently to different firm types, of course. Smaller firms have less capital, and less ability to raise capital, as well as a lower ability to diversify against the risk of long-shot projects. They are also far less able to invest in manufacturing. Larger firms, on the other hand, are far better at regulatory capture and finding ways to benefit from programs that could allow wasteful spending.\n\nSeveral classes of solutions have historically been used to foster innovation: prizes, government programs, guaranteed demand, and for vaccines specifically, advance market commitments. While all are limited in their ability to compress timelines, each has significant advantages in addressing at least one of the market failures, or is useful for some firm types. A new approach we propose, option-based guarantees, seems particularly well suited to mitigating the risk tolerance mismatch for smaller and newer firms, some of which are pursuing novel ideas. This is critical to ensure a diversity of COVID-19 vaccine candidates, and may foster innovation in vaccines more broadly as well.\n\nGovernments and private philanthropists can offer financial rewards for breakthroughs or solutions to a scientific problem. In the current scenario, a prize could be offered to companies that have a vaccine approved. Alternatively, companies could compete to offer the best idea for rapidly scaling up vaccines, and a contract to produce them could be part of the prize. By only paying out for successful solutions – or perhaps not any, if certain criteria are unmet – prizes are a fairly inexpensive option that incentivises innovation. This approach has a long history4, and has been used by governments recently in the US5 and abroad6.\n\nThese cases illustrate that prizes are useful for spurring new areas of research, but primarily attract large firms or well-funded new entrants. This is partly because they require investors to supply and risk capital, while paying nothing to firms that “lose,” thereby failing to address (or worsening) risk tolerance mismatches. This is particularly difficult for smaller firms. The hope is that many firms would participate due to the increased gain in case of success, but there will not be adequate financial incentive for them to do so unless the probability of success is high, the investment is nearly viable without a prize, or the size of the prize makes the investment an expected gain despite the risk. These conditions seem unlikely to be met in the current circumstances, where it is likely that only a few of the many vaccines will ever be made widely available.\n\nAnother disadvantage relates to timing. Prizes are often appropriate for early-stage investments in projects that have little short-term chance of profitability, but which have a clear path to success. In the current situation, where speed is critical, the incentive would ideally be immediate and certain, rather than contingent and post-success. This means that they do not address the risk mismatch issue. Not only this, but unlike the alternatives outlined below, prizes cannot be used or borrowed against to fund the project.\n\nThese drawbacks mean that, in the present case, prizes are unlikely to lead to a diversity of high-risk approaches.\n\nAnother alternative is the “Apollo program” or “Manhattan Project” approach, where the government directly invests massively in projects. Government agencies often enter into agreements, as the US Biomedical Advanced Research and Development Authority and others have done for COVID-19, to develop treatments, vaccines, and diagnostic tools. But this approach requires government expertise, and requires selecting one or a few projects to focus on. As a consequence, projects that are not funded, which are the majority of projects, languish unless and until the primary projects fail. It also favours well-connected and larger firms.\n\nA common variant of direct government investment is public-private partnerships (PPPs). These share the risk between government and private companies, which allows the government to leverage private companies’ expertise. PPPs can be an effective means of achieving social objectives, but such deals generally take a long time to negotiate and implement, are complex in ways that can make regulatory capture a larger problem, and are usually best deployed when a single approach is needed.\n\nThus, a government-led approach is promising for relatively predictable projects, such as building test-and-trace infrastructure. In contrast, a key goal of higher-risk investments in vaccine candidates is to build a diverse portfolio of investments with an overall high probability of ensuring needs are met more quickly than markets allow on their own. Because multiple approaches are needed and the negotiation and development process is limited by government capacity, neither direct government funding nor PPPs are likely to be the best option for creating a large and diverse portfolio.\n\nAnother approach is to pre-order vaccines. This has been done successfully by governments in the past, and in this case it could provide capital well before efficacy or safety is established. This would legally guarantee that producers have a market and that the company will supply the product, thereby reducing risk to both parties.\n\nKremer, Levin, and Snyder (2020)7 present a version of this called advance market commitments (AMCs), which are purchase orders contingent on successful development. These have been used successfully for “technologically close products,” such as a pneumococcal conjugate vaccine. They are a particularly valuable tool when few of those who would benefit from a vaccine are able to afford it, in which case development is only economically feasible with an outside funder, though this does not apply to COVID-19.\n\nHowever, this approach has several drawbacks, which the authors identify. First, both purchase orders and AMCs require choosing which approaches to fund. Governments’ track record of “picking winners” is less than stellar, and any such decisions would inevitably be highly politicized. Second, the government or other funder would need to negotiate prices and contractual details before companies would be able to start. Not only might this be a lengthy process, but vaccine manufacturers have a significant advantage in such negotiations, and there is potentially significant room for regulatory capture or windfall profits on the part of companies.\n\nIn addition to these problems, for COVID-19 this approach requires the government or other sponsor to commit to purchases of many still-unproven products. For this reason, it would need to contract with many different companies – the more the better, to improve the chances of a viable product – but committing to purchasing many vaccines when only a few will be needed would be very wasteful. In addition, purchase orders would potentially involve commitments to purchasing products not shown to be safe, which may be illegal for government agencies. It also gives far less incentive for producers to improve quality, speed, or cost-effectiveness through innovation. In the present crisis, these shortcomings are especially pressing.\n\nA variant on AMC proposed by Athey et al.8 to address COVID-19 would combine the direct investments (“push”) of PPPs with the typical AMC mechanism of a precommitment to purchase (“pull”) the first resulting product to come to market. This push-pull approach improves on both direct investment and PPPs; but because the government funding for purchasing may be exhausted before it reaches market, it has drawbacks similar to prize competitions in that it leaves companies with the bulk of the risk from overproduction if they are not first to market. This means that AMCs are an option regardless of which “push” option is selected.\n\n\nOption-based guarantees\n\nWe suggest a new approach for governments to “push” vaccine production, which is to enter into agreements with companies using put options. A put option (as in “put up for sale”) gives the holder the right, but not the obligation, to sell an asset, by (or on) a specified date, to the provider of the put. In this case, the put option would give companies the right to sell a portion of their investment in vaccine production to the government, i.e. at a guaranteed loss. Because there is no obligation to exercise the put, companies could sell viable products as usual, and would only use the option if their product turns out to be non-viable. If structured well, options can also align incentives in several other ways.\n\nTo understand how this would work, we start with an example, then note possible variants on the idea. Following this, we discuss the advantages of the proposal, and the implementation and political challenges it may face.\n\nSuppose, optimistically, that a manufacturer thinks it will be able to produce 100 million doses of a vaccine within six months, but is delaying investment in production facilities because the vaccine’s Phase 3 trial results will not be available for a year. Once the result is known, it will begin to invest in the production, and if there are no unforeseen obstacles, have the vaccine available six months later.\n\nUnder the proposed scheme, the company can approach the government with a budget and a timeline, and the government can agree to provide a put option that allows the company to recoup, say, 90% of its eventual costs, capped at the company's initial project cost estimate, in exchange for the facility and equipment. If the vaccine is viable, the company would not exercise the option, the government would pay nothing, and the company would be able to sell the vaccine normally. If found non-viable, however, the company would have an incentive to stop production and exercise its option as soon as possible. When the option is exercised, a financial audit of costs would take place, and the government would accept delivery of any items purchased, built, and/or produced in exchange for 90% of costs. Delivery upon contract termination is both a potential avenue for the government to recoup costs, and a means to ensure companies do not gain windfall profits from declaring a program a failure, then selling assets.\n\nA number of variants of this approach are possible, and three are worth highlighting: declining payouts, priced contracts, and conditions on sales. The first two modify the incentives, while the final variant addresses additional concerns about the availability and price of the vaccine.\n\nFirst, the payout for the put options could be declining over time, so that the payment is, say, 95% at the outset, and declines by a specified percentage, say 1%, each month. This will incentivize companies to exit as soon as possible if they think the project will fail.\n\nSecond, instead of providing options to companies for free, the government could charge for the contracts. This would further dissuade unqualified or undercapitalized companies from taking huge immediate risks with small probabilities of success. Prices for such contracts would still need to be a small fraction of the actuarially fair price, otherwise the scheme does not actually provide the needed incentives.\n\nThe last of the variants, conditions on sales, is somewhat different, since it is largely unrelated to the options themselves. Put options do not ensure the final vaccine is available at a reasonable price, but nor do they preclude other policy solutions. Because recipients of these options benefit from the program’s guarantee, in exchange for participation the government may claim priority for purchases, cap the profit margin on sales to the government, or cap the price paid by the government for a product. This is reasonable, but care should be taken not to either significantly reduce the incentive to invest, or greatly slow down the process of agreeing to deals. Note also that governments that took on risk to ensure investment in a product might also want to prioritize domestic purchases, rather than allowing them to be sold internationally. While posing additional challenges for international cooperation, this does not differ from other solutions, and can be addressed in similar ways, such as through international coalitions and agreements. Finally, we note that it may be less than ideal to pursue multiple goals with a single policy. If price controls or similar constraints are desired, they do not need to be tied to funding mechanisms.\n\n\nDiscussion\n\nThe use of put options to accelerate vaccine production is a somewhat novel idea. Though it creates incentives and shapes markets in ways similar to other policy tools, it is worth looking at the unique advantages and drawbacks of this approach.\n\nFirst, commercial companies can continue to use traditional methods for financing and operating their businesses without unnecessary government supervision or contracting. New companies can also use these options to help them secure funding from private investors, making the program more equitable to newer firms without requiring direct government investments.\n\nSecond, it provides incentives for starting production earlier, but preserves normal market mechanisms to provide high-quality products. Because the market is competitive, and it will be unclear whether other firms will be earlier to market or have a safer product, perhaps with higher efficacy, having an earlier and/or better product on the market will increase sales, and therefore profit.\n\nThird, this approach is guaranteed to have lower cost than direct investment to pay for high-risk products, while preserving market incentives. Increasing the possible cost savings, the government may also be able to resell some items. For example, a plant or equipment designed to manufacture an ineffective vaccine could be resold and adapted to produce a different one. There have been intermittent shortages of other vaccines, so excess capacity may not be entirely wasted.\n\nFourth, unlike advance payments or contracts, put options do not subsidize companies to undertake projects that they expect cannot succeed, but do allow them to take additional risks in order to accelerate production.\n\nLastly, it can be implemented more quickly than the alternatives. Private funding that relies in part on the known risk reduction from the government guarantee could replace direct payment by the government. Not only that, but because the calculation of the payment is deferred, the approach could potentially be implemented without extensive and slow negotiations – a very important consideration in the current circumstances where speed is critical.\n\nThe most critical decision for option-based guarantees is the structure of the payments. There is a tradeoff between payment amounts and incentives for firms. The ideal percentage of costs to reimburse with such a program requires economic analysis, weighing the cost of such a program, which likely involves payment of all or all but one of the put options, against the public benefit of a more rapidly available vaccine.\n\nHowever, these challenges are not unique to put options. Any incentive for production requires the government to choose projects to fund, and then pick a level of funding. By guaranteeing the payout will be less than the investment, providing incentives for early termination, and enabling cost-recovery through reselling assets, put options reduce the risk of corporate profiteering relative to direct investment and PPPs. In addition, as noted above, AMCs and prizes are compatible with put options as well as those two alternatives.\n\nPerhaps a greater disadvantage is that this is a novel suggestion: similar programs have not, to our knowledge, been tried before. While the proposal has attempted to consider implementation, unforeseen challenges may arise.\n\nAn option-based guarantee is a potential political liability. The program may appear wasteful because, perhaps counterintuitively, payments are only made for unused products and abandoned projects. This may make the program politically unpopular, especially if no successful vaccines are generated, or if the costs outweigh the value of the successes. In addition, put options do not address pricing or local supply, so they do not guarantee that any viable vaccines would become widely available. This potentially makes options far better for funding research or manufacturing capacity, rather than the products themselves.\n\nAt least four factors should mitigate the political risk. First, as explained above, the payment structure should help to minimise waste, giving less ammunition to opponents. Second, the program could (quite accurately) be presented as evidence of government action to combat the pandemic, which is likely to be popular in the current climate. Third, the government could of course claim credit for any successful products emerging from the program – products that would save or improve many of their constituents’ lives. Fourth, additional mechanisms or clauses in the contract, such as the variants described above, could ensure that the products are sold at a reasonable price.\n\n\nConclusion\n\nThe optimal approach – or more likely, combination of approaches – to developing healthcare products will vary by disease, time period, and urgency. For this reason, we conclude with a discussion of a few contexts in which option-based guarantees seem most useful, as well as areas where the other reviewed approaches are likely to be superior.\n\nIn the case of vaccines for COVID-19, we think that option-based guarantees for constructing production facilities are the best alternative for candidates that are promising but whose viability, large-scale manufacturing methods, and/or quantity required are substantially uncertain. These types of guarantees are potentially useful in other areas as well, but the selection of funding mechanisms should be made on the basis of the needs and characteristics of each specific product type and need, both in combating COVID-19 and for future pandemics.\n\nWe also think that option-based approaches may be useful for funding very costly Phase 2 and 3 trials, perhaps in place of the current model of directly funding large firms. This is especially true for smaller firms that may otherwise delay or under-power vaccine trials to mitigate risks. In this case, they could be paid part of their costs if the product fails to gain approval.\n\nBased on our review, PPP or direct purchase orders are much more appropriate than option-based guarantees for low-risk products. For example, purchasing a large number of a certain vaccine that is already near approval would be ideal if the safety, accuracy, cost, and quantity required are known, the company is trusted, and the paperwork can be done quickly. Antivirals or antibiotics that are already being produced and are likely to be useful may also fall into this category. This approach can also mitigate the risk that a company will be slow to respond to anticipated but uncertain demand.\n\nAthey et al.’s advance market commitments are useful when products are higher risk, but are close enough to being ready that price and quantity negotiations can take place before the decision is made. In such a case, option-based guarantees are less helpful.\n\nIn other contexts, such as when innovative solutions with low capital costs but significant conceptual innovation are likely to be needed, prizes for successful innovation are another useful approach for creating incentives for investments. This would potentially be true for new types of point-of-care tests for active infection and new monitoring technologies. Both AMCs and prizes are also useful in combination with any of the proposed “push” funding approaches, including option-based guarantees.\n\nOver the coming weeks and months, choosing the right funding mechanisms could save tens of thousands of lives. The sooner companies start these investments, the sooner their products can reduce economic damage, mitigate ongoing risks to the safety of vulnerable communities, and staunch the very high human costs of the ongoing COVID-19 pandemic.\n\n\nData availability\n\nNo data is associated with this article.",
"appendix": "Acknowledgements\n\nDM is funded via non-specific grants from various nonprofits and foundations, and DF is funded by Rethink Priorities (www.rethinkpriorities.org).\n\n\nReferences\n\nEyal N, Lipsitch M, Smith PG: Human Challenge Studies to Accelerate Coronavirus Vaccine Licensure. J Infect Dis. Oxford Academic; 2020; 221(11): 1752–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMilstien J, Batson A, Meaney W: A systematic method for evaluating the potential viability of local vaccine producers. Vaccine. 1997; 15(12–13): 1358–63. PubMed Abstract | Publisher Full Text\n\nCenter for Biologics Evaluation and Research: Development and Licensure of Vaccines to Prevent COVID-19. U.S. Food and Drug Administration. FDA; 2020; [cited 2020 Aug 2]. Reference Source\n\nClancy MS, Moschini G: Incentives for Innovation: Patents, Prizes, and Research Contracts. Appl Econ Perspect Policy. 2013; 35(2): 206–41. Publisher Full Text\n\nGustetic J: This October, the White House Celebrates Over $150 Million in Prize Competitions Since 2010. whitehouse.gov. 2015; [cited 2020 Aug 2]. Reference Source\n\nMasters WA, Delbecq B: Accelerating Innovation with Prize Rewards: History and Typology of Technology Prizes and a New Contest Design for Innovation in African Agriculture. IFPRI, Discussion Paper No. 00835. 2008. Reference Source\n\nKremer M, Levin J, Snyder CM: Advance Market Commitments: Insights from Theory and Experience. AEA Pap Proc. 2020; 110: 269–73. Publisher Full Text\n\nAthey S, Kremer M, Snyder C, et al.: Opinion In the Race for a Coronavirus Vaccine, We Must Go Big. Really, Really Big. The New York Times. [cited 2020 Aug 2]. Reference Source"
}
|
[
{
"id": "73554",
"date": "09 Nov 2020",
"name": "Daniel L. Tortorice",
"expertise": [
"Reviewer Expertise Macroeconomics",
"Finance",
"Value of Vaccination"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper considers how governments can accelerate the availability of COVID-19 vaccines. Since investment in production facilities is a large risky investment, without government intervention vaccine manufacturers would be expected to delay this investment until after successful Phase 3 trials. Since this delay is socially costly, government intervention is justified. The authors propose a new financing mechanism to speed up investment. Specifically, the government will provide vaccine manufacturers with a put option, that if exercised, will require the government to purchase vaccine manufacturing equipment and facilities for a prespecified amount, for example 90% of the cost of developing the facilities. The paper then compares the proposed financing mechanism with existing financing mechanisms and discusses the relative advantages and disadvantages of the proposed mechanism.\nAccelerating the development of a COVID vaccine is one of the most pressing challenges of our times and I commend the authors for inventing and developing a new financial mechanism to aid in this acceleration. Their proposed mechanism is a clever compromise between the direct funding of manufacturing approach taken in the US and the advanced purchase commitment approach taken in Europe.\nI have two concerns about the proposed mechanism which I outline below and several minor points which might help the authors improve the article.\nThe article does not discuss how manufacturers will be chosen to receive this option. Since the authors suggest using this mechanism to fund vaccine trials as well as manufacture, the number of companies that one might reasonably consider as applicants is quite large. The Milken Institute estimated there are 214 Covid-19 vaccines in development. Additionally, the option has no downside for the vaccine developer, so there is no reason they would not apply. The mechanism by which vaccine producers are chosen is essential. Because the option is worth more, the more likely you are to fail, offering the option will select producers with the least viable prospects. Additionally, given the difficulty of separating out project specific costs and costs the firms would have paid anyway, the option would attract firms with lower probabilities of success that see the option as away to increase general funding for their operations. The authors should make clear if and how they intend to limit the number of producers who receive the option.\nAnother issue that could be discussed is the socially optimal number of vaccines. From a social perspective, we would like a large number of safe and effective vaccines. This is both because increased competition benefits consumers through lower prices and population subgroups may respond differently to various vaccines. This reason is why wealthy countries have invested in a diverse portfolio of vaccine and agreed to advanced purchase commitment. However, even if a vaccine is proven safe and effective, a vaccine manufacturer may exercise the put option if they view the vaccine market as crowded and not profitable enough even if additional vaccines would be socially beneficial.\nSome more minor points. [Note the first three points below are why I answered partly to the question “Are all factual statements correct and adequately supported by citations?”]\nA citation should be provided for the Michael Kremer op-ed article.\nThe estimate of the monthly costs of the pandemic “tens or hundreds of billions of dollars of lost GDP,” seems too conservative. It is clearly in the hundred of billions of dollars per month even if you are only counting the US. For example, Cutler and Summers (2020)1 puts it in excess of 800 billion per month.\nThe claim that a vaccine with 50% efficacy is far short of the level of protection needed for herd immunity does not seem to be supported by the literature. See for example, Gomes et al. (2020)2 which estimate 60-70% immunity as sufficient for heard immunity.\nI found the paragraph on page 4 about the timing of prizes lacking precise reasoning. For example, while they cannot be borrowed against, they can help vaccine developers find investors by increasing their profit if they produce a successful candidate. Additionally, the option the authors propose is both contingent and in the future, so this line of thought doesn’t really make a distinction between prizes and the option.\nIt is incorrect to say AMCs require governments to pick winners. The typical AMC requires the government purchase the vaccine only when it is approved as safe and effective by a regulatory body like the FDA or WHO. As such it tends to select firms who believe they have a high likelihood of success.\nOn page 5 there is a typographical error. “Put options do not ensure the final vaccine is available at a reasonable price, but nor do they preclude other policy solutions.” The word but is not needed.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": [
{
"c_id": "6195",
"date": "11 Jan 2021",
"name": "David Manheim",
"role": "Author Response",
"response": "Thank you for your review and comments. We have responded to the points raised below, and are in the process of posting a revision which clarifies many of these points in the text. We have responded to the two major points, then responded to each of the minor points below. First, the review notes: \"The article does not discuss how manufacturers will be chosen to receive this option. Since the authors suggest using this mechanism to fund vaccine trials as well as manufacture, the number of companies that one might reasonably consider as applicants is quite large. The Milken Institute estimated there are 214 Covid-19 vaccines in development. Additionally, the option has no downside for the vaccine developer, so there is no reason they would not apply. The mechanism by which vaccine producers are chosen is essential. Because the option is worth more, the more likely you are to fail, offering the option will select producers with the least viable prospects. Additionally, given the difficulty of separating out project specific costs and costs the firms would have paid anyway, the option would attract firms with lower probabilities of success that see the option as a[ ]way to increase general funding for their operations. The authors should make clear if and how they intend to limit the number of producers who receive the option.\" We agree that there is some amount of discretion needed on the part of governments, but as we noted in the paper “these challenges are not unique to put options. Any incentive for production requires the government to choose projects to fund, and then pick a level of funding.” While the existence of 214 candidates seems daunting, it is immediately clear that some are more advanced along the track than others (and at this point, many are finished) so the government's choices can be far better informed. Lastly, as a policy analysis, we think the paper is more helpful explicitly leaving this choice to policymakers and the political process, where it will inevitably happen. However, the mechanism is not costless for manufacturers, since less than the full cost is paid. The option is more valuable to a less viable candidate, but because they have some portion of the funds at risk, despite a guaranteed partial repayment upon failure, they will be less interested in making an investment that is nearly certain to lose money, even if it is only a fraction of the investment. Additionally, the put option does not provide money upfront, and this means that most smaller manufacturers would need loans. Because of the very high cost of building manufacturing capacity, an option which repays “90% of costs,” as the paper suggests, would mean that failure is still likely to bankrupt smaller firms. For larger firms, we agree that there is a risk that they use the put option to reduce their cost of capital rather than spend more. For smaller firms, lenders will plausibly be willing to make such loans if there is an option in place, but it by no means suggests that lenders will give money to projects which are near-certain to fail – despite the ability to recover most of their investment due to the option, they are unlikely to be repaid in full in that case. For this reason, we think that the use of options will be limited to firms with a reasonable chance of success – and since only a finite number of vaccines are needed, the number of such firms which can succeed is, by nature of the problem, small. Second, the review notes that \"Another issue that could be discussed is the socially optimal number of vaccines. From a social perspective, we would like a large number of safe and effective vaccines. This is both because increased competition benefits consumers through lower prices and population subgroups may respond differently to various vaccines. This reason is why wealthy countries have invested in a diverse portfolio of vaccine and agreed to advanced purchase commitment. However, even if a vaccine is proven safe and effective, a vaccine manufacturer may exercise the put option if they view the vaccine market as crowded and not profitable enough even if additional vaccines would be socially beneficial.\" We agree that this could be a concern, but suggest two reasons it is unlikely. First, the manufacturer would be publicly claiming failure in a way that is nearly certain to create public backlash. Second, the public is by no means powerless in such a scenario – any philanthropist or government that wishes to see the vaccine produced has every right, and a significant incentive, to offer funds or contracts to urge a producer of, say, the 5th or 6th vaccine which is found effective and safe to continue despite their otherwise unprofitable situation. Regarding the more minor points, we have responded below. A citation should be provided for the Michael Kremer op-ed article. Thank you - this oversight is now corrected. The estimate of the monthly costs of the pandemic “tens or hundreds of billions of dollars of lost GDP,” seems too conservative. It is clearly in the hundred of billions of dollars per month even if you are only counting the US. For example, Cutler and Summers (2020)1 puts it in excess of 800 billion per month. Thank you for pointing this out - we have clarified. At the time the paper was initially written, the situation was less clear, and the estimate was clearly very conservative/optimistic. The claim that a vaccine with 50% efficacy is far short of the level of protection needed for herd immunity does not seem to be supported by the literature. See for example, Gomes et al. (2020) which estimate 60-70% immunity as sufficient for [herd] immunity. The occasional discussion of successful immunization with lower numbers of vaccinated individuals rests on a number of very dicey assumptions. While the point is mostly irrelevant to the current paper, it is sufficient to note that, to be generous to the assumptions of those promoting the vaccination of superspreaders, it seems extremely optimistic to think that countries which cannot manage viable test and trace programs will nonetheless be able to identify the future superspreaders prospectively. This also ignores the fact that “superspreader” is a characteristic of the circumstances, and the constantly shifting set of contacts individuals have over time – not simply a property of an individual which can be found in advance. Beyond that, the assumption is that almost all of these (by assumption) foolhardy individuals – those most likely to contract and spread COVID-19 – would be willing to be vaccinated. For a very recent preprint that lays out some of these issues, see Fox et al. [1] I found the paragraph on page 4 about the timing of prizes lacking precise reasoning. For example, while they cannot be borrowed against, they can help vaccine developers find investors by increasing their profit if they produce a successful candidate. Additionally, the option the authors propose is both contingent and in the future, so this line of thought doesn’t really make a distinction between prizes and the option. The key difference between prizes and options is the combination of timing and contingency. As noted, prizes cannot be relied on or borrowed against, so even if they change the overall incentives to succeed, they won’t impact overall timelines as significantly. Options, because they are a financial guarantee which can be borrowed against, allow and encourage earlier and larger investments in production; the contingency ensures profitability, rather than enhancing it, as occurs in the case of prizes. It is incorrect to say AMCs require governments to pick winners. The typical AMC requires the government purchase the vaccine only when it is approved as safe and effective by a regulatory body like the FDA or WHO. As such it tends to select firms who believe they have a high likelihood of success. We agree that not all AMCs require picking winners, and this is a tradeoff between early funding, and picking winners. We have edited to clarify that this is a tradeoff. By the time a typical AMC can be put in place, market funding is likely available, and it does not reduce risks. If the purchase agreement is early, the company bears the risk of it not being approved - as seems to have occurred recently in Australia with the UQ/CSL vaccine. On page 5 there is a typographical error. “Put options do not ensure the final vaccine is available at a reasonable price, but nor do they preclude other policy solutions.” The word but is not needed. Thank you. The British and American usage seem to differ slightly, and it can be used since this contrasts a positive and negative point, but we have amended it to omit the word. 1. The COVID-19 herd immunity threshold is not low: A re-analysis of European data from spring of 2020 Spencer J. Fox, Pratyush Potu, Michael Lachmann, Ravi Srinivasan, Lauren Ancel Meyers medRxiv 2020.12.01.20242289; doi: https://doi.org/10.1101/2020.12.01.20242289"
}
]
},
{
"id": "74620",
"date": "26 Nov 2020",
"name": "Rino Rappuoli",
"expertise": [
"Reviewer Expertise infectious diseases",
"vaccines"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper proposes one alternative mechanism to increase the incentives to develop vaccines which do not have an attractive market. The authors call this mechanism “option-based guarantees.” Overall, proposing an additional model to develop vaccines when there is a market failure is fine, because we need as many options as possible.\n\nHere are some comments:\nThis proposal is too late for COVID-19 vaccines, as they have been already developed.\n\nThe authors argue for an option-based approach to help manufacturers in investing in new vaccines by de-risking eventual failures with public money. This approach might be relevant to incentivize development efforts in areas of high public health and medical need.\n\nOverall, the manuscript is well written and brings interesting insights on mechanisms that will likely be at the center of the public attention in the upcoming post-discovery phase of COVID-19 vaccines. However, its content seems to be completely detached from present day issues and it is very theoretical. It is not clear how the proposed solution integrates with - or differentiates from - the dozens of private and public initiatives that support COVID-19 vaccine development. Are there already option-based approaches being used for COVID-19 vaccines? If not, why? Are governments or public entities considering such approaches for some of the vaccines under development?\n\nMinor: the authors should update the introduction session as - as of today - two vaccine candidates have already demonstrated high efficacy in interim ph3 analysis without the need of human challenge trials and with much stringent timelines than reported in this manuscript.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": [
{
"c_id": "6196",
"date": "11 Jan 2021",
"name": "David Manheim",
"role": "Author Response",
"response": "Thank you very much for the helpful notes and concerns. We are in the process of finalizing a revised version of the manuscript that addresses the concerns. We respond to the individual points below. 1. This proposal is too late for COVID-19 vaccines, as they have been already developed. While there are now approved vaccines, manufacturing for COVID-19 vaccines is not obviously a fulfilled need, and still-unapproved second-generation vaccines may still have an important role in fulfilling global demand. At the same time, we agree that at the current time, the proposal is far less relevant. At the time it was proposed and initially discussed with policymakers, this was not yet true. The current paper is the culmination of an initial idea from the beginning of April [1], a presentation of the idea in mid-April [2], discussions with policymakers, and an initial writeup [3] which was discussed seriously, albeit privately, as a policy approach. We have now cited the writeup in the revision of the paper. The peer reviewed publication process is unfortunately far slower than the policy process, and the current manuscript has been a victim of that delay. We will note that the developed vaccines are still being manufactured far more slowly than would be ideal. An investment of several billion dollars several months ago, earlier during the trials, would have significantly alleviated the current lack of vaccine production capacity. For that reason, it still seems very relevant to the policy discussion to point out that mechanisms to allow this were being proposed in time for them to have been used. Moreover, it is clearly relevant as a possible approach for future crises. It would be a shame if promising ideas once again only gained traction in academic and policy circles when it was too late to avert catastrophic outcomes. 2.The authors argue for an option-based approach to help manufacturers in investing in new vaccines by de-risking eventual failures with public money. This approach might be relevant to incentivize development efforts in areas of high public health and medical need. We certainly agree that this approach is viable in some other contexts, and agree that the paper is useful for that context. As we discussed in the paper, the different approaches are relevant for different needs. We are skeptical, however, that the approach we suggest for COVID-19 is applicable widely outside of the urgent development of a new and promising technology to meet a widely recognized need. We can certainly envision scenarios where the mechanism would be viable, but because the primary advantage is de-risking capital investment, rather than funding development, it seems unlikely to spur great interest unless and until there is a new need for drastic action, or a niche use case can be found to pioneer the method on a smaller scale via a philanthropic investment. 3. Overall, the manuscript is well written and brings interesting insights on mechanisms that will likely be at the center of the public attention in the upcoming post-discovery phase of COVID-19 vaccines. However, its content seems to be completely detached from present day issues and it is very theoretical. It is not clear how the proposed solution integrates with - or differentiates from - the dozens of private and public initiatives that support COVID-19 vaccine development. Are there already option-based approaches being used for COVID-19 vaccines? If not, why? Are governments or public entities considering such approaches for some of the vaccines under development? We appreciate the feedback. The paper presents a mechanism, and requires policy-decision making and negotiation if it were used in the future to integrate with other mechanisms. The mechanisms which were used seem to have spurred sufficient investment, though it is unclear how much of the credit for this should go to the financing mechanisms. While we will not discuss the history and context extensively in the paper, we will make a few observations here in the reply. As the paper discusses, the proposal is novel, and while we think there was promise, we are unaware of any earlier suggestions or current plans for using the novel mechanism. This seems to be due to a combination of factors. First, given the prominent role that the US played in financing vaccines, and the political turmoil of the Trump administration, there was reticence on the part of policymakers to try anything which might draw attention to the potential risks. Second, once funding was made available widely for vaccine development, efficiency of the mechanism was relegated to a minor role. Given this, we note that despite thinking our proposed mechanism is promising for any future pandemics, over-supply of funding was potentially a better choice, and is certainly simpler. Almost any plausible level of overspending on the vaccines to spur faster development and production would have been worthwhile, post-hoc. At the same time, investment in production of vaccines was not as timely or as large as could have been achieved with this model. Soon after this model was first proposed, in April, the Gates Foundation announced it would fund production of the 7 leading vaccine candidates, and it did so by funding GAVI to enter into advance market commitments. The EU similarly used AMCs, which were paired with funding for production. Unfortunately, as we discussed in the manuscript, the funds are only useful for providing capital, in these cases, to already-large global firms with existing access to global capital markets. Because the purchases are contingent on approval, they do not decrease risk, and so production has been slow. This makes sense for firms; if the purchase agreement is early, the company bears the risk of it not being approved - as seems to have occurred recently in Australia with the UQ/CSL vaccine, where the advance purchase was cancelled. Moreover, because AMCs are only used after it seems clear a vaccine has a high chance of success, these were not entered into until after June. The stated rationale was that “this Strategy is therefore similar to an insurance policy, by transferring some of the risks from industry to public authorities in return for assuring Member States of equitable and affordable access to a vaccine.” The proposal put forward here is a more straightforward way to accomplish the first half. The ethics of ensuring supply to some countries at the expense of others aside, as we noted, this does not preclude also using purchase commitments to guarantee supply. Not only this, but provision of funds was slow. Despite early promises, it took until August for the Gates Foundation to give $150m to increase production in India. Minor: the authors should update the introduction session as - as of today - two vaccine candidates have already demonstrated high efficacy in interim ph3 analysis without the need of human challenge trials and with much stringent timelines than reported in this manuscript. This is now edited in the revision to reflect the timing of the publication. Citations: 1. Manheim, D. (2020, April 2). A Simple Proposal for Jumpstarting Vaccine Production. In about 12 months, the world will need to start producing massive quantities... [Tweet]. Twitter. https://twitter.com/davidmanheim/status/1245810066843983872 2. Manheim, D. (2020, April 13). A Proposal to Accelerate Vaccine Production Now [Video]. YouTube. https://www.youtube.com/watch?v=mVjqGh_Dmv8 3. Foster, D., & Manheim, D. (2020, April 27). Market-shaping approaches to accelerate COVID-19 response: A role for option-based guarantees? LessWrong Forum. https://www.lesswrong.com/posts/uXb4gcDP2fgBPcMHJ/market-shaping-approaches-to-accelerate-covid-19-response-a"
}
]
}
] | 1
|
https://f1000research.com/articles/9-1154
|
https://f1000research.com/articles/9-1304/v1
|
06 Nov 20
|
{
"type": "Research Article",
"title": "Psychological burden of the COVID-19 pandemic and its associated factors among frontline doctors of Bangladesh: a cross-sectional study",
"authors": [
"Lingkan Barua",
"Muhammed Shahriar Zaman",
"Fardina Rahman Omi",
"Mithila Faruque",
"Muhammed Shahriar Zaman",
"Fardina Rahman Omi",
"Mithila Faruque"
],
"abstract": "Background: Frontline doctors are the most vulnerable and high-risk population to get the novel coronavirus disease 2019 (COVID-19) infection. Hence, we aimed to evaluate the anxiety, depression, sleep disturbance and fear of COVID-19 among frontline doctors of Bangladesh during the pandemic, and the associated factors for these psychological symptoms. Methods: In total, 370 frontline doctors who were involved in the treatment of suspected or confirmed COVID-19 patients during the pandemic took part in an online cross-sectional study. Recruitment was completed using convenience sampling and the data were collected after the start of community transmission of COVID-19 in the country. Anxiety and depression, sleep disturbance, and fear of COVID-19 were assessed by the Patient Health Questionnaire-4, two-item version of the Sleep Condition Indicator, and the Fear of Coronavirus-19 scale, respectively. Socio-demographic information, health service-related information, co-morbidity, and smoking history were collected for evaluating risk factors. The proportion of psychological symptoms were presented using descriptive statistics and the associated factors were identified using multinomial logistic regression analysis. Results: Of the doctors, 36.5% had anxiety, 38.4% had depression, 18.6% had insomnia, and 31.9% had fear of COVID-19. In multinomial logistic regression, inadequate resources in the workplace were found as the single most significant predictor for all psychological outcomes: anxiety and/or depression (severe, OR 3.0, p=0.01; moderate, OR 5.3, p=0.000; mild, OR 2.3, p=0.003), sleep disturbance (moderate, OR 1.9, p=0.02), and fear of COVID-19 (severe, OR 1.9, p=0.03; moderate, OR 1.8, p=0.03). Conclusions: The study demonstrated a high burden of psychological symptoms among frontline doctors of Bangladesh during the COVID-19 pandemic situation. Inadequate resources are contributing to the poor mental health of Bangladeshi doctors. The supply of sufficient resources in workplaces and mental health counseling may help to mitigate the burden of the psychological symptoms identified among the respondents..",
"keywords": [
"COVID-19",
"mental health",
"doctors",
"risk factors",
"Bangladesh"
],
"content": "Introduction\n\nNovel coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), was first recognized in December 2019 in Wuhan City in central China1,2. The World Health Organization declared the COVID-19 outbreak as a global pandemic on March 11, 20203. Bangladesh confirmed its first COVID-19 outbreak on March 08, 2020, when the Institute of Epidemiology, Disease Control and Research (IEDCR) reported the first three confirmed cases4. As of July 31, 2020, IEDCR confirmed 234,889 COVID-19 cases in Bangladesh, including 3083 related deaths with a Case Fatality Rate of 1.31%5.\n\nThe COVID-19 pandemic has caused various challenges in Bangladesh's healthcare system. One of the biggest challenges is the spread of COVID-19 infections among frontline doctors6. Up to July 29, 2020, about 2453 doctors have been infected7, and 69 doctors have died8 because of COVID-19 infection in Bangladesh. The mortality rate due to COVID-19 among doctors in Bangladesh is about 4%, which is the highest in the world among doctors9, and this rate is also higher than that of Bangladesh's national mortality rate for COVID-195.\n\nIn addition to the surge of COVID-19 infection, the pandemic has caused mental health problems to rise among doctors in Bangladesh. Mental health problems during pandemics are common, and evidence has shown that the severe acute respiratory syndrome (SARS), Middle East respiratory syndrome (MERS), and H1N1 pandemics also impacted the mental health condition of healthcare workers10–12. A study showed that frontline healthcare workers feel tremendous mental pressure during a pandemic because of the diminution of personal protection equipment, extensive media reportage, lack of treatment resources, increasing pattern of cases, death tolls, tremendous workload and social stigmatization13. Recently, articles from Singapore, India, Greece and China have reported mental health issues of healthcare workers during the current rapidly evolving situation14–17.\n\nBangladesh is a lower-middle-income country where doctors have to provide services in an overburdened, understaffed, and insufficiently equipped setting due to massive shortage and disproportionate distribution of skilled health workers, which causes unusual mental stress18. Despite the challenges in their workplaces, during the COVID-19 pandemic Bangladeshi doctors have shown their competency and professionalism in providing the best care to the country's people. As part of their responsibility, they have to expose themselves to the risk of COVID-19 infection for the benefit of the mass population19. It is speculated that the risk of infection and professional stress has gradually worsened the mental health condition of doctors in Bangladesh as they are facing stigmatization, fear of spreading the infection to family members and fear of being isolated. Currently, there is no evidence in support of this assumption. Therefore, we conducted this study to evaluate the psychological burden among Bangladeshi frontline doctors during the COVID-19 pandemic. To assess psychological symptoms, we quantified the magnitude of anxiety, depression, sleep disturbance and fear of COVID-19. Besides, we explored the associated factors influencing the psychological outcome. The findings of the study could be used to identify potential gaps in practice that would need interventions.\n\n\nMethods\n\nWe conducted an online cross-sectional study among doctors working at different clinical settings to treat patients, either suspected or confirmed COVID-19 cases, during the pandemic. Participants’ recruitment was completed by convenience sampling. Doctors from the professional and personal networks of the researchers were initially contacted through Facebook messenger and email. Doctors who showed interest were invited to participate in the study through an online questionnaire. We excluded doctors who did not complete intern training after graduation or were not involved in direct patient care.\n\nA total of 370 frontline doctors took part in the study over two months from 1st April to 30th May 2020. Sample size was determined using the prevalence of anxiety, depression, insomnia, and distress among China's healthcare workers during the COVID-19 pandemic13. The highest sample number was taken using the prevalence of depression (31.8%) in the aforementioned study, i.e. 370 respondents.\n\nWe circulated the questionnaire through online among the interested participants after the Government of Bangladesh confirmed community transmission in Bangladesh on March 28, 202020. Data were collected by an online self-administered semi-structured questionnaire using the Google survey platform21. The questionnaire link was sent to participants electronically through Facebook and email.\n\nQuestionnaire content. The online questionnaire collected data on sociodemographic factors (age, gender, marital status, education, occupation), health service-related factors (the type of service, working place, professional designation, service level of health system, number of days of service provided, shifting duty or not, resource of working place), psychological parameters (anxiety, depression, sleep disturbance, fear), co-morbid conditions (diabetes, hypertension, asthma, chronic obstructive pulmonary disease, heart disease, chronic kidney disease, thyroid disorder), high-risk behavior as defined by tobacco use, and the living area of the physician where at least one COVID-19 case had been confirmed by the local authority.\n\nThe questionnaire was pre-tested before the final administration to detect any inconsistency and biases. To pre-test, 10 men and 10 women frontline doctors were selected randomly using the inclusion criteria (MBBS degree with completed intern training) and the questionnaire was sent to them through an online platform (Facebook messenger and email). The objective and importance of pre-testing were added with the questionnaire as an explanatory note. The researchers also informed that participation of the respondents was voluntary and they have the right to withdraw themselves at any time or refuse to answer any question. The collected responses were analyzed and interpreted based on the following: trends in responses; fundamental flaws with the design or format; attractiveness; comprehension; acceptance; and relevance.\n\nInstruments used to assess psychological symptoms. Anxious and depressive symptoms were assessed via the Patient Health Questionnaire-4 (PHQ-4)22, which was an ultra-brief self-report questionnaire with a 2-item anxiety scale, named Generalized Anxiety Disorder 2-item (GAD-2), and a 2-item depression scale, named Patient Health Questionnaire 2-item (PHQ-2). The total score was determined by adding the scores of each of the four items as 0, 1, 2, and 3. Scores were rated as normal (0-2), mild (3-5), moderate (6-8), and severe (9-12). Total score ≥3 for the first two questions suggested anxiety. Total score ≥3 for the last two questions suggested depression22.\n\nSleep disturbance was assessed via a two-item version of the Sleep Condition Indicator (SCI-02), an ultra-short clinical rating scale, which can be used to rapidly screen for insomnia in routine clinical practice23. Each item was scored on a 5-point scale as 0, 1, 2, 3, 4. By adding the item scores, the SCI total score was obtained, ranging from 0 to 8. A higher score means better sleep. To quantify the magnitude of severity, we categorized the sleep disturbance using percentiles of the SCI-02 score as follows: good sleep condition (score ≥7), moderate sleep condition (score 3-6) and insomnia (score 0-2). Here, the cut-off value of insomnia was kept the same as DSM-5 threshold criteria24.\n\nThe Fear of Coronavirus-19 Scale (FCV-19S) was used to measure one’s fear of COVID-1925. The FCV-19S consists of 7 items. Participants were asked to rate their agreement with each statement on a 5-point scale from ‘1 - strongly disagree’ to ‘5 - strongly agree'. A higher score indicated greater fear. Recently, this instrument was validated among the Bangladeshi population26. Currently, the FCV-19S has no classification of severity, and hence, we developed a severity scale using percentiles of FCV-19S score as follows: mild (score ≤17), moderate (score 18 to 23) and severe (score ≥24).\n\nThe data were entered in a pre-designed Microsoft Office Excel format, which was imported later into the software Statistical Package for Social Science version 20.0 for Windows (SPSS, Inc. Chicago. IL.USA). All the estimates of precision were presented at a 95% confidence interval (CI). Descriptive analysis included mean, standard deviation (SD), frequencies, and percentages. Background information (sociodemographic and professional) and the magnitude of psychological outcomes were presented using frequencies and percentages. The score of the instruments was presented using the mean with SD.\n\nThe associated factors of psychological outcomes were determined using multinomial logistic regression analysis. To find the factors that influenced the psychological outcomes, first, we run univariate analysis. Variables that showed p ≤0.25 in the univariate analysis were examined as an independent variable in the logistic regression27,28. We calculated odds ratios (OR) and 95% CI for each independent variable for multiple logistic regression analysis. In the regression table, factors that had OR >1 were presented for each outcome variable. We ensured no multicollinearity presence using the variance inflation factor (VIF) to run the regression analysis. The statistical tests were considered significant (2-sided) at a level of p ≤0.05.\n\nThe Ethical Review Committee of Bangladesh University of Health Sciences approved the study (identification number: BUHS/ERC/20/16).\n\nAn information and consent form (Extended data21) to take part in the study and for the publication of the participant’s anonymized information was provided prior to the questionnaire. Completion of the questionnaire implied consent.\n\n\nResults\n\nAs shown in Figure 1, 1000 individuals were contacted initially and 370 were included in the study after exclusion.\n\nThe mean (SD) age of the doctors was 30.5 (4.4) years. Most of them were men (60.3%) and married (66.8%). A total of 69.5% had been living in areas that were affected by the COVID-19 outbreak. About a quarter of participants (24.8%) had been suffering from at least one chronic disease; the proportion of more-than-one chronic diseases was 4.3%. The most commonly reported chronic disease was chronic bronchial asthma (15.9%). Table 1 presents the detailed demographic and health-related characteristics of the study participants.\n\nCOVID-19, coronavirus 2019\n\nMore than half of the total doctors (56.5%) had a Bachelor's (MBBS) degree, which is the entry-level degree for medical doctors in Bangladesh, and 19.7% had post-graduation degrees. The rest were post-graduate students (23.8%). The majority was employed in the private sector (55.4%), followed by the government sector (30.3%). Most of the doctors' primary working settings were a hospital (54.3%), and most of them worked at tertiary level healthcare settings (32.2%). The majority of the doctors had shifting duties (69.5%) and worked in a low resource setting (70.5%). On average, they worked five days a week during the pandemic (Table 2).\n\n*Representing mean and standard deviation. MBBS, Bachelor of Medicine and Bachelor of Surgery; NGO, non-government organization; DG, directorate general; CI, confidence interval.\n\nThe detailed result of psychological status is presented in Table 3. The mean (SD) score of PHQ4, GAD-2 score and PHQ-2 score were 4.5 (2.9), 2.3 (1.8) and 2.2 (1.6), respectively. Considering the total score of PHQ4, about 73% of doctors had anxiety and/or depression, of which the majority were affected by mild anxiety and/or depression (39.2%). Separately, the first two (GAD-2) and successive two (PHQ-2) items of PHQ-4 identified that 36.5% of the doctors had anxiety, and 38.4% had depression. Moreover, in the SCI-2 score, 18.6% of the doctors were found to be insomniac. Furthermore, the FCV-19S identified that 31.9% and 37.6% of the physicians had a severe and moderate level of fear regarding the COVID-19 pandemic, respectively.\n\n*Representing mean and standard deviation. PHQ, Patient Health Questionnaire; GAD-2, Generalized Anxiety Disorder 2-item; SCI, Sleep Condition Indicator; COVID-19, coronavirus 2019; FCV-19S, fear of coronavirus 2019 scale; CI, confidence interval\n\nThe univariate analysis (Chi-square test) showed association between PHQ4 (anxiety and/or depression) categories and several factors including gender (p=0.03), inadequate resources (p<0.001), presence of chronic disease (p=0.001), number of chronic diseases (p=0.003), asthma (p=0.002), and hypertension (0.005) (Table 4). However, in the multinomial regression model, only inadequate resources in a working setting was found to be a significant predictor for severe (OR:2.99, 95% CI: 1.25- 7.15, p=0.01), moderate (OR:5.30, 95% CI: 2.54- 11.09, p<0.001), and mild (OR:2.28, 95% CI: 1.33-3.92, p=0.003) anxiety and/or depression controlling gender, presence of chronic disease, number of chronic diseases, asthma, and hypertension (Table 5).\n\nPHQ, Patient Health Questionnaire; SCI, Sleep Condition Indicator; COVID-19, coronavirus 2019; FCV-19S, fear of coronavirus 2019 scale. p-value significant at the threshold of ≤0.05\n\nPHQ, Patient Health Questionnaire; SCI, Sleep Condition Indicator; COVID-19, coronavirus 2019; FCV-19S, fear of coronavirus 2019 scale; Ref., reference. p-value significant at the threshold of ≤0.05\n\nRegarding sleep disturbance, the univariate analysis found age (p=0.001), working area (p=0.01), shifting duty (p=0.04), inadequate resources (p=0.05), residence in a COVID-19 affected area (p=0.004), number of chronic diseases (p=0.01), and asthma (p=0.05) as the associated factors (Table 4). Among the associated factors, only asthma was found as a significant predictor of insomnia (OR: 4.06, 95% CI: 1.57-10.51, p=0.004) and moderate sleep condition (OR: 3.33, 95% CI: 1.47-7.54, p=0.004) controlling all other associated factors in a regression model. In addition, shifting duty (OR: 2.21, 95% CI: 1.24-3.94, p=0.007), inadequate resources (OR: 1.85, 95% CI: 1.08-3.16, p=0.02), and living in a COVID-19 affected area (OR: 2.38, 95% CI: 1.41-4.01, p=0.001) were also found as significant predictors for moderate sleep condition (Table 5).\n\nRegarding fear of COID-19, the univariate analysis found gender (p<0.001), primary working area (p=0.002), and inadequate resources (p=0.03) as associated factors (Table 4). However, in multinomial regression analysis, only inadequate resources was found as the significant predictor for severe (OR: 1.90, 95% CI: 1.05-3.47, p=0.03) and moderate (OR: 1.82, 95% CI: 1.05-3.16, p=0.03) fear of COVID-19 (Table 5).\n\n\nDiscussion\n\nThe study aimed to assess the psychological burden of frontline doctors in Bangladesh during the COVID-19 pandemic, and factors that predict their psychological status. The study identified that anxiety, depression, insomnia, and fear related to the COVID-19 outbreak are common among frontline doctors of Bangladesh during this unprecedented time. The paucity of resources for providing care to patients in workplaces was found as the single most common predictor for poor psychological status. In addition, having shifting duty, living in a COVID-19 affected area, and the presence of asthma predicted poor quality of sleep among the frontline doctors.\n\nA considerable proportion of frontline doctors in Bangladesh has experienced psychological symptoms due to the COVID-19 pandemic. The burden of psychological symptoms is higher than the burden of symptoms among healthcare workers of China, Singapore and India during the COVID-19 pandemic14,15,17. A meta-analysis study from China has presented the pooled prevalence of depression (22.8%), anxiety (23.2%), and insomnia (38.9%)17. Compared to the pooled prevalence of symptoms in China, the current study has shown a higher proportion of depression and anxiety, but a lower proportion of insomnia among Bangladeshi doctors. Furthermore, the prevalence of anxiety and depression were reported as 14.4% and 9%, respectively, in Singapore15 and 17.1% and 12.4%, respectively, in India14, which are also lower than the magnitude of anxiety and depression observed among Bangladeshi doctors in this study. The burden of psychological symptoms in the current study is also higher than the burden of psychological symptoms among China's general population during the pandemic29. Moreover, a comparison with mental health symptoms (anxiety 77.4%, depression 74.2%, and sleep problems 52.3%) among health workers during SARS pandemics in Taiwan shows a lower burden of psychological symptoms in Bangladesh during COVID-19 pandemics10. It is noteworthy that there are variations in the methods of measuring psychological symptoms across the studies.\n\nMany underlying factors for mental health problems among frontline health workers during the pandemic situation have been reported in the literature, including gender, age, living in a rural area, poor social support, poor self-efficacy, profession, place of work, disruption of routine clinical practice, fear of potential destabilization of health services, the sense of loss of control, having organic disease, and being at risk of contact with a patient with COVID-1913,16,29–32 Among all the reported causes, COVID-19 can be an independent risk factor for healthcare workers' poor mental health30. In Bangladesh, the burden of COVID-19 is among the top 20 countries in the world. Along with the general population, frontline health workers have also been overwhelmed by the surge of infection. It has been reported that doctors in Bangladesh have been experiencing the highest infection and mortality in the world due to the virus9. Experts have suggested that lack of infection control measures, monitoring, proper management at hospitals, inappropriate use and disposal of safety gear, and lack of training for dealing with patients with COVID-19 are contributing to the highest infection and mortality of the doctors9. It is also believed that COVID-19 infection and its underlying causes contribute to the doctors' poor mental health condition.\n\nThe current study identified several factors that contribute to the burden of psychological symptoms among Bangladeshi doctors - the paucity of resources in the workplace is the most significant. Limited resources in the workplace include materials, trained workforce or any other things that are required to provide services. The current study has confirmed the association of inadequate resources with the poor psychological status of doctors. Inadequate resources such as masks, sanitizer, and personal protective equipment (PPE) in workplaces increase the chance of getting COVID-19 infection and can cause profound psychological pressure on frontline doctors. The lack of resources in workplaces in Bangladesh has been widely reported in news media33. The news media has reported inadequate and inappropriate PPE as a cause of widespread COVID-19 infection among healthcare professionals in Bangladesh33. However, Bangladesh is not the only country that faced a shortage of resources during the pandemic. The shortage of such resources has also been reported in many other countries because of the distorted supply chain across the world34. Lack of resources is also considered as a cause of poor psychological status among healthcare workers in many countries during the pandemic16,35. Experts have recognized sufficient resources as an essential factor for healthcare professionals to be resilient during an unprecedented time36.\n\nThe lack of skilled and trained workforce in hospitals is another underlying cause of the high burden of psychological symptoms among frontline doctors in Bangladesh. Amid the workforce shortage, frontline doctors have to do long shifting duties for a certain period and then stay in quarantine for 14 days before they return to work. This atypical work schedule for doctors has been introduced to reduce the frequency of exposure to COVID-19 virus in workplaces. However, it is believed that the long shifting duties and being isolated during quarantine may have triggered mental health problems among doctors. The current study has found that those who did shifting duties were more likely to have sleep problems linked with poor mental health. Although identifying links between the quarantine period and poor mental health was not a scope of the current study, other studies have confirmed the link between quarantine period and mental health during this pandemic37.\n\nThe current study is the first study in Bangladesh that provides the burden and associated factors for doctors' poor mental health outcome during the COVID-19 pandemic. There are some limitations in the study. As it is a cross-sectional study, causal relation could not be established. Thus, the study presents the factors linked with the psychological outcomes as associated factors. Moreover, the study is an online-based questionnaire. Therefore, the possibility of selection bias cannot be ruled out. Again, a small sample size limited the generalization of the study findings. The participants of the study were mainly young doctors. This happened because the younger population is more exposed to online platforms than the elderly. However, a recent review has shown that younger doctors are more affected by psychological symptoms than elder doctors31. Thus, the study has reflected evidence of the high-risk group of doctors for a psychological problem.\n\n\nConclusions\n\nA high burden of COVID-19 related anxiety, depression, sleep disturbance, and fear among Bangladeshi frontline doctors demands policymakers' immediate attention to take appropriate preventive measures. An appropriate risk-reduction strategy should be developed and implemented to reduce the risk of getting COVID-19 infection. In addition, the supply of adequate PPE and the development of a trained workforce with infection control skills need to be considered to reduce the psychological impact. The substantial burden of different mental health outcomes elucidated in the current study demands mental health counsellors in hospital settings where appropriate. Considering low resource settings, this strategy could be implemented at least in COVID-19 dedicated hospitals in Bangladesh.\n\n\nData availability\n\nZenodo: The psychological burden of the COVID-19 pandemic and its associated factors among the frontline doctors of Bangladesh: A cross-sectional study, http://doi.org/10.5281/zenodo.411033738.\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).\n\nZenodo: The Psychological Burden of the COVID-19 Pandemic and Its Associated Factors among the Frontline Doctors of Bangladesh: A Cross-sectional Study-Extended Data, https://doi.org/10.5281/zenodo.405871521.\n\nThis project contains the following extended data within the file ‘Extended data file.pdf’:\n\n— Consent form (English Version)\n\n— Questionnaire (English Version)\n\nZenodo: STROBE checklist for ‘The Psychological Burden of the COVID-19 Pandemic and Its Associated Factors among the Frontline Doctors of Bangladesh: A Cross-sectional Study’, https://doi.org/10.5281/zenodo.406217039.\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).",
"appendix": "Acknowledgments\n\nThe authors would like to acknowledge the participants who gave their valuable time to develop this evidence for the doctor's community of Bangladesh. We would also like to acknowledge the frontline doctors of Bangladesh who died from COVID-19 to save others' lives during this pandemic.\n\n\nReferences\n\nWorld Health Organization: Coronavirus. [accessed March 06, 2020]. Reference Source\n\nCDC: Coronavirus Disease 2019 (COVID-19). Centers for Disease Control and Prevention 2020; [accessed April 07, 2020]. Reference Source\n\nWorld Health Organization: Coronavirus disease (COVID-19) pandemic. [accessed April 09, 2020]. Reference Source\n\nPaul R: Bangladesh confirms its first three cases of coronavirus. Reuters. 2020; (accessed April 09, 2020). Reference Source\n\nInstitute of Epidemiology, Disease Control and Research: Bangladesh Covid-19 Update. [accessed July 31, 2020]. Reference Source\n\nAnwar S, Nasrullah M, Hosen MJ: COVID-19 and Bangladesh: Challenges and How to Address Them. Front Public Health. 2020; 8: 154. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBangladesh Medical Association: District Wise Total Number of Affected Doctor, Nurse & Staff (COVID 19+) 29/7/2020. [accessed August 01, 2020].\n\nBangladesh Medical Association: List of doctors died because of COVID-19. [accessed August 01, 2020]. Reference Source\n\nUNB: COVID-19: Physicians’ mortality rate in Bangladesh ‘highest in the world.’ Prothomalo. 2020; [accessed June 25, 2020]. Reference Source\n\nChong MY, Wang WC, Hsieh WC, et al.: Psychological impact of severe acute respiratory syndrome on health workers in a tertiary hospital. Br J Psychiatry. 2004; 185: 127–33. PubMed Abstract | Publisher Full Text\n\nGoulia P, Mantas C, Dimitroula D, et al.: General hospital staff worries, perceived sufficiency of information and associated psychological distress during the A/H1N1 influenza pandemic. BMC Infect Dis. 2010; 10: 322. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee SM, Kang WS, Cho AR, et al.: Psychological impact of the 2015 MERS outbreak on hospital workers and quarantined hemodialysis patients. Compr Psychiatry. 2018; 87: 123–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLai J, Ma S, Wang Y, et al.: Factors Associated With Mental Health Outcomes Among Health Care Workers Exposed to Coronavirus Disease 2019. JAMA Netw Open. 2020; 3(3): e203976. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChew NWS, Lee GKH, Tan BYQ, et al.: A multinational, multicentre study on the psychological outcomes and associated physical symptoms amongst healthcare workers during COVID-19 outbreak. Brain Behav Immun. 2020; 88: 559–565. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTan BYQ, Chew NWS, Lee GKH, et al.: Psychological impact of the COVID-19 pandemic on health Care Workers in Singapore. Ann Intern Med. 2020; 173(4): 317–320. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTsamakis K, Rizos E, Manolis AJ, et al.: COVID-19 pandemic and its impact on mental health of healthcare professionals. Exp Ther Med. 2020; 19(6): 3451–3453. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPappa S, Ntella V, Giannakas T, et al.: Prevalence of depression, anxiety, and insomnia among healthcare workers during the COVID-19 pandemic: A systematic review and meta-analysis. Brain Behav Immun. 2020; 88: 901–907. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAhmed SM, Alam BB, Anwar I, et al.: Bangladesh health system review. Manila: WHO Regional Office for the Western Pacific; 2015. Reference Source\n\nMalm H, May T, Francis LP, et al.: Ethics, pandemics, and the duty to treat. Am J Bioeth. 2008; 8(8): 4–19. PubMed Abstract | Publisher Full Text\n\nDirectorate General of Health Services: National Guidelines on Clinical Management of Coronavirus Disease 2019 (Covid-19) Version 4.0. 2020. Reference Source\n\nBarua L, Zaman MS, Omi FR, et al.: The Psychological Burden of the COVID-19 Pandemic and Its Associated Factors among the Frontline Doctors of Bangladesh: A Cross-sectional Study-Extended Data (Version 02) [Data set]. Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4058715\n\nLöwe B, Wahl I, Rose M, et al.: A 4-item measure of depression and anxiety: validation and standardization of the Patient Health Questionnaire-4 (PHQ-4) in the general population. J Affect Disord. 2010; 122(1–2): 86–95. PubMed Abstract | Publisher Full Text\n\nLuik AI, Machado PF, Siriwardena N, et al.: Screening for insomnia in primary care: using a two-item version of the Sleep Condition Indicator. Br J Gen Pract. 2019; 69(679): 79–80. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAmerican Psychiatric Association: Diagnostic and statistical manual of mental disorders: DSM-5. 5th ed. Arlington, Va.: American Psychiatric Association; 2013. Publisher Full Text\n\nAhorsu DK, Lin CY, Imani V, et al.: The Fear of COVID-19 Scale: Development and Initial Validation. Int J Ment Health Addict. 2020; 1–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSakib N, Bhuiyan AKMI, Hossain S, et al.: Psychometric Validation of the Bangla Fear of COVID-19 Scale: Confirmatory Factor Analysis and Rasch Analysis. Int J Ment Health Addict. 2020; 1–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBendel RB, Afifi AA: Comparison of Stopping Rules in Forward “Stepwise” Regression. J Am Stat Assoc. 1977; 72(357): 46–53. Publisher Full Text\n\nCostanza MC, Afifi AA: Comparison of Stopping Rules in Forward Stepwise Discriminant Analysis. J Am Stat Assoc. 1979; 74(368): 777–85. Publisher Full Text\n\nWang C, Pan R, Wan X, et al.: Immediate Psychological Responses and Associated Factors during the Initial Stage of the 2019 Coronavirus Disease (COVID-19) Epidemic among the General Population in China. Int J Environ Res Public Health. 2020; 17(5): 1729. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHall H: The effect of the COVID-19 pandemic on healthcare workers’ mental health. JAAPA. 2020; 33(7): 45–8. PubMed Abstract | Publisher Full Text\n\nSpoorthy MS, Pratapa SK, Mahant S: Mental health problems faced by healthcare workers due to the COVID-19 pandemic-A review. Asian J Psychiatr. 2020; 51: 102119. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang WR, Wang K, Yin L, et al.: Mental Health and Psychosocial Problems of Medical Health Workers during the COVID-19 Epidemic in China. Psychother Psychosom. 2020; 89(4): 242–250. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMahmud F: Hundreds of doctors in Bangladesh infected with coronavirus. WwwAljazeeraCom, 2020. [accessed April 25, 2020]. Reference Source\n\nBegg R: Global Medical Equipment Shortage as COVID-19 Spreads. WwwMachinedesignCom 2020. [accessed April 25, 2020]. Reference Source\n\nKar SK, Yasir Arafat SM, Kabir R, et al.: Coping with Mental Health Challenges During COVID-19. Medical Virology: From Pathogenesis to Disease Control. 2020; 2020: 199–213. Publisher Full Text | Free Full Text\n\nKalaitzaki AE, Tamiolaki A, Rovithis M: The healthcare professionals amidst COVID-19 pandemic: A perspective of resilience and posttraumatic growth. Asian J Psychiatr. 2020; 52: 102172. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrooks SK, Webster RK, Smith LE, et al.: The psychological impact of quarantine and how to reduce it: rapid review of the evidence. Lancet. 2020; (10227): 912–920. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarua L, Zaman MS, Omi FR, et al.: The psychological burden of the COVID-19 pandemic and its associated factors among the frontline doctors of Bangladesh: A cross-sectional study (Version 1) [Data set]. Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4110337\n\nBarua L, Zaman MS, Omi FR, et al.: The Psychological Burden of the COVID-19 Pandemic and Its Associated Factors among the Frontline Doctors of Bangladesh: A Cross-sectional Study-STROBE checklist (Version 01). Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4062171"
}
|
[
{
"id": "74585",
"date": "24 Nov 2020",
"name": "Mohammed Mamun",
"expertise": [
"Reviewer Expertise Psychiatric Epidemiology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI would like to thank the authors for assessing such a cohort (i.e., Frontline fighters), as there is almost no study in Bangladesh to the cohort. Their studied mental health outcomes such as depression, anxiety, insomnia and fear of COIVD is appreciable. However, a few observations are provided herein, which may be considered.\nInstead of frontline doctors, frontline fighters may be used – because this word is widely used in the press media, and your article may get more visibility by using it. My comment is not mandatory as you only collected data from doctors, whereas fighters includes nurse, paramedics etc.\n\nafter the start of community transmission of COVID-19 in the country – may I suggest to add the exact time frame?\n\nRecently, articles from Singapore, India, Greece and China – please replace the word ‘articles’ with ‘studies’.\n\nThe authors precisely reported the relevance of the study (i.e., evaluating mental health suffering in doctors). However, I would like to suggest referring a case study reporting excessive fear of COVID-19 existing in the healthcare facilities in Bangladesh. Consequently, a non- COVID-19 patient committed suicide. I think this information can add an extra value to your study rationale. Ref: https://doi.org/10.1016/j.ajp.2020.102295\n\nBesides, there is a Bangladeshi case-control large scale study accessing suicidality can be discussed and compare what is new in your study. For example, the study got not statistical significance difference in suicidality across general people and healthcare professionals. Ref: https://doi.org/10.1016/j.heliyon.2020.e05259\n\nMore information regarding sampling is needed. Please add scale reliability (e.g., Cronbach's alpha) as the scales are not validated in Bangla.\n\nThe authors already mentioned, there is no cutoff score for Fear of COVID scale. However, they categorized it. It would be better if they provide how the category was considered – median?\n\nIt is requested to report the mean and SD of fear of COVID (along with percentage; as it was done for depression and anxiety) as the scale has no cutoff score yet.\n\nIs not it better to report psychological burdens as “probable depression” instead of “depression”, because the scales are short version (although their cutoff scores had higher specificity and sensitivity than the original one; for example, PHQ-2 and PHQ-9)? Similar suggestion for fear of COVID. Besides, the authors may like to perform ANOVA and linear regression – not mandatory if they refer how the cutoff scores were categorized.\n\nDiscussion is focused on the results. Discussing prevalence rate across countries is perfect, even the authors compared with the review article prevalence rate, which is appreciated. By the way, may I suggest to compare the Bangladeshi general people mental health problems rates, this may help the reader to be informed about how this study findings differ from the general people. Please refer Bangladeshi studies assessing the pandemic related psychological burdens (ref: https://doi.org/10.31234/osf.io/q4k5b).\n\nThe current study is the first study in Bangladesh that provides the burden and associated factors for doctors’ poor mental health outcome during the COVID-19 pandemic. – I request to avoid the first word, because there is a published paper on suicidal behavior of the cohort.\nOverall observations, the authors did a great job addressing mental health problems of the vulnerable cohort to virus infection, which may have influence in policy level. And the paper was written in a good flow, and also provided some of recommendations. Best of luck.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6178",
"date": "16 Dec 2020",
"name": "Lingkan Barua",
"role": "Author Response",
"response": "Response to the Reviewer1 We would like to thank you for your overall comment on our submitted manuscript. Here we have added our responses. The changes have presented using track changes. (A) Initial Comments I would like to thank the authors for assessing such a cohort (i.e., Frontline fighters), as there is almost no study in Bangladesh to the cohort. Their studied mental health outcomes such as depression, anxiety, insomnia and fear of COIVD is appreciable. However, a few observations are provided herein, which may be considered. Comment 1: Instead of frontline doctors, frontline fighters may be used – because this word is widely used in the press media, and your article may get more visibility by using it. My comment is not mandatory as you only collected data from doctors, whereas fighters includes nurse, paramedics etc. Response 1: Thanks for your valuable comment about our target population. To present the target population more precisely and specifically, the term ‘Frontline doctors’ is best suited and also aligned with the 2nd sentence of the reviewer comment (“whereas fighters include nurse, paramedics etc”). Comment 2: After the start of community transmission of COVID-19 in the country – may I suggest to add the exact time frame? Response 2: Thanks for your valuable comment. The time frame has already mentioned in the two separate sections of the ‘Methods’ part. In the ‘Study design and participants’ section, paragraph 02, we mentioned-“A total of 370 frontline doctors took part in the study over two months from 1st April to 30th May 2020”. Again, in the ‘Data collection’ section, 1st paragraph, and in the 1st line we mentioned: “the Government of Bangladesh confirmed community transmission in Bangladesh on March 28, 2020”. So community transmission was confirmed on March 28, 2020, and after that, we started our data collection on 1st April that ended on 30th May 2020. Comment 3: Recently, articles from Singapore, India, Greece and China – please replace the word ‘articles’ with ‘studies’. Response 3: Corrected as suggested. Comment 4 The authors precisely reported the relevance of the study (i.e., evaluating mental health suffering in doctors). However, I would like to suggest referring a case study reporting excessive fear of COVID-19 existing in the healthcare facilities in Bangladesh. Consequently, a non- COVID-19 patient committed suicide. I think this information can add an extra value to your study rationale. Ref: https://doi.org/10.1016/j.ajp.2020.102295 Response 4 We mentioned the case study as you suggested. Comment 5 Besides, there is a Bangladeshi case-control large scale study accessing suicidality can be discussed and compare what is new in your study. For example, the study got not statistical significance difference in suicidality across general people and healthcare professionals. Ref: https://doi.org/10.1016/j.heliyon.2020.e05259 Response 5 We mentioned the study you suggested to discuss. Comment 6 More information regarding sampling is needed. Please add scale reliability (e.g., Cronbach's alpha) as the scales are not validated in Bangla. Response 6 We applied a convenience sampling technique that was clearly mentioned. Moreover, a flow chart was used to show the pool of samples from where the subjects were recruited based on inclusion criteria. Among the tools applied, only FCV-19S validated among the Bangladeshi population and we mentioned it in the manuscript. For PHQ-4 and SCI-02, as per your suggestion, we added the Cronbach's alpha as appropriate. Comment 7 The authors already mentioned, there is no cutoff score for Fear of COVID scale. However, they categorized it. It would be better if they provide how the category was considered – median? Response 7 We have already mentioned that we used ‘PERCENTILES’ to categorize the severity of the scales that have not yet developed their severity grades (see in the section ‘Instruments used to assess psychological symptom’). For fear of COVID-19, the FCV-19S score was categorized using percentiles of “mild” (score ≤25th percentile, ≤17), “moderate” (score >25th percentile and <75th percentile, 18–23), and “severe” (score ≥75th percentile, ≥24). To convince the readers, in the revised version, we have elaborated the method of categorization adding the percentiles with each cut-off value. This method of categorization also applied for another health-related tool also (https://doi.org/10.1111/jdi.13331). Comment 8 It is requested to report the mean and SD of fear of COVID (along with percentage; as it was done for depression and anxiety) as the scale has no cutoff score yet. Response 8 We have added. Comment 9 Is not it better to report psychological burdens as “probable depression” instead of “depression”, because the scales are short version (although their cutoff scores had higher specificity and sensitivity than the original one; for example, PHQ-2 and PHQ-9)? Similar suggestion for fear of COVID. Besides, the authors may like to perform ANOVA and linear regression – not mandatory if they refer how the cutoff scores were categorized. Response 9 As the generic version of PHQ-2 avoided the term “probable” to represent depression and showed an acceptable level of reliability, we used ‘depression’ to align with the main paper (A 4-item measure of depression and anxiety: Validation and standardization of the Patient Health Questionnaire-4 (PHQ-4) in the general population, https://doi.org/10.1016/j.jad.2009.06.019). Regarding fear of COVID-19, we presented the severity using the scientific method as described above. As this severity scale has not yet developed and validated, hence we pretested it (as part of the whole questionnaire) in a small sample of frontline doctors as mentioned in the section entitled ‘Questionnaire content’. We used multinomial logistic regression based on the categorization of the tools as explained in the revised version and also stated above to respond to the reviewer inquiry. Comment 10 Discussion is focused on the results. Discussing prevalence rate across countries is perfect, even the authors compared with the review article prevalence rate, which is appreciated. By the way, may I suggest to compare the Bangladeshi general people mental health problems rates, this may help the reader to be informed about how this study findings differ from the general people. Please refer Bangladeshi studies assessing the pandemic related psychological burdens (ref: https://doi.org/10.31234/osf.io/q4k5b). Response 10 Thank you for your valuable input. As the suggested study is a pre-print version, we avoided it and cited a nation-wide study that reported psychological symptoms among the general population of Bangladesh (https://doi.org/10.1016/j.jad.2020.10.036). We compared the prevalence of depression as appropriate. Comment 11 The current study is the first study in Bangladesh that provides the burden and associated factors for doctors’ poor mental health outcome during the COVID-19 pandemic. – I request to avoid the first word, because there is a published paper on suicidal behavior of the cohort. Response 11 As per our knowledge, this study is absolutely first as it reported the most commonly reported psychological symptoms that were not previously reported by any study of Bangladesh among frontline doctors. We respect the comment of the reviewer and want to mention that the previous study assessed suicidal behavior, not the ‘anxiety, depression, fear, and sleep disturbance’. Moreover, this study solely conducted among frontline doctors (not intern doctors) who were involved in treating the confirmed or suspected cases of COVID-19. (D) Overall Comments Overall observations, the authors did a great job addressing mental health problems of the vulnerable cohort to virus infection, which may have influence in policy level. And the paper was written in a good flow, and also provided some of recommendations. Best of luck. Response Thanks for the overall comments."
}
]
},
{
"id": "74583",
"date": "07 Dec 2020",
"name": "Ganiyu Adeniyi Amusa",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGeneral Comments:\n\nAn interesting study, well written, appropriate methodology/statistical analysis and evidence-based conclusion derived from the results. The tables are appropriate, clear and precise, no redundancy observed. The references are appropriate, recent and properly cited; no redundancy noted in the referencing\n\nAbstract:\nWell written, established the basis for the study and highlighted main findings of the study. The study revealed that 36.5% had anxiety, 38.4% had depression, 18.6% had insomnia, and 31.9% had fear of COVID-19. Inadequate resources in the workplace were found as the single most significant predictor for all psychological outcomes.\n\nIntroduction:\nThe origin of the pandemic was well captured as well as the attendant societal and individual consequences. The story of the pandemic in Bangladesh was also well captured and the basis for the need to do the study was adequately elucidated bearing in mind the unique characteristics of the Bangladesh community.\n\nMethods:\nThe methodology was described in detail, clear enough and can be reproduced elsewhere. The questionnaire content was detailed and appropriate and was well able to meet the objectives of the study. The instruments used were appropriate - Patient Health Questionnaire-4 (PHQ-4) to assess anxious and depressive symptoms, Sleep Condition Indicator (SCI-02) to assess sleep disturbance and Fear of Coronavirus-19 Scale (FCV-19S) was used to measure one’s fear of COVID-19. The cut-off used and methods of determining the score are appropriate and scientific. The statistics (descriptive and inferential) used were appropriate for the results and conclusion reached.\n\nResults:\nThe results were presented in appropriate formats with clear details. No redundancy was observed.\n\nDiscussion:\nThis was extensive and relevant examples were used in the discussion as related to the findings of the study. The result of a metanalysis of similar studies was comparable to that of the study. Similar findings were obtained in India, Singapore and Taiwan.\n\nConclusion:\nThe study concluded that a high burden of COVID-19 related anxiety, depression, sleep disturbance, and fear existed among Bangladeshi frontline doctors. This demands policymakers' immediate attention to take appropriate preventive measures. The substantial burden of different mental health outcomes elucidated in the current study demands mental health counselors in hospital settings where appropriate.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6179",
"date": "16 Dec 2020",
"name": "Lingkan Barua",
"role": "Author Response",
"response": "Dear reviewer, thank you for your positive observations regarding our manuscript."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1304
|
https://f1000research.com/articles/9-650/v1
|
26 Jun 20
|
{
"type": "Study Protocol",
"title": "The effect of two clinical criteria in the assessment of caries lesions around restorations in children (CARDEC-03): study protocol for a diagnostic randomized clinical trial",
"authors": [
"Bruna Lorena Pereira Moro",
"Cácia Signori",
"Raiza Dias Freitas",
"Laura Regina Antunes Pontes",
"Tathiane Larissa Lenzi",
"Tamara Kerber Tedesco",
"Daniela Prócida Raggio",
"Mariana Minatel Braga",
"Kim Rud Ekstrand",
"Maximiliano Sérgio Cenci",
"Fausto Medeiros Mendes",
"CARDEC collaborative group",
"CaCIA collaborative group",
"Bruna Lorena Pereira Moro",
"Cácia Signori",
"Raiza Dias Freitas",
"Laura Regina Antunes Pontes",
"Tathiane Larissa Lenzi",
"Tamara Kerber Tedesco",
"Daniela Prócida Raggio",
"Mariana Minatel Braga",
"Kim Rud Ekstrand",
"Maximiliano Sérgio Cenci"
],
"abstract": "Introduction: The detection of caries lesions around restoration can be challenging. Therefore, the use of some criteria has been proposed in order to give more objectivity to the diagnosis process. Two of them are the International Dental Federation (FDI) and the Caries Associated with Restorations and Sealants (CARS) criteria. Both methods have a different approach to caries, and it is not possible to know which one of them is the best to use in clinical practice to assess restorations in children. Thus, the present protocol aims to evaluate the effect of the use of the FDI and CARS criteria in the assessment of caries lesions around restorations in primary teeth on outcomes related to oral health in children and costs resulting from the assessments. Methods and analysis: A total of 626 restorations of children from three to 10 years were randomly assessed and are being treated following the FDI criteria (FDI group) or CARS criteria (CARS group). Participants will be followed-up after six, 12, 18, and 24 months. The primary outcome will be the need for a new intervention in the evaluated restorations. This outcome consists of several components, and each of these events will be analyzed separately as secondary outcomes. The changes in children’s oral health-related quality of life and the cost of the restoration dental treatments will also be analyzed as secondary outcomes. The methods will be compared using the Cox regression model with shared frailty. A significance level of 5% will be adopted for all statistical analyses. Discussion: This will be the first randomized clinical study carried out regarding the detection of caries lesions around restorations in primary teeth. Trial registration: The study underwent registration in Clinicaltrials.gov (NCT03520309) on 9 May 2018.",
"keywords": [
"Randomized Controlled Trial",
"Dental Caries",
"Diagnosis",
"Permanent Dental Restoration",
"Dental Restoration Repair",
"Pediatric Dentistry"
],
"content": "Introduction\n\nCaries lesions around restoration, also known as secondary caries or recurrent caries, are the main reason for restoration failure1. However, the detection of these lesions can be challenging for a few reasons, as the presence of gaps between the restoration and tooth surface2 and the presence of stained margins on resin-based composite restorations makes it difficult to differentiate between lesions and demineralization3. For this reason, the use of some criteria has been proposed to give more objectivity to the diagnosis process.\n\nOne such set of criteria is the International Dental Federation (FDI) criteria4, developed in 2007. Although largely used to assess restorations, it evaluates some aspects that might not be directly related to caries lesions, such as marginal staining and marginal adaptation. Using these criteria may lead to a more interventional approach. Another set of criteria is the Caries Associated with Restorations and Sealants (CARS) criteria, which has been integrated into the International Caries Classification and Management System5 and its more recent update, named CariesCare 4D6. The CARS criteria5 focus on aspects related to caries and not on other possible reasons for restoration failure. This method is probably more conservative when it comes to restoration reintervention.\n\nWhen it comes to the management of restorations in primary dentition, it is not possible to know if a more conservative or invasive approach would bring more benefits to children. Restorations that are repaired seem to be more likely to have an additional treatment compared to restorations that are replaced7. On the other hand, replacement often causes the loss of healthy dental structure8,9, leading to a repeated restorative cycle10, increasing the professional time and costs for health systems8.\n\nIt would be preferable that the criteria for assessing caries around restorations in children is in line with the philosophy of minimal intervention dentistry11. However, the majority of studies about the detection of these lesions were performed in vitro, assessed caries lesion in permanent teeth, and did not evaluate relevant aspects to the clinical practice12,13. This lack of evidence inspires the conduction of a third study, which is part of an initiative that aims to build scientific evidence for diagnostic strategies in children - CARies DEtection in Children nº 3 (CARDEC-03).\n\nThus, this trial aims to evaluate the effect of the use of two different visual criteria, the FDI and CARS criteria, for assessing caries lesions around restorations in primary teeth on outcomes related to children’s oral health and costs resulting from the assessments. We hypothesize that the diagnostic criteria that lead to a more conservative approach would bring more benefits to children’s oral health, decreasing the treatment costs and professional time.\n\n\nMethods\n\nA controlled, triple-blind (participant, care provider, outcomes assessor), randomized clinical trial with two parallels arms (1:1) is being carried out. The present protocol is reported according to the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) guidelines14. The completed checklist can be accessed on Figshare15.\n\nThe local ethics committee from the School of Dentistry of the University of São Paulo, São Paulo, Brazil, previously approved the study (registration no. 2.291.642) on 22 September 2017. The participants of the study were recruited from 16 November 2017 to 30 November 2018. The trial was retrospectively registered on Clinicaltrials.gov (NCT03520309) on 9 May 2018 due to of a lack of awareness that registration must occur before enrollment begins. No changes were made to the study after approval by the local ethics committee in 2017, and no results were analyzed before the trial registration on Clinicaltrials.gov. The authors are aware of possible causes of publication bias and selective reporting, and are committed to promoting complete transparency in our research.\n\n\nParticipants, interventions, and outcomes\n\nThis trial is being conducted at the School of Dentistry Dental Clinic of the University of São Paulo, Brazil. The participants (three to 10 years-old) were randomly selected from a list of patients who sought dental treatment at the School of Dentistry. A random sequence was generated using the website “Sealed Envelope” through the tool “Create a randomisation list”. Only patients who fulfilled the eligibility criteria were included in the study after their legal guardians signed the informed consent form and literate children signed an assent form. Both documents are available as Extended data in English16,17 and the original language18,19.\n\nThe inclusion criteria for the present study are children:\n\na) Who have sought treatment at the School of Dentistry;\n\nb) From three to 10 years-old;\n\nc) Presenting at least one restoration of any restorative material (composite resin, amalgam or glass ionomer cement) on a primary tooth (anterior or posterior) regardless of its condition.\n\nThe exclusion criteria for the present study are children:\n\na) Whose parents refuse to participate in the study;\n\nb) Who did not agree to participate, or showed behavior problems during the first appointment.\n\nAll children’s restorations were included for the assessment, except restorations on teeth with fistula, abscess, pulp exposure, history of spontaneous dental pain or mobility. Children presenting these conditions in one or more teeth, but also presenting at least one eligible tooth fitting the inclusion criteria were included in the study.\n\nRandomization was stratified by blocks of different sizes (2, 4, 6 or 8). The strata considered were: (1) children aged 3 to 6 years presenting three restorations or less; (2) children aged 7 to 10 years presenting three restorations or less; (3) children aged 3 to 6 years presenting more than three restorations; (4) children aged 7 to 10 years presenting more than three restorations.\n\nThe random sequence was generated using the website “Sealed Envelope” through the tool “Create a randomisation list”. It was done by an external examiner and to guarantee allocation confidentiality, blocks with allocation sequences were kept in opaque sequential envelopes.\n\nA preliminary visual inspection was performed to assess all participants’ dental surfaces according to the International Caries Detection and Assessment System (ICDAS)20 described in the CariesCare 4D to detect and assess the caries lesions stage and activity6. The assessment was performed by an examiner (LRAP) who is not participating in the subsequent phases of the study. All the assessments of the study are being conducted under a dental clinic setting using a dental chair and artificial illumination. Participants’ teeth receive a professional oral hygiene using a rotating bristle brush, pumice/water slurry and dental floss. A plane buccal mirror and a ball-point probe are being used for all visual inspection and tactile examination of the clinical trial.\n\nThen, children meeting the inclusion criteria were classified into subgroups for further block stratification, according to the number of restorations present in mouth (0 to 3 restorations vs. more than three restorations) and age (3 to 6 years old vs. 7 to 10 years old).\n\nThe children included in the study were randomly allocated in two groups to have their restorations evaluated and treated according to different clinical criteria for caries lesion around restoration:\n\na) FDI group: diagnosis and treatment decision based on the International Dental Federation (FDI) criteria4 (Table 1).\n\nb) CARS group: diagnosis according to the Caries Associated with Restorations and Sealants (CARS) detection criteria, described in the ICCMS5 and in CariesCare 4D6 (Table 2), and proposed treatment decision (Table 3). The definitions and characteristics of activity for primary caries from CariesCare International 4D will also be used in association (Table 4).\n\nThis table was created based on information from Hickel et al. 20104.\n\nThis table was created based on information from Pitts et al. 20165 and Martignon et al. 20196.\n\nThis table was created based on information from Pitts et al. 20165 and Martignon et al. 20196.\n\nThis table was created based on information from Pitts et al. 20165 and Martignon et al. 20196.\n\nThe restorations assessment was performed by an examiner (BLPM), who was trained and calibrated before the beginning of the study. Calibration involves a lecture of clinical criteria, and training was carried out using photos of clinical cases. The web-based training and calibration tool ICDAS Calibration for ICCMS(TM) by ICCMS e-learning was used for this purpose.\n\nAfter these procedures, the examiner evaluated restorations in 10 children who did not participate in the clinical trial. The examiner repeated the same evaluation for intra-examiner agreement. A benchmark examiner (TLL) also performed the tests to assess inter-examiner reproducibility. The assessment of children included in the study started after the intra-examiner and inter-examiner weighted kappa value reached values greater than 0.75 for both FDI and CARS criteria.\n\nFor examinations using the FDI criteria, all tooth surfaces are dried before. When using the CARS criteria, teeth are examined firstly wet and then dried for 5 seconds with a dental 3-in-1 air water syringe.\n\nThe first assessment was performed with the participant’s allocated group (FDI or CARS). After reaching the diagnosis and treatment decision according to the allocated group, the same examiner performed a second assessment according to the other criteria. This procedure aims to compare the methods of the study, and the second assessment did not influence or change the classification and treatment decision proposed by the criteria the participant is allocated. If a legal guardian presents a complaint related to any children’s restoration, it can be repaired or replaced independently of the criteria used. The scores obtained with the restoration assessment were collected using a specific sheet that can be found as Extended data in English21 and Portuguese22.\n\nAt the first appointment, legal guardians were asked to answer a questionnaire to assess the impact on children’s oral health-related quality of life. The instrument used was the Brazilian version23,24 of the Early Childhood Oral Health Impact Scale (ECOHIS)25. Besides that, an anamnesis related to children’s health and medical history was carried out (this form is available as Extended data in English26 and original language27). At the end of the first appointment, oral hygiene instructions were delivered, showing the correct use of toothbrush and fluoride toothpaste (1000 to 1500 ppm of fluoride)28. Dietary advice was also given to all participants and their parents or legal guardians to reduced intake of free sugars throughout the life course29.\n\nFor all appointments, the time spent and materials used on patient care are collected using a specific sheet that can be found as Extended data in English30 and original language31. Parents or guardians are asked about transportation and absenteeism in the workplace.\n\nIn the subsequent appointments, dental treatments following a predefined protocol are being performed by postgraduate dental students in Pediatric Dentistry, who are blind to the criteria used to reach the treatment decision. In all situations, if there is active dentine tissue, it is removed using dentin excavators. Diamond burs are used to remove the restorations, if necessary.\n\nThe treatment decisions for the restorations evaluated according to the FDI and CARS criteria are being classified into:\n\nNo treatment: no intervention needed and the restoration will be followed-up;\n\nProfessional topical fluoride application: a treatment for non-cavitated active caries lesions detected by the CARS criteria;\n\nRefurbishment: restorations finishing and polishing;\n\nRepair: minimally invasive approach resulting in the addition of a restorative material, with or without a preparation of the restoration and/or dental hard tissues32. Composite resin or glass ionomer cement will be used as a restorative material;\n\nReplacement: complete removal of the restoration present on the tooth32. Composite resin will be used as restorative material for the new restoration.\n\nThe presence or absence of soft or hard carious tissue is evaluated and recorded after the restoration removal when replacement is indicated. This procedure is performed to record a possible false-positive diagnosis for dentine caries lesion around the restoration since the authors will also develop an accuracy study nested in this clinical trial.\n\nThe same operators are performing additional dental treatment needs (not related to the restorations included in the study). Treatment plan related to additional dental treatment was carried out by the examiner responsible for children's initial clinical examination. Details of the pre-established treatment protocols can be found in Figure 1.\n\nAfter the completion of the treatment plan, participants will be followed up considering the outcome evaluation after six, 12, 18, and 24 months. At the follow-up visits, if a new dental treatment is needed (related or not to the restorations), necessary procedures will be carried out. Hygiene and dietary instructions will be given to children at each follow-up visit.\n\nThe treatment decisions for the restorations evaluated during the follow-up visits will be decided according to the FDI or CARS criteria, considering the child’s allocation group. The same trained and calibrated examiner (BLPM) who conducted the assessments at the beginning of the study will perform the evaluations.\n\nDuring the 24 months follow-up visit, a new ECOHIS questionnaire will be applied for parents or legal guardians who had previously answered at the time the child was included in the study.\n\nStimuli for participants' adherence to the treatment and follow-up sessions are happening via mobile and social networks. Facebook and Instagram profiles were created to stay in touch with patients through social media. Humanized care is provided for all participants. Explanations about the importance of participation for their benefit are also being given.\n\nThe primary outcome of this trial will be the need for a new intervention during the follow-up of restorations evaluated by different criteria. This outcome consists of several components. Thus, the outcome occurrence will be considered if any of the following conditions are detected:\n\nPresence of secondary caries lesion exposing dentin;\n\nNeed for repair;\n\nNeed for restoration replacement;\n\nNeed for extension of the existing restoration on the examined tooth due to a tooth fracture or caries lesion development exposing dentin;\n\nAn episode of pain or need for endodontic treatment;\n\nExtraction requirement (except in the case of prolonged retention).\n\nThe occurrence of any of these conditions at any time of follow-up will be considered as an event related to the primary outcome. Each of the events that make up the primary outcome will be analyzed separately as secondary outcomes. Changes in children's oral health-related quality of life after two years will be considered as a secondary outcome. The costs and effects per child of the treatments performed during the follow-up, considering the teeth included in our sample, are also going to be analyzed as a secondary outcome.\n\nThe occurrence of the outcomes will be evaluated according to predetermined criteria from two other criteria during the follow-up visits of six, 12, 18, and 24 months. Different criteria will be used according to the number of surfaces the restoration involves:\n\nFor one-surface restorations: the criteria used will be according to Frencken et al.33;\n\nFor a multi-surface restoration: the criteria used will be according to Roeleveld et al.34.\n\nAccording to Frencken et al.33 criteria, scores related to restoration success will be 0, 1 or 7. Those considered to have failed will be scored as 2, 3, 4 or 8; while those considered being unrelated to success and failure will be scored as 5, 6 or 9. Concerning the Roeleveld et al.34 criteria, restoration success will be scored as 00 or 10. Those considered to have failed will be scored as 11, 12, 13, 20, 21, 30 or 40; while those considered being unrelated to success and failure will be scored as 50, 60, 70 or 90.\n\nThe follow-up evaluations will be carried out by an examiner (TKT) blind to children’s allocation group who was previously trained and calibrated for both criteria and not participating in the previous phases of the trial.\n\nThe sample size calculation was performed based on the primary outcome (percentage of restorations requiring reintervention). A failure rate of 10% after two years was considered for occlusal restorations35 and 30% for occlusal-proximal restorations36. It was also considered that approximately 10% of the replaced restorations and 14% of the restorations undergoing repair fail again37. Considering that half of the sample is occlusal restorations, an operative reintervention requirement rate of 24% is expected in two years. The minimum number of 522 restorations was reached, based on an absolute difference of 10% between the groups, using a two-tailed test. As a child can contribute with more than one restoration, 20% was added to the sample size (n = 626).\n\nConsidering that children with restored teeth have on average 3.7 restorations38, and adding 20% for possible participants loss, a minimum number of 204 children presenting at least one restored primary tooth is required to be included in this trial.\n\n\nData management and analysis\n\nClinical data will be entered directly into predetermined sheets. Data quality will be ensured by validation checks that include missing data, out-of-range values, and illogical and invalid responses.\n\nExaminers' reproducibility will be performed using the weighted kappa test, calculating the weighted value of kappa and also the 95% confidence intervals. The primary outcome of the study is a dichotomous variable (with or without the need for intervention); therefore, the unit of analysis is the restored tooth. As children can have more than one tooth included in the study, the comparison between the groups will be carried out using survival analysis, considering the cluster-effect. Kaplan-Meyer graphs will be constructed, and the methods will be compared using the Cox regression model with a shared frailty.\n\nSecondary clinical outcomes will also be analyzed using the same statistical tests. Quality of life will be analyzed using Poisson regression analysis and the unit of analysis will be the child.\n\nA trial-based economic evaluation will be performed considering the difference of the inputs (costs) and outputs (effects) of the two diagnostic criteria (FDI and CARS) after two years. Further details regarding the economic evaluation will be described on a health economic analysis plan.\n\nA p-value of 5% as the level of significance will be considered for all tests. The analyses will be performed using the statistical package Stata 13.0 (Stata Corp, College Station, USA).\n\n\nParticipant recruitment and timeline\n\nRecruitment took place at the School of Dentistry of the University of São Paulo from November 2017 to November 2018. Each allocated participant will have an average treatment period of one month and will be followed-up for 24 months, resulting in a total of 25 months of enrollment. The detailed timeline for data collection is summarized in Figure 2.\n\nECOHIS, Early Childhood Oral Health Impact Scale; FDI, International Dental Federation; CARS, Caries Associated with Restorations and Sealants.\n\n\nMonitoring\n\nNo data monitoring committee is needed in this trial since adverse events are unlikely to happen during restoration evaluation and dental treatments. For this reason, the chief investigator of the study (FMM) will assume an independent oversight of trial data collection, management, and analysis.\n\nThe effects expected in this study are the ones listed as trial outcomes. All of them are usually expected to happen during pediatric dentistry clinical practice. Any other adverse event is unlikely to happen.\n\nThe data will be periodically subjected to audit by the coordinator of the study. Any discrepancies will be verified, corrected and registered.\n\n\nEthics and dissemination\n\nSequential numbers will be used to identify and ensure participant confidentiality. Participants’ identifiable information will be stored in filing cabinets in a locked secure room.\n\nThe full data generated from this trial will be placed in a public repository (University of São Paulo Data Repository).\n\nParticipants included in this trial will have dental treatments provided at the School’s dental clinic during and after the completion of the trial if necessary.\n\nAll the findings of this trial will be reported in peer-reviewed journals, patient newsletters and the School of Dentistry of University of São Paulo website.\n\n\nStudy status\n\nThe patient recruitment took place from 16 November 2017 to 30 November 2018. The follow-up evaluations of 6 and 12 months were concluded; however, the study is now temporarily suspended since 16 March 2020 due to COVID-19.\n\n\nDiscussion\n\nRestoration assessment is a challenge in dentistry, and the main point of debate is caries around restoration1,39. However, due to the scarcity of well-conducted studies, its diagnosis is not based on objective clinical criteria, and there is a considerable variation in the criteria used. As a consequence, a significant number of restorations presenting small defects are often indicated to be replaced since they can be misdiagnosed as caries lesions8. Also, there is no homogeneity on the treatment decision-making for secondary caries between dentists13,40, and studies based on clinical practice have shown that they tend to replace more restorations than necessary41.\n\nTwo recently published systematic reviews included around 20 accuracy studies of methods for detecting caries lesions around restorations12,13. The majority of these studies were performed in vitro, assessed caries lesions in permanent teeth, and did not evaluate relevant aspects to the clinical practice12,13. Nevertheless, the decision on what is the best method to be used should evaluate whether patients undergoing such methods would have greater health-related benefits than patients undergoing some other method42. For this assessment, ultimate health outcomes for patients must be considered. The experimental design to assess it is the randomized clinical trial (Phase IV question).\n\nRandomized clinical trials are considered the best study design on which clinicians and policy-makers rely most to determine whether an intervention is effective43. However, as far as we know, no randomized clinical study has been carried out regarding the detection of caries lesions around restorations in primary teeth. Besides that, no study compared the accuracy of FDI and CARS criteria clinically to detect caries around restoration on primary teeth, and the impact of the use of the criteria on the restorative treatment decisions for children. For this reason, an accuracy study (Phase III question) with the FDI and CARS methods will be developed nested to this trial.\n\nFor the present trial, the authors decided to use among the FDI criteria the subcategories marginal staining and marginal adaptation, beyond recurrence of caries. The decision was based on the fact that both aspects can be misinterpreted with secondary caries during restoration assessment44–46. Therefore, we tried to simulate what can clinically be a reason for restoration reintervention in the daily clinical practice. Regarding the CARS criteria, the system does not present any treatment decision linked to the evaluation method. For this reason, we adapted the decisions based on the ICCMS recommendations for treating primary caries lesions47.\n\nThe study’s limitation is that the first assessment performed with the participant’s allocation group (FDI or CARS criteria) and the second assessment according to the other criteria will be done at the same dental appointment. This will be done to reduce the number of dental appointments for the patients, enhancing their adherence to the clinical research. However, a carry-over effect could occur between the methods. Contrariwise, a strength of the study is the procedure used to avoid selection bias. The evaluations will be conducted in a sample of children randomly selected from a list of patients who sought dental treatment at our School. Besides that, the outcome assessor will be blinded regarding the allocation group to avoid assessment bias.\n\nThus, with the development of this clinical trial and expected results, we aim to define between FDI and CARS criteria the best approach for diagnosis and management of dental restorations in children, considering the impact on the treatment decision on clinically relevant outcomes for the patient and costs resulting from the treatments performed.\n\n\nData availability\n\nNo underlying data are associated with this article.\n\nFigshare: Consent form. https://doi.org/10.6084/m9.figshare.12327644.v116.\n\nFigshare: Consent form in the original language (Portuguese). https://doi.org/10.6084/m9.figshare.12327674.v118.\n\nFigshare: Assent form. https://doi.org/10.6084/m9.figshare.12327731.v117.\n\nFigshare: Assent form in the original language (Portuguese). https://doi.org/10.6084/m9.figshare.12327779.v119.\n\nFigshare: Restorations assessment form. https://doi.org/10.6084/m9.figshare.12331460.v121.\n\nFigshare: Restorations assessment form in the original language (Portuguese).\n\nhttps://doi.org/10.6084/m9.figshare.12331466.v122.\n\nFigshare: Anamnesis form. https://doi.org/10.6084/m9.figshare.12324212.v126.\n\nFigshare: Anamnesis form in the original language (Portuguese). https://doi.org/10.6084/m9.figshare.12327578.v127.\n\nFigshare: Time and cost form. https://doi.org/10.6084/m9.figshare.12327854.v130.\n\nFigshare: Time and cost form in the original language (Portuguese). https://doi.org/10.6084/m9.figshare.12331451.v131.\n\nFigshare: SPIRIT checklist. https://doi.org/10.6084/m9.figshare.12331484.v115.\n\nData are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).",
"appendix": "Acknowledgements\n\nThe authors would like to thank CARDEC-03 and CaCIA collaborative groups. Both collaborative groups have shared the ideas and collaborated with the planning and establishment of the present study. Members of each group can be found below. We also wish to thank the participants of the Post-Graduation in Pediatric Dentistry Seminar of University of São Paulo School of Dentistry (FOUSP) for the critical comments put forth.\n\nCaCIA collaborative group – Trial 1:\n\nAna Beatriz Lima de Queiroz, Alessandra Braga de Avila, Bruna Oliveira Souza Cácia Signori, Camila Raubach Dias, Camila Thurow Becker, Eduardo Trota Chaves, Eugênia Carrera Malhão, Elenara Ferreira de Oliveira, Juliana Lays Stolfo Uehara, Fernanda Gonçalves da Silva, Fernanda Srynczyk da Silva, Gabriel Ventura Lima Kucharski, Gabriele Ribeiro dos Santos, Julia Macluf Torres, Karoline Von Ahn Pinto, Laura Lourenço Morel, Leonardo Blank Weymar, Marcelo Pereira Brod, Maria Fernanda Gamborgi, Maximiliano Sérgio Cenci, Renata Uliana Posser, Thaís da Silva Vieira, Vitor Henrique Romero Digmayer, Wagner da Silva Nolasco and Wagner Martins da Silva Leal.\n\nCARDEC collaborative group – Trial 3:\n\nAna Laura Passaro, Annelry Costa Serra, Antonio Carlos Lopes Silva, Bruna Lorena Pereira Moro, Carolina de Picoli Acosta, Caroline Mariano Laux, Cíntia Saori Saihara, Daniela Prócida Raggio, Fausto Medeiros Mendes, Haline Cunha Medeiros Maia, Isabel Cristina Olegário da Costa, Isabella Ronqui de Almeida, Jhandira Daibelis Yampa Vargas, Jonathan Rafael Garbim, José Carlos P. Imparato, Julia Gomes Freitas, Karina Haibara De Natal, Kim Rud Ekstrand, Laura Regina Antunes Pontes, Mariana Bifulco, Mariana Minatel Braga, Mariana Pinheiro de Araújo, Mayume Amorim do Vale, Raiza Dias Freitas, Renata Marques Samuel, Rita Baronti, Rodolfo de Carvalho Oliveira, Simone Cesar, Tamara Kerber Tedesco, Tathiane Larissa Lenzi, Tatiane Fernandes Novaes and Thais Gimenez.\n\n\nReferences\n\nNedeljkovic I, De Munck J, Vanloy A, et al.: Secondary caries: prevalence, characteristics, and approach. Clin Oral Investig. 2020; 24(2): 683–691. PubMed Abstract | Publisher Full Text\n\nNassar HM, González-Cabezas C: Effect of gap geometry on secondary caries wall lesion development. Caries Res. 2011; 45(4): 346–52. PubMed Abstract | Publisher Full Text\n\nMJÖR IA: Clinical diagnosis of recurrent caries. J Am Dent Assoc. 2005; 136(10): 1426–33. PubMed Abstract | Publisher Full Text\n\nHickel R, Peschke A, Tyas M, et al.: FDI World Dental Federation - clinical criteria for the evaluation of direct and indirect restorations. Update and clinical examples. J Adhes Dent. 2010; 12(4): 259–72. PubMed Abstract | Publisher Full Text\n\nPitts NB, Ismail AI, Martignon S, et al.: ICCMSTM Guide for Practitioners and Educators. 2016; 1–45. Reference Source\n\nMartignon S, Pitts NB, Goffin G, et al.: CariesCare practice guide: consensus on evidence into practice. Br Dent J. 2019; 227(5): 353–362. PubMed Abstract | Publisher Full Text\n\nGordan VV, Riley JL, Rindal DB, et al.: Repair or replacement of restorations. J Am Dent Assoc. 2015; 146: 895–903. Publisher Full Text\n\nDemarco FF, Collares K, Correa MB, et al.: Should my composite restorations last forever? Why are they failing? Braz Oral Res. 2017; 31(suppl 1): e56. PubMed Abstract | Publisher Full Text\n\nEltahlah D, Lynch CD, Chadwick BL, et al.: An update on the reasons for placement and replacement of direct restorations. J Dent. 2018; 72: 1–7. PubMed Abstract | Publisher Full Text\n\nElderton RJ: Clinical Studies Concerning Re-Restoration of Teeth. Adv Dent Res. 1990; 4: 4–9. PubMed Abstract | Publisher Full Text\n\nFrencken JE, Peters MC, Manton DJ, et al.: Minimal Intervention Dentistry for Managing Dental Caries - A Review: Report of a FDI Task Group. Int Dent J. 2012; 62(5): 223–43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrouwer F, Askar H, Paris S, et al.: Detecting Secondary Caries Lesions: A Systematic Review and Meta-analysis. J Dent Res. 2016; 95(2): 143–51. PubMed Abstract | Publisher Full Text\n\nSignori C, Gimenez T, Mendes FM, et al.: Clinical relevance of studies on the visual and radiographic methods for detecting secondary caries lesions - A systematic review. J Dent. 2018; 75: 22–33. PubMed Abstract | Publisher Full Text\n\nChan AW, Tetzlaff JM, Altman DG, et al.: SPIRIT 2013 statement: Defining standard protocol items for clinical trials. Ann Intern Med. 2013; 158(3): 200–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoro BLP, Signori C, Freitas R, et al.: CARDEC-03 - SPIRIT_checklist. 2020. http://www.doi.org/10.6084/m9.figshare.12331484.v1\n\nMoro BLP, Signori C, Freitas R, et al.: CARDEC-03 - Patient Consent Form. 2020. http://www.doi.org/10.6084/m9.figshare.12327644.v1\n\nMoro BLP, Signori C, Freitas R, et al.: CARDEC-03 - Assent Form. 2020. http://www.doi.org/10.6084/m9.figshare.12327731.v1\n\nMoro BLP, Signori C, Freitas R, et al.: CARDEC-03 - Patient Consent Form in Portuguese. 2020. http://www.doi.org/10.6084/m9.figshare.12327674.v1\n\nMoro BLP, Signori C, Freitas R, et al.: CARDEC-03 - Assent Form in Portuguese. 2020. http://www.doi.org/10.6084/m9.figshare.12327779.v1\n\nShivakumar KM, Prasad S, Chandu GN: International Caries Detection and Assessment System: A new paradigm in detection of dental caries. J Conserv Dent. 2009; 12(1): 10–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoro BLP, Signori C, Freitas R, et al.: CARDEC-03 - Restorations Assessment Form. 2020. http://www.doi.org/10.6084/m9.figshare.12331460.v1\n\nMoro BLP, Signori C, Freitas R, et al.: CARDEC-03 - Restorations Assessment Form in Portuguese. 2020. http://www.doi.org/10.6084/m9.figshare.12331466.v1\n\nTesch FC, De Oliveira BH, Leão A: Equivalência semântica da versão em português do instrumento Early Childhood Oral Health Impact Scale. Cad Saude Publica. 2008; 24(8): 1897–909. Publisher Full Text\n\nMartins-Júnior PA, Ramos-Jorge J, Paiva SM, et al.: Validação da versão brasileira do Early Childhood Oral Health Impact Scale (ECOHIS). Cad Saude Publica. 2012; 28(2): 367–74. PubMed Abstract | Publisher Full Text\n\nPahel B, Rozier RG, Slade GD: Parental perceptions of children’s oral health: The Early Childhood Oral Health Impact Scale (ECOHIS). Health Qual Life Outcomes. 2007; 5: 6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoro BLP, Signori C, Freitas R, et al.: CARDEC-03 - Anamnesis Form.pdf. 2020. http://www.doi.org/10.6084/m9.figshare.12324212.v1\n\nMoro BLP, Signori C, Freitas R, et al.: CARDEC-03 - Anamnesis Form in Portuguese. 2020. http://www.doi.org/10.6084/m9.figshare.12327578.v1\n\nWalsh T, Worthington HV, Glenny AM, et al.: Fluoride toothpastes of different concentrations for preventing dental caries in children and adolescents. Cochrane Database Syst Rev. 2010; 20(1): CD007868.PubMed Abstract | Publisher Full Text\n\nHarris R, Gamboa A, Dailey Y, et al.: One-to-one dietary interventions undertaken in a dental setting to change dietary behaviour. Cochrane Database Syst Rev. 2012; 2012(3): CD006540. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoro BLP, Signori C, Freitas R, et al.: CARDEC-03 - Time and Cost Form. 2020. http://www.doi.org/10.6084/m9.figshare.12327854.v1\n\nMoro BLP, Signori C, Freitas R, et al.: CARDEC-03 - Time and Cost Form in Portuguese. 2020. http://www.doi.org/10.6084/m9.figshare.12331451.v1\n\nHickel R, Brüshaver K, Ilie N: Repair of restorations – Criteria for decision making and clinical recommendations. Dent Mater. 2013; 29(1): 28–50. PubMed Abstract | Publisher Full Text\n\nFrencken JE, Makoni F, Sitholea WD: Atraumatic restorative treatment and glass-lonomer sealants in a school oral health programme in zimbabwe: Evaluation after 1 year. Caries Res. 1996; 30(6): 428–33. PubMed Abstract | Publisher Full Text\n\nRoeleveld AC, van Amerongen WE, Mandari GJ: Influence of residual caries and cervical gaps on the survival rate of Class II glass ionomer restorations. Eur Arch Paediatr Dent. 2006; 7(2): 85–91. PubMed Abstract | Publisher Full Text\n\nQvist V, Poulsen A, Teglers PT, et al.: The longevity of different restorations in primary teeth. Int J Paediatr Dent. 2010; 20(1): 1–7. PubMed Abstract | Publisher Full Text\n\nTedesco TK, Calvo AFB, Lenzi TL, et al.: ART is an alternative for restoring occlusoproximal cavities in primary teeth - evidence from an updated systematic review and meta-analysis. Int J Paediatr Dent. 2017; 27(3): 201–209. PubMed Abstract | Publisher Full Text\n\nGordan VV, Riley JL, Rindal DB, et al.: Repair or replacement of restorations. J Am Dent Assoc. 2015; 146(12): 895–903. Publisher Full Text\n\nGimenez T, Moro BLP, Camargo LB: Impact of visual inspection and radiographs for caries detection in children: a 2-year randomized clinical trial. J Am Dent Assoc. 2020; 151(6): 407–415.e1. Publisher Full Text\n\nNedeljkovic I, Teughels W, De Munck J, et al.: Is secondary caries with composites a material-based problem? Dent Mater. 2015; 31(11): e247–77. PubMed Abstract | Publisher Full Text\n\nAlomari Q, Al-Saiegh F, Qudeimat M, et al.: Recurrent Caries at Crown Margins: Making a Decision on Treatment. Med Princ Pract. 2009; 18(3): 187–92. PubMed Abstract | Publisher Full Text\n\nGordan VV, Riley JL 3rd, Geraldeli S, et al.: Repair or replacement of defective restorations by dentists in the dental practice-based research network. J Am Dent Assoc. 2012; 143(6): 593–601. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSackett DL, Haynes RB: The architecture of diagnostic research. BMJ. 2002; 324(7336): 539–41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChassé M, Fergusson DA: Diagnostic Accuracy Studies. Semin Nucl Med. 2019; 49(2): 87–93. PubMed Abstract | Publisher Full Text\n\nMjör IA, Toffenetti F: Secondary caries: a literature review with case reports. Quintessence Int. 2000; 31(3): 165–79. PubMed Abstract\n\nKidd EA, Beighton D: Prediction of Secondary Caries around Tooth-colored Restorations: A Clinical and Microbiological Study. J Dent Res. 1996; 75(12): 1942–6. PubMed Abstract | Publisher Full Text\n\nFoster LV: Validity of clinical judgements for the presence of secondary caries associated with defective amalgam restorations. Br Dent J. 1994; 177(3): 89–93. PubMed Abstract | Publisher Full Text\n\nPitts NB, Ekstrand KR: International Caries Detection and Assessment System (ICDAS) and its International Caries Classification and Management System (ICCMS) - methods for staging of the caries process and enabling dentists to manage caries. Community Dent Oral Epidemiol. 2013; 41(1): e41–52. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "65636",
"date": "20 Jul 2020",
"name": "Andrea Zandoná",
"expertise": [
"Reviewer Expertise Cariology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe aim of this study is to evaluate the effect of two different visual criteria, the FDI and CARS criteria for the assessment of caries around restorations in primary teeth on outcomes related to children’s oral health and costs.\nThis is a timely study that is assessing an important topic from the outcomes point of view, which makes it very clinically relevant. The study protocol is well written, provides a rationale, and clearly describes the objectives of the study. The study design is appropriate for the research question. There are however insufficient details of the methods to allow replication by others. The following should be clarified: Given the impact that caries risk can have on the outcomes, it is important to understand if caries risk was assessed and played any role in randomization. Additionally, although caries lesion activity was assessed, it is not clear if randomization considered caries lesion activity status. It is also not clear if exfoliation status was considered and how that affected inclusion/exclusion. Regarding the 4 stratification blocks, please clarify if the number of restorations considered were only those in the primary dentition or if the restorations in the permanent dentition were also considered for stratification. On examiner calibration – please clarify if the repeat examinations included all 10 children. It is stated that examinations of study patients only occurred after weighted values were greater than 0.75. Please clarify if the same 10 children were repeatedly examined to achieve these values for both intra- and inter-examiner reliability. Were new Kappa scores calculated for new exams or were the exams compounded and the Kappa scores re-calculated? After removal of the restoration when the replacement was indicated the hardness of dentin was evaluated, please clarify if this was done by the provider or by a different examiner and if the calibration was conducted prior to the assessments. On Adherence, the authors state that “humanized care is provided for all participants. Please provide a definition of “humanized care”. Management: Figure 1 provides a very clear flow chart of how lesions were managed according to the ICDAS criteria. A similar chart with the FDI criteria and management would be helpful. Additionally, the predetermined protocols need to be overlaid with the FDI and CARS to better illustrate how the different cases were/are handled.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Partly\n\nAre the datasets clearly presented in a useable and accessible format? Not applicable",
"responses": [
{
"c_id": "5982",
"date": "02 Oct 2020",
"name": "Fausto Mendes",
"role": "Author Response",
"response": "Dear Professor Andrea Zandoná. We are grateful for the comments and suggestions provided. In order to facilitate the review process, your comments are summarized before the respective answers. Reviewer: Given the impact that caries risk can have on the outcomes, it is important to understand if caries risk was assessed and played any role in randomization. Additionally, although caries lesion activity was assessed, it is not clear if randomization considered caries lesion activity status. It is also not clear if exfoliation status was considered and how that affected inclusion/exclusion. Our response: Dear Professor Andrea Zandoná, the following paragraph was added to the manuscript to clarify these points (Methods, page 7, 1st paragraph): “All participants of the study could be classified as having a high caries risk since past caries experience is the most important component for the development of caries lesions. However, stratified randomization was performed considering the children's number of restorations to subdivide them in children with higher and lower caries experience. The caries lesion activity was not considered for randomization. On the other hand, the children's age was a parameter for stratification in order to consider the different time of exfoliation of the evaluated teeth. In this way, the number of teeth with different times of exfoliation was balanced between the FDI and CARS criteria”. A new reference was also added to this paragraph: Twetman S, Fontana M, Featherstone JD. Risk assessment - can we achieve consensus?. Community Dent Oral Epidemiol. 2013;41(1):e64-e70. doi:10.1111/cdoe.12026 Regarding considering the exfoliation status in inclusion and exclusion criteria, we excluded from the sample restorations on teeth with mobility. This is how the exfoliation status affected the exclusion criteria of the study (Methods, page 6, 3rd paragraph). Reviewer: Regarding the 4 stratification blocks, please clarify if the number of restorations considered were only those in the primary dentition or if the restorations in the permanent dentition were also considered for stratification. Our response: Dear reviewer, we added a sentence to explain that “the number of restorations considered for stratification were those placed in primary and permanent teeth” (Methods, page 6, 4th paragraph). Reviewer: On examiner calibration – please clarify if the repeat examinations included all 10 children. It is stated that examinations of study patients only occurred after weighted values were greater than 0.75. Please clarify if the same 10 children were repeatedly examined to achieve these values for both intra- and inter-examiner reliability. Were new Kappa scores calculated for new exams or were the exams compounded and the Kappa scores re-calculated? Our response: Dear reviewer, thank you for your suggestions. Two sentences were added to answer your question: “After these procedures, the examiner evaluated restorations in 10 children who did not participate in the clinical trial. The examiner repeated the same evaluation, in the same 10 children, for intra-examiner agreement. A benchmark examiner (TLL) also performed the tests to assess inter-examiner reproducibility twice in the same sample of children. In this way, the exams were compounded, and the weighted kappa scores were re-calculated. The assessment of children included in the study started after the intra-examiner and inter-examiner weighted kappa value reached values greater than 0.75 for both FDI and CARS criteria” (Methods, pages 8 and 9, 4th paragraph). Reviewer: After removal of the restoration when the replacement was indicated the hardness of dentin was evaluated, please clarify if this was done by the provider or by a different examiner and if the calibration was conducted prior to the assessments. Our response: Dear reviewer, the missing information was provided on Methods section, page 11, and paragraph 1st: “The presence or absence of soft or hard carious tissue is evaluated and recorded by the postgraduate dental student who provides dental care after the restoration removal when replacement is indicated. Training and calibration were conducted before the assessments. An experienced researcher in Cariology performed a theoretical lecture about the clinical characteristics of caries lesions, and training was carried out using photos of clinical cases. The procedure of evaluating the carious tissue is performed to record a possible false-positive diagnosis for dentine caries lesion around the restoration since the authors will also develop an accuracy study nested in this clinical trial”. Reviewer: On Adherence, the authors state that “humanized care is provided for all participants. Please provide a definition of “humanized care”. Our response: Dear reviewer, a definition of “humanized care” was provided on the “Adherence” section: “Humanized care is provided for all participants, focusing on the patient's well-being and providing empathy, affection, and familiarity between the CARDEC collaborative group and children and their families” Reviewer: Figure 1 provides a very clear flow chart of how lesions were managed according to the ICDAS criteria. A similar chart with the FDI criteria and management would be helpful. Additionally, the predetermined protocols need to be overlaid with the FDI and CARS to better illustrate how the different cases were/are handled. Our response: Dear reviewer, thank you for your suggestion. We created three new flowcharts to illustrate how restorations are managed according to the FDI criteria (Figure 1) and the CARS system (Figure 2). The third flowchart (Figure 3) was done to illustrate how the clinical assessment of restoration is handled differently according to both criteria used in our study. We added on the “Interventions” section of the study protocol the figure numbers after the explanation of the FDI and CARS group. We also added a new sentence to the paper: “A clinical example of the restoration assessment performed with both FDI and CARS criteria is illustrated in Figure 3”."
}
]
}
] | 1
|
https://f1000research.com/articles/9-650
|
https://f1000research.com/articles/9-1349/v1
|
19 Nov 20
|
{
"type": "Brief Report",
"title": "‘The long tail of Covid-19’ - The detection of a prolonged inflammatory response after a SARS-CoV-2 infection in asymptomatic and mildly affected patients",
"authors": [
"Ivan Doykov",
"Jenny Hällqvist",
"Kimberly C. Gilmour",
"Louis Grandjean",
"Kevin Mills",
"Wendy E. Heywood",
"Ivan Doykov",
"Jenny Hällqvist",
"Kimberly C. Gilmour",
"Louis Grandjean",
"Kevin Mills"
],
"abstract": "‘Long Covid’, or medical complications associated with post SARS-CoV-2 infection, is a significant post-viral complication that is being more and more commonly reported in patients. Therefore, there is an increasing need to understand the disease mechanisms, identify drug targets and inflammatory processes associated with a SARS-CoV-2 infection. To address this need, we created a targeted mass spectrometry based multiplexed panel of 96 immune response associated proteins. We applied the multiplex assay to a cohort of serum samples from asymptomatic and moderately affected patients. All patients had tested positive for a SARS-CoV-2 infection by PCR and were determined to be subsequently positive for antibodies. Even 40-60 days post-viral infection, we observed a significant remaining inflammatory response in all patients. Proteins that were still affected were associated with the anti-inflammatory response and mitochondrial stress. This indicates that biochemical and inflammatory pathways within the body can remain perturbed long after SARS-CoV-2 infections have subsided even in asymptomatic and moderately affected patients.",
"keywords": [
"Sars-CoV-2",
"mass spectrometry",
"inflammation",
"biomarker",
"proteomics"
],
"content": "Introduction\n\nAs more and more people are recovering from SARS-CoV-2 infection, one of the growing concerns is the increasing reports of the post viral fatigue symptoms or ‘long Covid’. This phenomenon is defined as not recovering for several weeks or months following the start of symptoms and whereby patients present with chronic and recurrent fatigue for weeks and even many months after a SARS-CoV-2 infection1,2. Understanding the effects and complications of ‘long Covid’, and then managing it, is the next challenge for public health services. Currently the UK is increasing its testing capacity for virus detection and antibody detection, but there still remains a gap in the understanding and diagnosis of long Covid.\n\nWork has been performed to characterise the inflammatory response to SARS-CoV-2 infection in relation to disease severity. There has been controversy as to whether severity is associated with a hyperinflammatory cytokine storm or failure of host protective immunity that results in unrestrained viral dissemination and organ injury. What has made addressing this question challenging has been the lack of diagnostic tools to evaluate immune function in Covid-19 infections. There are sets of simple but expensive immunoassay panels that are available to look at known key inflammatory proteins such as cytokine panels; however, these only give information on known pathways and limit discovery of novel or less defined inflammatory responses. Targeted proteomics using mass spectrometry to quantitate multiple diagnostic proteins can help address this. Novel assays for virus detection have been already developed using targeted mass spectrometry3,4 but no assays are available to look at the symptoms for the diagnosis or understanding of ‘long Covid’.\n\nFrom a previous study (unpublished reports) we developed a custom targeted mass spectrometry based assay panel that looks at up to 96 pro- and anti-inflammatory associated proteins (Figure 1a; see Table 1 on protocols.io5). Our hypothesis was long Covid symptoms could be related to a lingering ‘tail’ and an abnormal inflammatory response to an infection, by a type of virus the body has not seen before. We applied this assay to a cohort of samples taken from healthcare workers who had tested positive for SARS-CoV-2 infection by PCR and were either asymptomatic or had only a mild infection. Samples were taken at least 40–45 days post infection and demonstrated a positive antibody test. We compared these with serum from healthcare workers with a negative antibody test, no reported infection and no positive PCR test.\n\na) Representative overlaid chromatogram of the multiplex inflammation panel. Protein identifiers indicated by name and followed by first three amino acids of the peptide. b) Principle component analysis score plot of 10 SARS-Cov-2 infected patients >40 days post infection and 10 negative controls.\n\n\nMethods\n\nSamples were identified from the Health Research Authority approved project Co-Stars (Great Ormond Street Hospital NHS Trust COSTARS, IRAS 282713, ClinicalTrials.gov Identifier: NCT04380896, registered May 8th 2020) and all participants provided informed written consent.\n\nA pilot group of 10 positive and 10 negative samples covering a broad age range was selected as proof of principle for this assay. The negative group was 60% female with an age-range of 21–57, median 38 years. The positive group was 69% female, with an age range 31–66 and median age of 44 years. Of the positive patients, seven were asymptomatic and six had loss of taste/smell or had abnormal taste/smell. None were admitted to hospital or reported other symptoms.\n\nThe detailed method for the multiplex assay is published and available at protocols.io5. Briefly, serum sample proteins were precipitated and trypsin digested to peptides. Peptides were desalted, separated by reverse phase chromatography and analysed on a Waters Aquity UPLC system coupled to a Xevo TQ-S mass spectrometer.\n\nRaw data was acquired using MassLynx v 4.1 in multiple reaction monitoring mode. Raw files were processed using Skyline v 19. Protein-Peptide sequences were obtained from www.uniprot.org and settings optimised using custom synthesised peptides (Genscript USA). Normalised peak intensity data were exported to Microsoft Excel and data analysed using SIMCA v 15 (Umetrics, Sweden) for multivariate analysis and Graphpad prism v 6 was used for statistical analysis.\n\n\nResults\n\nMultivariate analysis of all inflammatory proteins measured in the control and SARS-CoV-2 positive patients are shown in Figure 1b and 1c. The score plot shows a clear separation of the positive and negative samples indicating the serum immune profile from people infected with SARS-CoV-2 is still significantly affected even 40 days post-infection. Figure 2 shows the univariate analysis of six proteins from our panel that were significantly altered. The majority of these proteins are either anti-inflammatory or associated with the stress response. Two proteins originate from the mitochondria, peroxiredoxin 3 (PRDX3) and carbamoyl phosphate synthase (CPS1). PRDX3 is a known antioxidant. Its increase in serum of patients infected with SARS-CoV-2 is likely indicative of continued mitochondrial stress response. CPS1 is a major mitochondrial urea cycle enzyme in hepatocytes. Serum CPS1 originates from the bile duct and is usually rapidly cleared by peripheral blood mononuclear cells6. It is possible that basal levels of CPS1 in serum are reduced in patients infected by SARS-CoV-2 due to increased circulation and activity of peripheral blood mononuclear cells.\n\nProteins significantly affected (p< 0.001) by non-parametric statistical analysis in the serum of >40 day post SARS-Cov-2 infected healthcare workers **** p<0.0001, ***p<0.001, **p<0.01, *p<0.01.\n\nN-Myc downstream regulated gene 1 (NDRG1) is a cytosolic protein with many biological functions7. Its role in the immune response is undefined but deficiency of NDRG1 affects the differentiation process of macrophages8 and maturation of mast cells9. Collagen triple helix repeat containing 1 (CTHRC1) is anti-inflammatory and promotes wound healing by recruiting M2 macrophages and regulating the TGF-β and Notch pathways10. This increase of CTHRC1 indicates tissue damage has occurred even in moderately affected patients.\n\nCystatin C is a protease inhibitor and extracellular levels are used as a biomarker for disease prognosis in cancer, cardiovascular disease, and inflammatory lung disorders11. In mice serum cystatin C is controlled by the anti-inflammatory cytokine IL10 of which increasing levels suppress cystatin C expression11. A longitudinal study looking at immune mediators show IL10 levels are significantly elevated in only severe cases of SARS-CoV-2 infection at four weeks post infection and are not affected at four weeks in mild cases12. This would corroborate with what we observe for cystatin C as the mild patients have increased cystatin C that is not being suppressed by higher IL10 levels. We also observe a slight reduction in serum progranulin. Progranulin plays a fundamental role in the immune response which is better defined within its role in neurodegenerative disorders13 but the relevance of serum progranulin is not fully understood. It appears to have a pro-inflammatory role in adipocytes in diabetes14 and an anti-inflammatory protective role in the vascular endothelium against inflammatory reactions15.\n\n\nConclusions\n\nRemarkably, even in patients who have suffered from an asymptomatic or mild SARS-CoV-2 infection, after 40 days post-infection they still exhibit a significantly raised group of biomarkers involved in inflammation and the stress response. This initial data using a custom designed inflammatory marker panel applied to mildly affected patients identifies potential drug targets, provides insight into the post infection inflammatory response. This approach using targeted proteomic technology has potential for application on further well-defined sample cohorts to understand what is abnormal about post infection inflammatory response in ‘long Covid’ patients.\n\n\nData availability\n\nProteomeXchange: Underlying mass spectrometry data on ProteomeXchange. Accession number PXD022159.\n\nUnderlying mass spectrometry data is also available on PanoramaWeb at https://panoramaweb.org/x1eZmn.url.",
"appendix": "Acknowledgments\n\nWe wish to thank Annabelle Lea Mai Immunology department Great Ormond Street Hospital for help with samples and the Peto Foundation for their continuing support.\n\nThis work is partly funded by the NIHR GOSH BRC. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.\n\n\nReferences\n\nNabavi N: Long covid: How to define it and how to manage it. BMJ. 2020; 370: m3489. PubMed Abstract | Publisher Full Text\n\nMarshall M: The lasting misery of coronavirus long-haulers. Nature. 2020; 585(7825): 339–341. PubMed Abstract | Publisher Full Text\n\nWandernoth P, Kriegsmann K, Groh-Mohanu C, et al.: Detection of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) by Mass Spectrometry. Viruses. 2020; 12(8): 849. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIhling C, Tänzler D, Hagemann S, et al.: Mass Spectrometric Identification of SARS-CoV-2 Proteins from Gargle Solution Samples of COVID-19 Patients. J Proteome Res. 2020; 19(11): 4389–4392. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMRM-LC-MS/MS Assay for inflammatrory associated proteins in serum. protocols.io. 2020. Publisher Full Text\n\nPark MJ, D'Alecy LG, Anderson MA, et al.: Constitutive release of CPS1 in bile and its role as a protective cytokine during acute liver injury. Proc Natl Acad Sci U S A. 2019; 116(18): 9125–9134. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMelotte V, Qu X, Ongenaert M, et al.: The N-myc downstream regulated gene (NDRG) family: diverse functions, multiple applications. FASEB J. 2010; 24(11): 4153–4166. PubMed Abstract | Publisher Full Text\n\nWatari K, Watari K, Watari K, et al.: Impaired differentiation of macrophage lineage cells attenuates bone remodeling and inflammatory angiogenesis in Ndrg1 deficient mice. Sci Rep. 2016; 6: 19470. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTaketomi Y, Sunaga K, Tanaka S, et al.: Impaired mast cell maturation and degranulation and attenuated allergic responses in Ndrg1-deficient mice. J Immunol. 2007; 178(11): 7042–7053. PubMed Abstract | Publisher Full Text\n\nQin S, Zheng JH, Xia ZH, et al.: CTHRC1 promotes wound repair by increasing M2 macrophages via regulating the TGF-β and notch pathways. Biomed Pharmacother. 2019; 113: 108594. PubMed Abstract | Publisher Full Text\n\nXu Y, Schnorrer P, Proietto A, et al.: IL-10 controls cystatin C synthesis and blood concentration in response to inflammation through regulation of IFN regulatory factor 8 expression. J Immunol. 2011; 186(6): 3666–3673. PubMed Abstract | Publisher Full Text\n\nZhao Y, Zhang P, Li K, et al.: Longitudinal COVID-19 profiling associates IL-1RA and IL-10 with disease severity and RANTES with mild disease. JCI Insight. 2020; 5(13): e139834. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaushter DH, Du H, Feng T, et al.: The lysosomal function of progranulin, a guardian against neurodegeneration. Acta neuropathologica. 2018; 136(1): 1–17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMatsubara T, Mita A, Minami K, et al.: PGRN is a key adipokine mediating high fat diet-induced insulin resistance and obesity through IL-6 in adipose tissue. Cell Metab. 2012; 15(1): 38–50. PubMed Abstract | Publisher Full Text\n\nHwang HJ, Jung TW, Hong HC, et al.: Progranulin protects vascular endothelium against atherosclerotic inflammatory reaction via Akt/eNOS and nuclear factor-κB pathways. PLoS One. 2013; 8(9): e76679. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "75166",
"date": "07 Dec 2020",
"name": "William J. griffiths",
"expertise": [
"Reviewer Expertise Mass spectrometry"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting study looking at a panel of inflammation-related protein signatures in a small cohort of patients 40+ days post infection. The authors see some changes in inflammatory proteins even this far after infection.\nThe methods and underlying data are available at open access repositories.\nI just have a few minor suggestions:\nAnalysis section - It would be good to read a few lines on how the data was normalised (i.e. normalisation of protein concentration). Results - I think the authors mean Figure 1a and 1b?\nIn summary, a nice brief study with some interesting results.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6212",
"date": "08 Jan 2021",
"name": "Wendy Heywood",
"role": "Author Response",
"response": "Many thanks for reviewing our brief report. Additional information about how proteins were normalised has been included and the figure reference has been corrected in the results."
}
]
},
{
"id": "75167",
"date": "16 Dec 2020",
"name": "Paul Skipp",
"expertise": [
"Reviewer Expertise Precision Medicine",
"Proteomics",
"Clinical Proteomics",
"Data Science",
"Biochemistry"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present valuable preliminary data investigating inflammatory signatures in patients previously infected with SARS-CoV2, 40 days after a positive diagnosis. Using a comprehensive mass spectrometry based multiplexed panel of 96 proteins associated with immune response, the study provides clear evidence of perturbed biochemical and inflammatory signature(s) 40 days after infection, even in asymptomatic patients. This provides important findings and a very useful methodology for future follow-up studies in Long COVID-19 cohorts.\nMinor suggestion for improvement of the manuscript:\nPlease provide a reference(s) supporting the statement,\"Work has been performed to characterise the inflammatory response to SARS-CoV2 infection in relation to disease severity\".\n\nUltimately, the mass spectrometry based MRM assay used in this study also only measures a defined panel of targets and hence pathways. In this light, I would suggest expanding or amending the statement, \"There are sets of simple but expensive immunoassay panels that are available to look at known key inflammatory proteins such as cytokine panels; however, these only give information on known pathways and limit discovery of novel or less defined inflammatory responses\".\n\nIt would be useful to have a couple of sentences to give the reader an overview of the pathways that the 96 protein panel targets (i.e. NF-kB, MAPK, JAK-STAT, etc).\n\nThe study provides a very useful and valuable methodology and although documented on protocols.io, a more comprehensive description in the main body of the manuscript would be useful.\n\nPlease clarify Figure 1. Figure 1C does not exist. Was this figure a loading plot to show how strongly each characteristic influences a principal component?\n\nIn the manuscript the authors state the following: \"Figure 2 shows the univariate analysis of six proteins from our panel that were significantly altered\". Please can you clarify in the text whether these were the only six proteins that were significantly altered or a selected subset.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6214",
"date": "08 Jan 2021",
"name": "Wendy Heywood",
"role": "Author Response",
"response": "Many thanks for taking the time to review our brief report. We have made all the suggested corrections including rewording and expanding of some comments as well as added additional reference. We have tried to improve the method description but the journal requirements require the methods section to be as brief as possible."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1349
|
https://f1000research.com/articles/10-9/v1
|
08 Jan 21
|
{
"type": "Case Report",
"title": "Case Report and Literature Review: COVID-19 and status epilepticus in Dyke-Davidoff-Masson syndrome",
"authors": [
"Lourdes de Fátima Ibañez Valdés",
"Jerry Geroge",
"Sibi Joseph",
"Mohamed Alshmandi",
"Wendy Makaleni",
"Humberto Foyaca Sibat",
"Lourdes de Fátima Ibañez Valdés",
"Jerry Geroge",
"Sibi Joseph",
"Mohamed Alshmandi",
"Wendy Makaleni"
],
"abstract": "Dyke-Davidoff-Masson syndrome (DMMS) is a non-inherited rare condition with a clinical constellation of hemiparesis/hemiplegia, facial asymmetry, intellectual disability, and epilepsy. The radiological features can be including unilateral cerebral atrophy, calvarial thickening, and hyper pneumatization of the paranasal sinuses. The condition can either be congenital or acquired. The presentation usually occurs during childhood or early adolescents, but there have been adult cases reported. Here we report a 48-year-old male who was a known poorly controlled epileptic that contracted SARS-CoV-2 with subsequently developed status epilepticus and, when worked up, was shown to have features of DDMS. This case is unique as the patient had hemiatrophy and epilepsy but managed to lead a normal, physically demanding, and high functioning academic career and presented late in life. Perhaps only due to coronavirus disease 2019 (COVID-19) was this diagnosis picked up. This report contains a case presenting atypical DDMS in status epilepticus and COVID -19 plus other complications. From our knowledge, this is the first case presenting these comorbidities reported to the medical literature.",
"keywords": [
"Covid-19",
"Dyke-Davidoff-Mason Syndrome",
"status epilepticus",
"hyperglycaemic hyperosmolar syndrome"
],
"content": "Introduction\n\nIn 1580, the first respiratory pandemic was reported1. Up to date, millions of people died, most of them during the 20th and 21st centuries. The most devastating epidemic and outbreaks were the Spanish Flu (500 million infected) during the early 20th century, even bigger than Hong Kong flu, swine flu, SARS-CoV-1 (2003), and the MERS-CoV outbreak (2012)1–3. However, the first pandemic causing encephalitis was reported soon after 15803.\n\nSince 1965, when human coronaviruses were discovered4, several types of coronavirus (CoV) have been reported, including SARS-CoV-2, SARS-CoV-1, and MERS-CoV, which are all responsible for three epidemics, plus others four types that also infect many human beings (HCoV-229E, HCoV-OC43, HCoV-NL63, and HCoV-HKU1)5. Based on several mechanisms, coronaviruses affect the peripheral and the central nervous system. Even before SARS-CoV-2, other types of coronavirus, such as SARS-CoV-1, HCoV-229E, and HCoV-OC43, also damage the nervous system5.\n\nWuhan is a large city and the capital of Hubei Province in the People's Republic of China. Wuhan has a population of around 11 million persons. At the beginning of December 2019, an outbreak of many persons presenting viral pneumonia of an unknown agent was reported. In the following month (January 7, 2020), some Chinese authors identified the etiological agent of that respiratory disease and called it by 2019-nCoV (for 2019 novel coronavirus)6–8.\n\nSince December 2019, documented an increasing number of cases presenting the novel coronavirus disease of 2019 (nCOVID-19) and associated neurological manifestation are published every month. CoV also caused neurological lesions like anosmia and ageusia with different prevalence in China (5%)6 or in Italy (88%)9.\n\nIn a recent study by Dorche et al., the following list of neurological complications were observed: headache and dizziness (the most common on initial presentation), fatal encephalitis with HCoV-OC43 (two immunosuppressed infants), acute disseminated encephalomyelitis (one 15-year-old boy with HCoV-OC43 and four adults with SARS-CoV-2), acute flaccid paralysis (HCoV 229E) and OC43 (one 3-year-old girl), ischemic (1.3%) and hemorrhage (0.5%) strokes, encephalitis with SARS-CoV-1 RNA (one 39-year-old patient), different presentation of Guillain-Barré syndrome, cerebral venous sinus thrombosis (13 patients in nine studies), acute encephalomyelitis (four patients), acute myelitis (five patients), optic neuritis (one patient) altered level of consciousness (nonconvulsive status epilepticus, infections, parenchymal lesions, electrolyte disturbances, hypoxic, toxic and metabolic encephalopathies), leukoencephalopathy (18 patients in three studies), acute necrotizing encephalopathy (eight patients), other encephalitis (22 patients out of 13 reviews), mild encephalitis/encephalopathy with a reversible splenial lesion(MERS), posterior reversible encephalopathy syndrome (PRES), and Bickerstaff's encephalitis (BBE),10–14. In patients infected by SARS-CoV-1 and SARS-CoV-2, epileptic seizures have been reported13 also in infected patients with MERS-CoV12 and SARS-CoV-2. In COVID-19, (48 epileptic patients out of 20 studies), visual impairments (12 patients out of 3 reviews), impaired eye movement mainly due to Abducens nerve palsy (12 patients out of 4 reviews), trigeminal neuropathy (in 9 patients out of 2 studies), Miller-Fisher syndrome (52 patients out of 36 studies), skeletal muscle injury and muscular diseases Have been reported10.\n\nNepal and colleagues15 reported one case of Bell's palsy, as a neurological presentation of COVID-19. Another group of authors published two new patient cases but did not include enough supporting information to draw firm conclusions16. At the same time, other authors have published cases presenting generalized epileptic seizures17–22.\n\nOne case of focal status epilepticus (SE) was reported by Vollono et al.23 and acute epileptic encephalopathy by others24–26, including the treatment for these conditions27.\n\nA systematic review done by Ghannam et al. found two cases of SE, one of which had a past medical history of epilepsy from another cause28. Gelisse et al. established that some patients with severe SARS-CoV-2 infection are at risk of subclinical epileptic seizures or even nonconvulsive status epilepticus (NCSE) and recommend video EEG monitoring in some cases29.\n\nRecently, some authors have speculated that acute epileptic seizures may be due to swelling of the brain cortex (encephalitis) and the direct damage of the brain cortex by the virus because SARS-CoV-2 can be present in the cerebrospinal fluid (CSF) of some patients18,30,31.\n\nIn other extensive studies involving several hundreds of COVID-19 patients, the authors concluded that none of their cases had acute symptomatic seizures or SE20,32–42.\n\nNevertheless, the retrospective case series published by Somani and collaborators25 deserves special mention. These investigators published the electroencephalographic findings and clinical manifestations of two COVID-19 patients with new-onset SE without a previous history of epilepsy or acute epileptic seizures. Both patients had SARS-CoV-2 pneumonia confirmed by CT scan and PCR; however, the authors did not perform CSF and could not rule out meningoencephalitis. The second patient presented a new-onset refractory status epilepticus. The same author established the neurovirulence of SARS-CoV-1, finding the presence of viral antigen in the thalami, hippocampus, medulla oblongata, and mesencephalic regions that regulate cardiorespiratory functions in a human autopsy series25. Some recent good news is the excellent response of SE to levetiracetam reported by two investigators24,25.\n\nOther investigators also recommend the use of verapamil in patients presenting SE stage III and SARS-Cov-2 infection43. The same authors reported the first patient affected by PRES and SARS-CoV-2 without SE. In contrast, Mohammad et al. wrote about a 32-year-old male with tonic-clonic generalized SE44. In the meantime, other investigators delivered essential recommendations to improve the management of SE during the pandemic despite the lack of ventilators and ICU facilities45.\n\nAcquired or congenital (infantile) cerebral hemiatrophy, otherwise referred to as Dyke-Davidoff-Masson syndrome (DDMS), was first described in 1933 by Dyke and colleagues46–48. DDMS is a non-inherited rare condition49, with an unknown frequency; most of the literature stems from either case reports or series50. DDMS is a diagnostic constellation made up of hemiparesis/hemiplegia, facial asymmetry, intellectual disability, and treatment-resistant epilepsy, classically with distinct neuroimaging features48. However, according to Ayaz et al., the syndrome has varied clinical and radiological spectrum presenting at different life stages51. The classical imaging findings are hypoplasia of one brain hemisphere (hemiatrophy), often accompanied by volume reduction of corresponding cranial fossa and thickening of nearby bony structures and equilateral enlargement paranasal sinuses, the frontal sinus being the most involved or hyperpneumotisation of mastoid air cells. The congenital type can be due to insults suffered during fetal or early childhood development, such as ischemia, trauma, infarction, hemorrhage, and infections. However, the acquired type is usually associated with trauma, infectious diseases, or hemorrhages after one month of age47. We know that hemispherectomy is the best treatment for patients who have drug-resistant and disabling seizures.\n\nAt the time of writing, the coronavirus disease-19 (COVID-19) pandemic continues infecting peoples worldwide. COVID-19, caused by SARS-CoV-2, has thus far claimed 23,057,288 cases worldwide52 and 607 045 patients in South Africa53. Up to date, 46 medical doctors died in the Eastern Cape province alone.\n\nWe performed an extensive search of the medical literature to answer our research question: \"What is the reported frequency of status epilepticus in patients with DDMS and coronavirus infections?\n\n\nCase presentation\n\nA 48-year-old African male patient was admitted to Nelson Mandela Academic Central Hospital (NMACH) in Mthatha, South Africa. He was born out of a non-consanguineous marriage and was referred from a regional hospital with tonic-clonic-generalized status epilepticus. On initial presentation to the base hospital, he was given diazepam 10 mg IV stat (dose repeated twice) and then loaded with phenytoin 750 mg.\n\nThis patient had a past medical history of chronic epilepsy for many years but was well-controlled on valproate acid CR 500 mg PO Bd, levetiracetam 750 mg PO BD per day. There was no facial asymmetry, no hemiplegia, the rest of the cognitive functions were average, and there are no mental retardation signs.\n\nHe was also a chronic hypertensive. He worked as a police officer on further inquiry and he did not smoke, consume drugs or alcohol. We did not obtain remarkable information on birth history, developmental milestones, education history, prior admissions to hospital, and childhood illnesses.\n\nWe found no noticeable body asymmetry on examination. The patient had pink mucous membranes, was well hydrated, and afebrile with a GCS 11/15 (E3V3M5); his motor examination revealed a power 3/5 with spastic hypertonia on left upper and lower limbs, and no fits noted.\n\nHe was in respiratory distress with tachypnea of 30 breaths/minute saturating at 84% on a 40% venture face mask. The rest of the vital signs showed a BP 118/88 mmHg and Pulse 98 bpm. He had scattered crepitation on the chest bilaterally. Table 1 shows all blood test results.\n\nSerum levels of interleukine-6 were not available. PCR confirmed SARS-CoV infection, but we did not perform ferritin and procalcitonin investigations.\n\nThe patient presented with a one-day SE type II (established), characterized by recurrent tonic-clonic generalized seizures with impaired awareness. As he did not recover, progressive doses of anti-seizure medication (ASM) were administered, reaching a total of 20 mg of diazepam (10 mg IV twice), 1500 mg of phenytoin (IV bolus), and 1500 mg of valproate (25 mg/kg), without recovering. An urgent cranial CT scan of the brain revealed atrophy on the right cerebral hemisphere with associated thickening of the calvarium on the same side without hyperpneumotisation of paranasal sinuses or mastoid air cells (Figure 1, Figure 2 and Figure 3), suggestive of Dyke-Davidoff-Masson Syndrome; otherwise, there was no bleed or area of infarct, and there was no space-occupying lesion.\n\nShows a notable atrophy of the right cerebral hemisphere with enlargement of the ipsilateral lateral ventricle.\n\nShows asymmetry of the lateral ventricles (right to left) with a notable atrophy of the right cerebral hemisphere.\n\nShows a marked thickness on the right side of the skull.\n\nOn the second day of admission, the patient was admitted to the COVID ward, put on high-flow nasal oxygen (60% at 15 L/min), dexamethasone 8 mg IV daily, Clexane 60 mg SC 12-hourly, ceftriaxone 1 g IV every day, azithromycin 500 mg Po daily, vitamin D 50000 U PO weekly, vitamin C 250 mg PO 8-hourly, diazepam 10 mg IV if fitting, valproate 500 mg IV 12-hourly, phenytoin 100 mg IV 8-hourly, amlodipine 10 mg orally daily, Ridaq 25 mg orally daily and intravenous fluids (1 L Ringers lactate IV 8-hourly).\n\nAfter two days of admission, the patient improved neurologically and presented no more seizures, but his respiratory distress continued progressively getting worse, and his septic markers were rising. Two days later, blood levels showed values of Na 159 mmol/L, K 5.1 mmol/L, urea 42 mmol/L, creatinine 275 µmol/L, CRP 117 mg/dL, T protein 88 g/L, Alb 38 g/L, ALT 48 U/L, AST 168 U/L, GGT 190 U/L, ALP 58U/L, white cell count 14.30 × 109/L, Hb 17.1g/dl, platelet count 316 × 109/L cholesterol level no done, Blood gas showed Ph. 7.46, PaCO2 36 mmHG, PaO2 66mmHg, HCO3 27mmol/L, Na 160 mmol/L, K 3.5 mmol/L, Ca 1.05 mmol/L, Hgt 25.6 mmol/L, blood oxygen saturation 84%. The patient's urine did not contain ketone bodies.\n\nOn the fifth day after admission, the patient was assessed as having ARDS secondary to COVID-19 and hyperglycemic hyperosmolar state (HHS), and high-flow nasal O2 was increased to 100% concentration at 20 L/min. One and a half hours after the onset of the symptoms, the patient had not recovered yet. The patient began to fit again and under the suspicion of refractory status epilepticus secondary to HHS and neuro-COVID 19; when another round of 20 mg of diazepam (10 mg IV twice), 1500 mg of phenytoin (IV bolus), and 1500 mg of valproate (25 mg/kg) was started, the patient developed cardiac arrest and demised.\n\n\nDiscussion and literature review\n\nOur literature review utilized the Preferred Reporting Items for Systemic review and Meta-Analysis statement. However, we did not conduct a classical systematic review.\n\nWe reviewed the databases published before August 20, 2020, such as Medline EMBASE, Scopus online databases, Google Scholar, to identify articles evaluating COVID-19 and SE in DDMS. All items about \"neurologic complications* OR epilepsy* OR brain* OR status epilepticus* OR fits* OR neuronal lesion* OR Neuro-Covid* OR cortical lesions* OR DDMS OR * OR seizure* OR COVID-19* OR unconsciousness* OR acute epileptic seizure*, OR Duke Davidoff Mason Syndrome*\" where * is the PubMed wildcard for every possible word beginning or ending. Other neurological combinations were considered beyond the scope of the current work and no included. Finally, we did not find a publication related to COVID-19, SE and DDMS.\n\nOur patient complained of chronic arterial hypertension, and this condition and diabetes mellitus is associated with a significant risk of lung disease leading to COVID-19 severity. Despite the patient's condition, antihypertensive therapy should continue in COVID-19 patients54. Concerning our patient, it's important to highlight that diabetes mellitus by itself is one of the most relevant comorbidities associated with the severity of all coronavirus infections, including the current SARS-CoV-2, and affected cases have an increased risk to develop severe complications such as acute respiratory distress syndrome and systemic organ failure55. COVID-19 patients with hyperglycemia are at risk of developing other infections, including influenza and pneumonia with increasing mortality rate; this is also applicable to other SARS coronavirus, pandemic influenza A 2009 (H1N1), and middle east respiratory syndrome coronavirus56–59.\n\nHere we discuss elevated urea and creatinine in our patient. Therefore, it is essential to mention that apart from diabetes and hypertension, acute kidney injury has also been documented in some patients with COVID-19. ACE2 gene expression in renal cells and bladders cells has been investigated, and the results confirmed damage of the renal proximal tubule cells and the bladder epithelial cells by COVID-19 infection60. SARS-CoV-2 affects the kidneys61, which has been confirmed by examining viral nucleocapsid protein accumulated in the renal tubules by post-mortem examination proved that62.\n\nDDMS is due to atrophy of one cerebral hemisphere and usually occurs due to an insult to the brain in utero or an early period of childhood63. In the first description, Dyke, Davidoff, and Masson described nine patients who had a constellation of seizures, facial asymmetry, mental retardation, and hemiparesis with bare skull X-ray changes (ipsilateral osseous hypertrophy and calvarial thickening)64. We can understand intracranial pathology with MRI and CT scans, which results in such clinical presentation. Our patient is atypical because he did not complain of weakness on the left half body, was strong enough to work as a police officer, and there was no evidence of mental retardation was observed.\n\nSome patients with DDMS can complain of psychiatric manifestations in rare instances65. The radiological features can include unilateral cerebral atrophy, calvarial thickening, and hyperpneumotisation of the paranasal sinuses66. As with our patient, mental retardation does not always need to be present, and the seizures may develop years after the initial insult67.\n\nSome authors classify DDMS as either congenital/primary or acquired/secondary. The congenital form occurs due to an insult that happens in utero. It could be infections or vascular disorders occurring during the gestational period (unilateral cerebral artery pathologies or mid aortic arch coarctation)68. The congenital forms usually present during the perinatal period. The acquired form occurs due to early childhood infections, trauma, tumors, asphyxia, intracranial ischemia, or hemorrhage.\n\nTo better understand the anatomical changes occurring, it is crucial to understand the brain's growth and surrounding structure. The mail sulci form around 3 months’ gestation up to approximately eight months of pregnancy. If there are no prominent sulci visible on imaging, the congenital form of DDMS is present. Most brain and skull development occurs during the first three years of life (reaching 75% of adult size). The outward pressure of the brain parenchyma on the skull contributes to this growth. If there is unilateral atrophy, then the surrounding structures will grow inwards (calvarial thickening, enlarged sinuses, increased width of diploid spaces69,70. Hangmen et al. proposed that the congenital form of DDMS be named unilateral cerebral hypoplasia because there is hypoplasia instead of atrophy71.\n\nA literature review done by Unal et al. showed that in the pediatric presentations, there is a male predisposition towards DDMS, and the left hemisphere is more commonly affected; the mean age at diagnosis was 11 in this review72. However, a literature review done by Diestro et al. in 2018 comprising 21 patients with a mean age at presentation being 31 years old showed a slight female despondence, and these adult presentations more commonly involving the right cerebral hemisphere. In 28% of cases (6/21), there was no mental retardation, and in 14% (3/21), it was unknown whether there was mental retardation73.\n\nThe association of SARS-CoV and seizures was known even before the current pandemic. In 2003 authors published the first observed case74. The following year, other patients were reported75, and later another case was associated with a different coronavirus76. However, the total number of reported cases is small. SE and encephalopathy have been reported in children as a presentation of COVID-19. The mechanism of production of seizures is also known77. These authors77,78 proposed that epileptic seizures can be due to several mechanisms, such as direct infection of the virus, a post-infectious mechanism, an autoimmune response, hematogenous pathway and thrombosis79, by dysregulated cytokine storm79, and by the retrograde neural way, hypoxia, and via the ACE-2 enzyme80.\n\nAssociation between DDMS and SE is hugely uncommon, and before the current COVID-19 pandemic, only three adolescent patients have been reported: one in 201581 and the other two in 201882. Other recent systematic reviews, case series, and case reports did not mention any association of DDMS and SE51,63,73,83–90. At the time of writing (August 25, 2020), no patients presenting DDMS and SE infected by COVID-19 have been reported to the medical literature.\n\nDifferential diagnoses, such as Silver-Russell syndrome, basal ganglia germinoma, neurofibromatosis, Parry-Romberg Syndrome, Sturge-Weber syndrome, Rasmussen encephalitis, Fishman syndrome, linear nevus syndrome, and Rasmussen encephalitis, should be considered during the management of these patients.\n\nFinally, we want to highlight some precautions to be considered when treating patients in SE and COVID-19. As aforementioned, DDMS causes epilepsy, epileptic seizures, and even SE. On the other hand, COVID-19 can also cause epileptic seizures and SE. However, the use of some ASM and anti-COVID medicines may cause complications. Therefore, some medications should be used with caution. For example, lacosamide is recommended for the adjunctive treatment of partial-onset seizures, diabetic neuropathic pain, and to control attacks in refractory SE. However, it can prolong the PR interval on an electrocardiogram. Hydroxychloroquine extends the QT interval91. Elongation of the QT interval can also be caused by azithromycin, phenytoin, carbamazepine, and rufinamide, leading to cardiac conduction disturbances13. Combining ASM with hydroxychloroquine and azithromycin can be harmful. Therefore, we recommend EKG monitoring.\n\n\nConclusion\n\nIn our opinion, the SE in our patient had a multifactorial origin, including HHS and atypical DDMS. This could have created a susceptible environment in which the new coronavirus's disease acted as a SE trigger. These hypotheses make this case a unique report.\n\nTo our knowledge, this patient is the first case of SARS-CoV-2 infection leading to TCG-SE type 2 on a DDMS patient published in the medical literature.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nConsent\n\nWritten informed consent for publication of their clinical details and clinical images was obtained from the relatives of the patient.",
"appendix": "Author contributions\n\n\n\nStudy concept and design: HFS, LFIV, and JG Data collection: WM, JG, and MS Analysis and interpretation of the references: LFIV. Drafting of the manuscript: LFIV, HFS. Revising the manuscript: SJ, MS, WM, and LFIV. Supervision of research and manuscript writing process: HFS and LFIV. All named authors meet the International Committee of Medical Journal Editors (ICMJE) criteria for authorship for this article, take responsibility for the work's integrity, and have given their approval for this version to be published.\n\n\nAcknowledgments\n\nSpecial thanks to Dr. A Anwary from the Department of Radiology, NMACH. Mthatha, South Africa, for his report on images.\n\n\nReferences\n\nMartin PMV, Martin-Granel E: 2,500-year evolution of the term epidemic. Emerg Infect Dis. 2006; 12(6): 976–980. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTroyer EA, Kohn JN, Hong S: Are we facing a crashing wave of neuropsychiatric sequelae of COVID-19. Neuropsychiatric symptoms and potential immunologic mechanisms. Brain Behav Immun. 2020; 87: 34–39. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCunha BA: Influenza: historical aspects of epidemics and pandemics. Infect Dis Clin North Am. 2004; 18(1): 141–155. PubMed Abstract | Publisher Full Text\n\nKahn JS, McIntosh K: History and recent advances in coronavirus discovery. Pediatr Infect Dis J. 2005; 24(11 Suppl): S223– S227. PubMed Abstract | Publisher Full Text\n\nZhu N, Zhang D, Wang W, et al.: A novel coronavirus from patients with pneumonia in China, 2019. N Engl J Med. 2020; 382(8): 727–733. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchwartz DA, Graham AL: Potential Maternal and Infant Outcomes from Coronavirus 2019-nCoV (SARS-CoV-2) Infecting Pregnant Women: Lessons from SARS, MERS, and Other Human Coronavirus Infections. Viruses. 2020; 12(2): 194. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMa K, chen T, Han MF: [Management and clinical thinking of Coronavirus Disease 2019]. Zhonghua Gan Zang Bing Za Zhi. 2020; 28(0): E002. PubMed Abstract | Publisher Full Text\n\nMao L, Jin H, Wang M, et al.: Neurologic Manifestations of Hospitalized Patients With Coronavirus Disease 2019 in Wuhan, China. JAMA Neurol. 2020; 77(6): 683–690. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLechien JR, Chiesa-Estomba CM, Cabaraux P, et al.: Features of Mild-to-Moderate COVID-19 Patients with Dysphonia. J Voice. 2020; S0892-1997(20)30183-1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDorche MS, Huot Ph, Osherov M, et al.: Neurological complications of coronavirus infection; a comparative review and lessons learned during the COVID-19 pandemic. J Neurol Sci. 2020; 417: 117085. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPinna P, Grewal P, Hall JP, et al.: Neurological manifestations, and COVID-19: Experiences from a tertiary care center at the Frontline. J Neurol Sci. 2020; 415: 116969. PubMed Abstract | Publisher Full Text\n\nJain R, Young M, Dogra S, et al.: COVID-19 related neuroimaging findings: A signal of thromboembolic complications a strong prognostic marker of poor patient outcome. J Neurol Sci. 2020; 414: 116923. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAsadi-Pooya AA: Seizures associated with coronavirus infections. Seizure. 2020; 79: 49–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen T, Wu D, Chen H: Clinical characteristics of 113 deceased patients with coronavirus disease 2019: a retrospective study. BMJ. 2020; 368: m1091. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNepal G, Rehrig JH, Shrestha GS, et al.: Neurological manifestations of COVID-19: a systematic review. Crit Care. 2020; 24(1): 421. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPinzon RT, Wijaya VO, Buana RB, et al.: Neurologic Characteristics in Coronavirus Disease 2019 (COVID-19): A Systematic Review and Meta-Analysis. Front Neurol. 2020; 11: 565. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHao X, Zhou D, Li Z, et al.: Severe psychological distress among epilepsy patients during the COVID-19 outbreak in southwest China. Epilepsia. 2020; 61(6): 1166–1173. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoriguchi T, Harii N, Goto J, et al.: A first case of meningitis/encephalitis associated with SARS-Coronavirus-2. Int J Infect Dis. 2020; 94: 55–58. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDuong L, Xu P, Liu A: Meningoencephalitis without respiratory failure in a young female patient with COVID-19 infection in downtown Los Angeles, early April 2020. Brain Behav Immun. 2020; 87: 33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLu L, Xiong W, Liu D, et al.: New-onset acute symptomatic seizure and risk factors in coronavirus disease 2019: a retrospective multicenter study. Epilepsia. 2020; 61(6): e49–e53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFasano A, Cavallieri F, Canali E, et al.: First motor seizure as presenting symptom of SARS-CoV-2 infection. Neurol Sci. 2020; 41(7): 1651–1653. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeonardi M, Padovani A, McArthur JC: Neurological manifestations associated with COVID-19: a review and a call for action. J Neurol. 2020; 267(6): 1573–1576. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVollono C, Rollo E, Romozzi M, et al.: Focal status epilepticus as a unique clinical feature of COVID-19: a case report. Seizure. 2020; 78: 109–112. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBalloy G, Leclair-Visonneau L, Péréon Y, et al.: Non-lesional status epilepticus in a patient with coronavirus disease 2019. Clin Neurophysiol. 2020 ; 131(8): 2059–2061. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSomani S, Pati S, Gaston T, et al.: De Novo status epilepticus in patients with COVID-19. Ann Clin Transl Neurol. 2020; 7(7): 1240–1244. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKatal S, Balakrishnan S, Gholamrezanezhad A: Neuroimaging and neurologic findings in COVID-19 and other coronavirus infections: A systematic review in 116 patients. J Neuroradiol. 2020; S0150-9861(20)30204-2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOrsucci D, Ienco EC, Nocita G, et al.: Neurological features of COVID-19 and their treatment: a review. Drugs Context. 2020; 9: 2020-5-1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGhannam M, Alshaer Q, Al-Chalabi M, et al.: Neurological involvement of coronavirus disease 2019: a systematic review. J Neurol. 2020; 267(11): 3135–3153. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGelisse P, Rossetti AO, Genton P, et al.: How to carry out and interpret EEG recordings in COVID-19 patients in ICU. Clin Neurophysiol. 2020; 131( 8): 2023–2031. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNiazkar HR, Zibaee B, Nasimi A, et al.: The neurological manifestations of COVID-19: a review article. Neurol Sci. 2020; 41(7): 1667–1671. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen X, Laurent S, Onur OA, et al.: A systematic review of neurological symptoms and complications of COVID-19. J Neurol. 2020; 1–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang L, Shen Y, Li M, et al.: Clinical manifestations and evidence of neurological involvement in 2019 novel coronavirus SARS-CoV-2: a systematic review and meta-analysis. J Neurol. 2020; 267(10): 2777–2789. PubMed Abstract | Publisher Full Text | Free Full Text\n\nButala N: Neurological Aspects of Coronavirus Infectious Disease 2019 (COVID-19). Innov Clin Neurosci. 2020; 17(4–6): 13–15. PubMed Abstract | Free Full Text\n\nGarcía-Azorín D, Martínez-Pías E, Trigo J, et al.: Neurological Comorbidity Is a Predictor of Death in Covid-19 Disease: A Cohort Study on 576 Patients. Front Neurol. 2020; 11: 781. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBridwell R, Long B, Gottlieb M: Neurologic complications of COVID-19. Am J Emerg Med. 2020; 38(7): 1549.e3–1549.e7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlomari SO, Abou-Mrad Z, Bydon A: COVID-19 and the Central Nervous System. Clin Neurol Neurosurg. 2020; 198: 106116. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYavarpour-Bali H, Ghasemi-Kasman M: Update on neurological manifestations of COVID-19. Life Sci. 2020; 257: 118063. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNeedham EJ, Chou SHY, Coles AJ, et al.: Neurological Implications of COVID-19 Infections. Neurocrit Care. 2020; 32(3): 667–671. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSheraton M, Deo N, Kashyap R, et al.: A Review of Neurological Complications of COVID-19. Cureus. 2020; 12(5): e8192. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMao L, Jin H, Wang M, et al.: Neurologic manifestations of hospitalized patients with coronavirus disease 2019 in Wuhan, China. JAMA Neurol. 2020; 77(6): 683–690. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEllul MA, Benjamin L, Singh B, et al.: Neurological associations of COVID-19. Lancet Neurol. 2020; 19(9): 767–783. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGarg RK: Spectrum of Neurological Manifestations in Covid-19: A Review. Neurol India. 2020; 68(3): 560–572. PubMed Abstract | Publisher Full Text\n\nGómez-Enjuto S, Hernando-Requejo V, Lapeña-Motilva J, et al.: Verapamil as treatment for refractory status epilepticus secondary to PRES syndrome on a SARS-Cov-2 infected patient. Seizure. 2020; 80: 157–158. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAbdulsalam MA, Abdulsalam AJ, Shehab D: Generalized status epilepticus as a possible manifestation of COVID-19. Acta Neurol Scand. 2020; 142(4): 297–298. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKinney MO, Brigo F, Kaplan PW: Optimizing status epilepticus care during the COVID-19 pandemic. Epilepsy Behav. 2020; 109: 107124. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDyke C, Davidoff L, Masson C: Cerebral hemiatrophy & homolateral atrophy of the skull and sinuses. Surg Gynecol Obs. 1932; 57: 588–600.\n\nDilber B, Sahin S, Eyüboğlu I: Two Different Manifestations of Neonatal Vascular Injury: Dyke-Davidoff-Masson Syndrome and Crossed Cerebellar Atrophy. J Stroke Cerebrovasc Dis. 2020; 29(3): 104600. PubMed Abstract | Publisher Full Text\n\nAdebayo PB, Bakare A, Bello MM, et al.: Dyke-Davidoff-Masson syndrome in a Nigerian. Epilepsy Behav Case Rep. 2016; 7: 10–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKalaskar R, Kalaskar AR: Classical oral manifestations of Dyke-Davidoff-Masson syndrome: A case report with review of the literature. J Korean Assoc Oral Maxillofac Surg. 2018; 44(4): 198–203. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRashid AMA, Md Noh MSF: Dyke-Davidoff-Masson syndrome: a case report. BMC Neurol. 2018 [cited 2020 Aug 15]; 18(1): 76. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBjelan M, Kozic D, Njagulj V, et al.: DYKE-DAVIDOFF-MASSON SYNDROME - TYPICAL IMAGING FEATURES. Med Pregl. 2016; 69(11–12): 373–5. PubMed Abstract | Publisher Full Text\n\nCoronavirus disease (COVID-19). [cited 2020 Aug 25]. Reference Source\n\nCOVID-19 South African coronavirus news and information. [cited 2020, August 20]. Reference Source\n\nSanchis-Gomar F, Lavie CJ, Perez-Quilis C, et al.: Angiotensin-Converting Enzyme 2 and Antihypertensives (Angiotensin Receptor Blockers and Angiotensin-Converting Enzyme Inhibitors) in Coronavirus Disease 2019. Mayo Clin Proc. 2020; 95(6): 1222–1230. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBornstein SR, Rubino F, Khunti K, et al.: Practical recommendations for the management of diabetes in patients with COVID-19. Lancet Diabetes Endocrinol. 2020; 8(6): 546–550. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGupta R, Ghosh A, Singh AK, et al.: Clinical considerations for patients with diabetes in times of COVID-19 epidemic. Diabetes Metab Syndr. 2020; 14(3): 211–212. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang JK, Feng Y, Yuan MY, et al.: Plasma glucose levels and diabetes are independent predictors for mortality and morbidity in patients with SARS. Diabet Med. 2006; 23(6): 623–628. PubMed Abstract | Publisher Full Text\n\nSchoen K, Horvat N, Guerreiro NFC, et al.: Spectrum of clinical and radiographic findings in patients diagnosed with H1N1 and correlation with clinical severity. BMC Infect Dis. 2019; 19(1): 964. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSong Z, Xu Y, Bao L, et al.: From SARS to MERS, thrusting coronaviruses into the spotlight. Viruses. 2019; 11(1): 59. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLin W, Hu L, Zhang Y, et al.: Single-cell analysis of ACE2 expression in human kidneys and bladders reveals a potential route of 2019-nCoV infection. bioRxiv. 2020; (Reviewed on August 23, 2020). Publisher Full Text\n\nDiao B, Feng Z, Wang C, et al.: Human kidney is a target for novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. medRxiv. 2020. Publisher Full Text\n\nPerico L, Benigni A, Remuzzi G: Should COVID-19 concern nephrologists? Why and to what extent? The emerging impasse of angiotensin blockade. Nephron. 2020; 144(5): 213–221. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJilowa CS, Meena PS, Rohilla J, et al.: Dyke-Davidoff-Masson syndrome. Neurol India. 2017; 65(2): 413–4. PubMed Abstract | Publisher Full Text\n\nDyke CG, Davidoff LM, Masson CB: Cerebral hemiatrophy and homolateral hypertrophy of the skull and sinuses. Surg Gynecol Obstet. 1933; 57: 588–600.\n\nAmann B, Garcia de la Iglesia C, Mckenna P, et al.: Treatment-refractory schizoaffective disorder in a patient with dyke-davidoff-masson syndrome. CNS Spectr. 2009; 14(1): 36–9. PubMed Abstract | Publisher Full Text\n\nPendse NA, Bapna P, Menghani V, et al.: Dyke-Davidoff-Masson syndrome (DDMS). Indian J Pediatr. 2004; 71(10): 943. PubMed Abstract | Publisher Full Text\n\nSener RN, Jinkins JR: MR of craniocerebral hemiatrophy. Clin Imaging. 1992; 16(2): 93–7. PubMed Abstract | Publisher Full Text\n\nGraham A, Molnar Z: Development of the nervous system. In Standring S (ed): Gray's Anatomy, ed 40. London, Churchill Livingstone, Elsevier. 2008; 385.\n\nSolomon GE, Hilal SK, Gold AP, et al.: Natural history of acute hemiplegia of childhood. Brain. 1970; 93(1): 107–120. PubMed Abstract | Publisher Full Text\n\nSharma S, Goyal D, Negi A, et al.: Dyke–Davidoff–Masson syndrome. Indian J Radiol Imaging. 2006; 16(2): 165–6. Publisher Full Text\n\nHageman G, Gooskens RH, Willemse J: A cerebral cause of arthrogryposis: Unilateral cerebral hypoplasia. Clin Neurol Neurosurg. 1985; 87(2): 119–22. PubMed Abstract | Publisher Full Text\n\nUnal O, Tombul T, Cirak B, et al.: Left hemisphere and male sex dominance of cerebral hemiatrophy (Dyke-Davidoff-Masson Syndrome). Clin Imaging. 2004; 28(3): 163–5. PubMed Abstract | Publisher Full Text\n\nDiestro JDB, Dorotan MKC, Camacho AC, et al.: Clinical spectrum of Dyke-Davidoff-Masson syndrome in the adult: an atypical presentation and review of literature. BMJ Case Rep. 2018; 2018: bcr2018224170. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHung ECW, Chim SSC, Chan PKS, et al.: Detection of SARS coronavirus RNA in the cerebrospinal fluid of a patient with severe acute respiratory syndrome. Clin Chem. 2003; 49(12): 2108–2109. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLau KK, Chu CM, Lau ST, et al.: Possible central nervous system infection by SARS coronavirus. Emerg Infect Dis. 2004; 10(2): 342–344. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDominguez SR, Robinson CC, Holmes KV: Detection of four human coronaviruses in respiratory infections in children: a one-year study in Colorado. J Med Virol. 2009; 81(9): 1597–1604. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSwarz JA, Daily S, Niemi E, et al.: COVID-19 Infection Presenting as Acute-Onset Focal Status Epilepticus. Pediatr Neurol. 2020; 112: 7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcAbee GN, Brosgol Y, Pavlakis S, et al.: Encephalitis Associated with COVID-19 infection in an 11-Year-Old Child. Pediatr Neurol. 2020; 109: 94. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPedersen SF, Ho YC: SARS-CoV-2: a storm is raging. J Clin Invest. 2020; 130(5): 2202–2205. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWu Y, Xu X, Chen Z, et al.: Nervous system involvement after infection with COVID-19 and other coronaviruses. Brain Behav Immun. 2020; 87: 18–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZawar I, Khan AA, Sultan T, et al.: Dyke-Davidoff-Masson Syndrome. An unusual cause of status epilepticus. Neurosciences (Riyadh). 2015; 20(4): 385–7. PubMed Abstract | Free Full Text\n\nAlam M, Haq MAU, Ali F, et al.: Dyke-Davidoff-Masson Syndrome: An Unusual Cause of Status Epilepticus and Refractory Seizures. J Coll Physicians Surg Pak. 2018; 28(6): S99–101. PubMed Abstract\n\nRoy U, Panwar A, Mukherjee A, et al.: Adult Presentation of Dyke-Davidoff-Masson Syndrome: A Case Report. Case Rep Neurol. 2016; 8(1): 20–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShahid R: An unusual presentation of Dyke-Davidoff Masson syndrome. Neurosciences (Riyadh). 2018; 23(3): 254–257. PubMed Abstract | Publisher Full Text\n\nMalik P, Garg R, Gulia AK, et al.: Dyke-Davidoff-Masson Syndrome- a rare cause of refractory epilepsy. Iran J Psychiatry. 2014; 9(1): 42–4. PubMed Abstract | Free Full Text\n\nPark KI, Chung JM, Kim JY: Dyke-Davidoff-Masson Syndrome: cases of two brothers and literature review. J Epilepsy Res. 2014; 4(1): 24–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKumar NV, Gugapriya TS, Guru AT, et al.: Dyke-Davidoff-Masson syndrome. Int J App Basic Med Res. 2016; 6(1): 57–59. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAyas ZÖ, Asil K, Öcal R: The clinico-radiological spectrum of Dyke-Davidoff-Masson syndrome in adults. Neurol Sci. 2017; 38(10): 1823–8. PubMed Abstract | Publisher Full Text\n\nSarangi P, Mangaraj PD, Mohanty J: Dyke-Davidoff-Masson Syndrome (DDMS): A rare preventable cause of refractory epilepsy. American Journal of Diagnostic Imaging. 2017; 1(1): 28. Publisher Full Text\n\nLi Y, Zhang T, Li B, et al.: A potential cause of adolescent onset Dyke-Davidoff-Masson syndrome: A case report. Medicine (Baltimore). 2019; 98(51): e18075. PubMed Abstract | Publisher Full Text | Free Full Text\n\nhttps://www.drugbank.ca/drugs/DB06218/. accessed on August 25 2020."
}
|
[
{
"id": "76973",
"date": "29 Jan 2021",
"name": "Emilia Virginia Noormahomed",
"expertise": [
"Reviewer Expertise Infectious diseases and health professionals education"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting and unique case report presentation documenting a COVID-19 and status epilepticus in Dyke-Davidoff-Masson syndrome, its management and outcomes. It also discusses and highlights some precautions to be considered when treating patients in SE and COVID-19. The infection by SARS-COV-2 is a very recent disease which possesses many challenges in terms of diagnosis and clinical management of its less common clinical presentation. This case report and review of the literatures provide scientific information that will enrich the few existing data, thus contributing to better understand the disease and associated co-morbidities, especially in case of associated neurological diseases.\nFurther, this case report is of potential importance to a global audience. My suggestions for further improvement are as follows:\nMake corrections to language and grammar in the manuscript.\n\nSpell out some acronyms the first time they write it.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes",
"responses": [
{
"c_id": "6321",
"date": "29 Jan 2021",
"name": "Humberto Foyaca-Sibat",
"role": "Author Response",
"response": "Dear Prof Noormahoomed Thanks a lot for your review and constructive suggestions. We will correct by what you suggested. Again, thanks for your kind attention and professionalism Regards Dr Foyaca"
}
]
},
{
"id": "134971",
"date": "21 Jun 2022",
"name": "Rizwana Shahid",
"expertise": [],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIt is an interesting case, highlighting a rare disease with the co-occurrence of the current pandemic. As we are aware, any infection can precipitate status epilepticus in epileptic patients, but with the emergence of the new COVID-19 virus, this report gives the reader good insight.\nIn this patient, the reason for status epilepticus is multifactorial rather than being secondary to only DDMS/COVID-19 infection, and the authors have addressed this issue in their discussion.\nThere is one issue that needs to be addressed:\nThe patient was chronic epileptic, as per the authors, and was on many AED but there is no imaging done before. Can you explain why the patient wasn’t worked up before?\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-9
|
https://f1000research.com/articles/10-8/v1
|
08 Jan 21
|
{
"type": "Data Note",
"title": "Dataset: local government mask orders preceding statewide orders by US states",
"authors": [
"Philip Jacobs",
"Arvi Ohinmaa",
"Arvi Ohinmaa"
],
"abstract": "We present a database listing local government mask orders for COVID-19 that were enacted between April and September, 2020, prior to the date that the governors issued statewide mask wearing mandates. We obtained data from a Google search of web pages of local and national commercial and public broadcasters and newspapers, and of the orders themselves. In the database, we present data identifying the county, municipality or tribal council, date of the order, and the source’s internet address. In the 34 states with statewide orders, local governments in 21 of these states issued mandates in 218 municipalities, 155 counties, and 1 tribal council. The dataset can be accessed from https://doi.org/10.7939/DVN/NDFEHK",
"keywords": [
"City mask orders",
"County mask orders",
"COVID-19 masks",
"local government prevention",
"COVID-19",
"health communication",
"health policy"
],
"content": "Introduction\n\nDuring the Spring of 2020, the use of face masks in public places emerged as an important determinant of the prevention of COVID-19. By August, 2020 public health officers in 34 US states had issued statewide orders for occupants to wear masks in public places1. In many of these states, local governments issued their own mask orders prior to the statewide orders. When we are considering the impact of mask wearing orders, we need to know the full extent to which local governments required occupants to wear masks in public. We developed a dataset of mask orders by local government units (counties and cities) in the states which eventually enacted statewide orders, and the dates which these orders came into effect.\n\n\nMethods\n\nOur initial sample consisted of 34 states whose governments issued statewide mask wearing mandates by 1 September, 2020. Starting with the date that each state issued statewide orders, and going backwards until early April, we conducted Google searches with the following search terms: state AND city or county or tribal group (general and specific terms) AND COVID-19 AND “mask order” or “mask mandate” AND date (backwards from state order date). From the resulting articles we searched first for website news articles from local newspapers, commercial TV and radio stations, and local Public Radio (NPR) or television (PBS) stations that listed government mask orders. If there was no statewide list, we then searched for articles on orders from key counties and cities in all of the remaining states. From these items, and for each state, we developed a list of cities and counties where orders had been reported. We recorded the date on which each order came into effect, and also the internet address of the mask order or news source reporting on a mask order.\n\n\nDataset description\n\nAmong the 34 states that issued statewide orders, counties, cities or tribal councils in 21 states issued orders prior to the statewide mandates in 21 states. We could not find any early local orders in the following 13 states: Connecticut, Delaware, Hawaii, Kentucky, Maine, Maryland, Nevada, New Jersey, New Mexico, New York, Pennsylvania, Virginia, and West Virginia. In the accompanying Excel file (Underlying data), we present the state name, the name of the local area, the designation of the area as a county (C), municipality (M) or Tribal Council (T), and the date the local mask order came in effect and the reference for the mask order.\n\nWe present data on the number of orders by C, M, and T, along with the date of the state order going into effect in Table 1.\n\nWe should note that although we list Mississippi state as having local mask orders, in fact it was the Governor who issued the counties’ orders: counties were exempt from the orders if they had incidences of COVID-19 below rates set by the Governor’s office and the State Health Officer.\n\n\nSummary\n\nOur dataset shows the number of local government units that established mask orders prior to the states issuing statewide orders. In Table 1 we show the number of local government orders for each state. In the 34 states, 218 municipalities, 155 counties and 1 tribal council issued orders.\n\n\nData availability\n\nUniversity of Alberta Library Dataverse: Local mask orders pre Statewide, https://doi.org/10.7939/DVN/NDFEHK2.\n\nThe database contains detailed collected data for 21 states with local orders that were in effect prior to statewide orders:\n\nA. County, Municipality or Tribal Council\n\nB. State\n\nC. Identification of locality as county (C), City or town (M), or Tribal Council (T) + source data embedded.\n\nD. Date the local order came into effect\n\nData are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).",
"appendix": "References\n\nJacobs P, Ohinmaa A: Dataset: percent of population covered by local government mask orders in the US [version 1; peer review: awaiting peer review]. F1000Res. 2020; 9: 1267. Publisher Full Text\n\nJacobs P: Local mask orders pre Statewide. UAL Dataverse, V1, UNF: 6:ulvNXl/tEMknDEhbD5gGFw== [fileUNF]. 2020. http://www.doi.org/10.7939/DVN/NDFEHK"
}
|
[
{
"id": "77018",
"date": "26 Jan 2021",
"name": "Ying-chu NG",
"expertise": [
"Reviewer Expertise health care services research"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nNo further comments. The study seems to be done in an appropriate manner. The issue addressed in the article is timely needed in these days. Any studies about commenting on government policy on preventing COVID-19 can make use of the data for analysis. With persistence of COVID-19, the data can be updated in the coming months.\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes",
"responses": []
},
{
"id": "82423",
"date": "23 Apr 2021",
"name": "Meg Seymour",
"expertise": [
"Reviewer Expertise epidemiology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a well-done dataset and could be very useful to other researchers. It should be accepted. Our only suggestion is that perhaps listing the dates numerically on the spread sheet would be easier to read. On the other hand, having the months listed as words might make it easier to search.\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes",
"responses": []
},
{
"id": "81356",
"date": "28 Apr 2021",
"name": "Karen Lee",
"expertise": [
"Reviewer Expertise Health economics",
"Health policy"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe format for this publication is appropriate as a \"data note\". The research reported is timely and likely to evolve over the next few months - potentially warranting an update. Information on validation of the data directly with sources might have been helpful, but likely would require more time given other priorities with the governments.\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-8
|
https://f1000research.com/articles/9-380/v1
|
19 May 20
|
{
"type": "Research Article",
"title": "Pictures over words: a cross-sectional study reporting short term memory abilities in children",
"authors": [
"Aysha Rooha",
"Malavika Anakkathil Anil",
"Jayashree S. Bhat",
"Aysha Rooha",
"Jayashree S. Bhat"
],
"abstract": "Background: An impressive amount of research has been conducted studying modality-specific Short Term Memory (STM) skills in children differing in the type of tasks used. In the present study, we aimed to profile the STM abilities based on modality, in typically developing children using a story-based assessment. Methods: The study followed a cross-sectional design and comprised of 80 participants between the ages of 6 years to 9 years 11 months. An animated story was shown to the children, following which a word recall task was performed. In this task, children were asked to recall the words mentioned in the story from a pictorial array. Results: One-way analysis of variance revealed a significant difference in the overall recall abilities of children. The recall performance was strongly related to the modality of the presentation of words. A marginal difference was observed for the recall of auditory-visual words in comparison to recall of words in the auditory modality; wherein older children recalled better in comparison to younger children. The findings of the study could be attributed to the \"visual superiority effect\", \"encoding specificity principle of memory\" and \"multimedia effect.\" Conclusion: STM abilities were observed to increase with age, with the existence of asynchrony in the auditory-visual and auditory recall scores indicating the firm reliance on the modality of presentation of word. The study implications emphasize on the use of visual stimuli for teaching new vocabularies, skills, and concepts in younger children. These findings also highlight the use of visual stimuli while assessing speech, language, and cognitive skills in younger children.",
"keywords": [
"Cognitive communication",
"Recall",
"Short Term Memory",
"Stories"
],
"content": "Introduction\n\nShort term memory (STM) skills are dependent on the modality of presentation of stimuli. However, the literature on modality has shown contradictory results. Pillai & Yathiraj (2017) reported that auditory recall is superior when compared to visual recall in children aged 7- and 8-years old. On the contrary, Vuontela et al. (2003) observed that visual memory reaches functional maturity earlier than the auditory memory system in children between 6 and 13 years. These studies differed in the type of task used, and neither were context-based nor curriculum-oriented. Academics make children “context-bound” by exposing them to a rich curriculum via textbooks and curricular activities (Walker & Lombrozo, 2017). Given these observations, using stories would be both curriculum-oriented and context-bound in assessing STM skills. The present study aimed to profile modality-specific STM skills in children between the age range of 6 years to 9 years 11 months using a story-based assessment.\n\n\nMethods\n\nThe study followed a cross-sectional design that was approved by the Institutions Ethics Committee (Ethical Reference Number - IEC KMC MLR 11-18/463). The study was conducted between December 2018 and January 2020. The study was conducted within the classroom setting of the approached school. Written informed consent was procured from parents of children who agreed to take part in the study.\n\nEnglish medium schools affiliated to the Central Board of Secondary Education (CBSE), Mangalore city, were approached, after obtaining authorization from the Block Education officer, to recruit participants for the study. Those schools that provided permission to conduct the study were considered for data collection. Typically developing children who passed the WHO Ten-Question Disability Screening Checklist (Singhi et al., 2007) were recruited for the study. Children with a history of any transfer from more than one school; a history of any shift in the medium of instruction; or a history of academic failures were excluded from the study.\n\nA sample size of 80 was determined with respect to the study done by Appose & Karuppali (2018) using the formula: n = Zα2σ2/d2 where, Zα = 1.96 at 95% confidence level, d = 20% of the mean and, σ = standard deviation. The 80 participants were assigned equally into four groups (Group I: 6 years – 6 years 11 months; Group II: 7 years – 7 years 11 months; Group III: 8 years – 8 years 11 months; and Group IV: 9 years – 9 years 11 months ).\n\nA story, “The Wooden Box” (copyright © 2019, Anil and Bhat), was constructed as animated stimuli and a “Word Recall” task (Extended data (Rooha et al., 2020)) was formulated based on the story to assess STM skills. The final modified task included 12 pictorially represented words that had an equal number of words from the story (Gold coins, Cupboard, Keys, Traffic), words thematically related (Hut, Chair, Diamond, Bag), and words unrelated to the story (Apple, Chair, Frock, Flower). Among the four words from the story, two of the words were presented in the auditory-visual modality (Gold coins, Cupboard), while the other two words were presented in auditory modality alone (Traffic, Keys). The task was to identify the pictures representing the words from the story. A score of one was given for each word recall.\n\nThe animated story as well as the formulated task was content validated by three speech-language pathologists and three primary school teachers of CBSE. The suggestions provided included using Indian names for the characters of the story, to modify the instructions and to modify certain words in the formulated task. These suggestions were incorporated in the preparation of the final stimuli.\n\nEach child was evaluated individually in a classroom. The animated story was presented on a laptop screen, followed by the administration of the word recall task. The examiner presented the task verbally, and the child’s responses were scored simultaneously.\n\nThe responses were subjected to statistical analysis using SPSS software version 16.0 and significance was set at the 0.05 level (p<0.05). Descriptive statistics were used to obtain the mean and standard deviation of the data. A frequency measure was done to analyse the percentage of children recalling each of the words from each of the groups. One-way analysis of variance (ANOVA) test was used to test the difference in recall performance across the age groups. Further, the Bonferroni post-hoc test was done to assess the pair-wise differences in performance between the groups.\n\n\nResults\n\nThe results of the descriptive statistics revealed a steady increase in performance across the age groups (Figure 1). Group IV obtained the highest scores, followed by Group III, Group II, and Group I.\n\nGroup I – 6-6.11 years, Group II – 7-7.11 years, Group III – 8-8.11 years, and Group IV – 9-9.11 years.\n\nThe results of one-way ANOVA revealed a statistically significant difference in the Word Recall task [F(3,76) = 8.387, P=0.000] across the groups. Bonferroni Post-hoc results revealed that only Group I differed significantly from other groups, as depicted in Table 1. Frequency measure of children who recalled each of the words across the four age groups is depicted in Figure 2, which reveals that words ‘Gold coins’ and ‘Cupboard’ were recalled by almost all children. However, recall of words ‘Key’ and ‘Traffic’ increased with age, with drastic changes in the recall of word ‘Traffic’.\n\nGroup I – 6-6.11 years, Group II – 7-7.11 years, Group III – 8-8.11 years, and Group IV – 9-9.11 years.\n\n* Significance at the 0.05 level\n\nGroup I – 6-6.11 years, Group II – 7-7.11 years, Group III – 8-8.11 years, and Group IV – 9-9.11 years.\n\n\nDiscussion\n\nThe results revealed that the recall ability increases significantly with age. The findings agree with the study done by Belacchi et al. (2017), where they concluded that the word recall scores increased significantly from 6 to 12 years. The study also observed a difference in recall of the words ‘Gold coins’ and ‘Cupboard’ in comparison to the words ‘Keys’ and ‘Traffic’. The differences in the recall can be attributed to the inherent characteristic of the words, as the words ‘Gold coins’ and ‘Cupboard’ were presented in auditory-visual modalities in the story. In contrast, the words ‘Key’ and ‘Traffic’ were presented in auditory modality alone. Thus, it can be observed that auditory-visual recall was superior to auditory recall in these children. The findings of the present study can be attributed to various reasons which are discussed ahead.\n\nYounger children are fascinated by the illustrations of the story, and focus more on visual animations in comparison to the auditory narration of the story. Attention thereby forms a critical prerequisite to encode, store, and subsequently recall information. Reduced attention to auditory information may have contributed to poorer recall of words presented in the auditory modality. Hayes & Birnbaum (1980) observed similar behaviour and termed it as the “visual superiority effect,” i.e., younger children are more inclined to “look and not listen.”\n\nFurther, there exists a difference in processing the two types of sensory information. For the items to be stored in the STM, the brain has to cognitively create ‘mental images’ of these items, which are pictorial representations of words inside one’s mind. When processing visual stimuli, the brain functions to discover a ‘mental image,’ but when processing auditory stimuli, the brain has to create a mental image of the heard word for correct recall (Hilton, 2001). These brain functions are mediated by higher cognitive skills, which develop only with age. This could have contributed to better recall of auditory-visual words when compared to auditory words.\n\nThis principle states that recall of memory is optimal when the retrieval conditions replicate the conditions present when memory was created (Tulving & Thomson, 1973). In the current study, recall of auditory-visual memory was superior because the retrieval condition duplicated the conditions when the memory was formed, i.e. children had to identify the same pictures as seen in the story. However, the retrieval of auditory stimuli did not duplicate the conditions when the memory was formed, i.e. children had to identify pictures of words heard in the story, which could have contributed to poorer performance.\n\nIt is claimed that presenting multimedia information, i.e. presentations of material using words and pictures (Mayer, 2002) results in deeper comprehension (Boerma et al., 2016), subsequently improving recall. This can be considered as a contributing factor for better recall of auditory-visual words.\n\nLastly, Ferrara et al. (2017) reported that detailed visual memory capacity is present in children as young as six years of age, as a result of faster maturation of visual memory than the auditory memory. This can be considered as a contributing factor for observing no differences in auditory-visual recall performance across the age group.\n\nThese evidences supports the findings of the present study; an increase in STM skills with age, the existence of asynchrony in the auditory-visual and auditory recall scores, and recall performance strongly relating to the modality of presentation of information. These findings provide implications for the use of visual stimuli while teaching new vocabularies, skills, and concepts in younger children. These findings also highlight the use of visual stimuli while assessing speech, language, and cognitive skills in younger children as it will serve as a framework for maintaining their attention while evaluating various communicative skills.\n\n\nData availability\n\nHarvard Dataverse: Replication Data for: Short Term Memory abilities in Children, https://doi.org/10.7910/DVN/LLGPOX (Rooha et al., 2020).\n\nThis project contains the following underlying data:\n\nDATA 1.tab (Analysis data with raw scores for each participant)\n\nHarvard Dataverse: Replication Data for: Short Term Memory abilities in Children, https://doi.org/10.7910/DVN/LLGPOX (Rooha et al., 2020).\n\nThis project contains the following extended data:\n\nWord Recall task.docx (Stimulus – Formulated task)\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).",
"appendix": "References\n\nAppose A, Karuppali S: Decoding the macrostructural form of oral narratives in typically developing children between 6 - 11 years of age: Using story grammar analysis. Online Journal of Health and Allied Sciences. 2018; 17(1): 12. Reference Source\n\nBelacchi C, Palladino P, Giofré D, et al.: How semantic organization of information in Long Term Memory LTM) influences Working Memory (WM) recall in children from 6 to 12 years. In IASCL 14th International Congress for the study of the Child language. 2017; 292.\n\nBoerma IE, Mol SE, Jolles J: Reading pictures for story comprehension requires mental imagery skills. Frontiers in Psychology. 2016; 7: 1630. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFerrara K, Furlong S, Park S, et al.: Detailed Visual Memory Capacity Is Present Early in Childhood. Open Mind (Camb). 2017; 1(3): 136–147. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHayes DS, Birnbaum DW: Preschoolers' retention of televised events: Is a picture worth a thousand words? Developmental Psychology. 1980; 16(5): 410–416. Publisher Full Text\n\nHilton E: Difference in Visual and auditory short term memory. IU South Bend Undergaraduate Research Journal. 2001; 4: 47–50. Reference Source\n\nMayer RE: Multimedia learning. In Psychology of Learning and Motivation. Elsevier. 2002; 41. : 85–139. Publisher Full Text\n\nPillai R, Yathiraj A: Auditory, visual and auditory-visual memory and sequencing performance in typically developing children. International Journal of Pediatric Otorhinolaryngology. 2017; 100: 23–34. PubMed Abstract | Publisher Full Text\n\nRooha A, Anil MA, Bhat JS: \"Replication Data for:Short Term Memory abilities in Children\". Harvard Dataverse, V1, UNF:6:D/2yaBdOOTOuVK07VBPgkQ== [fileUNF]. 2020. http://www.doi.org/10.7910/DVN/LLGPOX\n\nSinghi P, Kumar M, Malhi P, et al.: Utility of the WHO Ten Questions Screen for Disability Detection in a Rural Community—the North Indian Experience. J Trop Pediatr. 2007; 53(6): 383–387. PubMed Abstract | Publisher Full Text\n\nThe Wooden Box(2019)© Copyright held by Anil MA, Bhat JS. Kasturba Medical College, Mangalore, Manipal Academy of Higher Education, Manipal.\n\nTulving E, Thomson DM: Encoding specificity and retrieval processes in episodic memory. Psychological Review. 1973; 80(5): 352. Publisher Full Text\n\nVuontela V, Steenari M, Carlson S, et al.: Audiospatial and Visuospatial Working Memory in 6 - 13 Year Old School Children. Learning & Memory. 2003; 10(1): 74–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWalker CM, Lombrozo T: Explaining the moral of the story. Cognition. 2017; 167: 266–281. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "63646",
"date": "18 Jun 2020",
"name": "Ramesh Kaipa",
"expertise": [
"Reviewer Expertise Speech and Language Learning"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIntroduction: I don’t really understand what the rationale is for the current study? The authors do not build an appropriate case for pursuing this line of study. I would like to see some extensive review of previous literature. The authors also need to mention how would the current study contribute to the empirical body of knowledge.\nMethods: I would like to see a detailed explanation of how the test was administered. The authors mentioned it was presented using a laptop but I would like to see specific information on how long each of the stimuli appeared on the screen. How long were the participants allowed to wait before providing a response? Was there a penalty if the participants waited for an extended period of time before they provided a response?\nDiscussion The authors make a case that words such as “gold coins” and “cupboard” were recalled by more number of participants. I wonder is it because the participants are exposed to these stimuli items more than the other items that were presented to them? I wish the authors had performed a familiarity rating on each of the stimuli items. Currently, it is unknown what played a role in better identification of specific words over others? This is a major confound of the current study,\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "5781",
"date": "14 Sep 2020",
"name": "Malavika Anil",
"role": "Author Response",
"response": "Pictures over words: a cross-sectional study reporting short term memory abilities in children Rooha A, Anil MA and Bhat JS INTRODUCTION: I don’t really understand what the rationale is for the current study? The authors do not build an appropriate case for pursuing this line of study. I would like to see some extensive review of previous literature. The authors also need to mention how would the current study contribute to the empirical body of knowledge. INTRODUCTION: The paper was designed considering a word limit of 1000 words, hence an extensive literature review was not included in the Introduction. However, the authors have built the need by citing relevant literature to peruse the study in the introduction. The contribution of the present study to the empirical body of knowledge has been discussed in the concluding paragraph of the discussion section. METHOD: I would like to see a detailed explanation of how the test was administered. The authors mentioned it was presented using a laptop but I would like to see specific information on how long each of the stimuli appeared on the screen. How long were the participants allowed to wait before providing a response? Was there a penalty if the participants waited for an extended period of time before they provided a response? METHOD: As the study focused only on the accuracy of the recall and not the reaction time measures, time-related aspects were not considered in the methodology section. DISCUSSION: The authors make a case that words such as “gold coins” and “cupboard” were recalled by more number of participants. I wonder is it because the participants are exposed to these stimuli items more than the other items that were presented to them? I wish the authors had performed a familiarity rating on each of the stimuli items. Currently, it is unknown what played a role in better identification of specific words over others? This is a major confound of the current study, DISCUSSION: The story included words that are familiar to children, ascertained through a textbook analysis of vocabulary. Content validation was also done by primary school English teachers who approved the familiarity of the words considered in the story by all students. The words considered were all nouns, having only a single frequency of occurrence in the whole story which was also content validated by primary school English teachers and Speech-Language Pathologist."
}
]
},
{
"id": "65393",
"date": "24 Jul 2020",
"name": "Raju Sapkota",
"expertise": [
"Reviewer Expertise Visual short-term memory",
"cross-modal binding",
"early dementia",
"aging effects in human visual short-term memory"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe major issue with this article is that the background is poorly developed. There is abundant evidence in literature on the effect of visual, auditory or cross-modal effects in STM.\nAlso, while the authors associate the findings to aging effect, it could well be that the effects are merely the effect of learning, since children become more experienced learners with age.\nThe findings are not discussed adequately in context of the existing literature. I suggest addressing these issues before the article is accepted for indexing.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "5782",
"date": "14 Sep 2020",
"name": "Malavika Anil",
"role": "Author Response",
"response": "Pictures over words: a cross-sectional study reporting short term memory abilities in children Rooha A, Anil MA and Bhat JS INTRODUCTION: The major issue with this article is that the background is poorly developed. There is abundant evidence in the literature on the effect of visual, auditory or cross-modal effects in STM INTRODUCTION: The paper was designed considering a word limit of 1000 words, hence an extensive literature review was not included in the Introduction. However, the authors have built the need by citing relevant literature to peruse the study in the introduction. DISCUSSION: Also, while the authors associate the findings to the aging effect, it could well be that the effects are merely the effect of learning since children become more experienced learners with age. DISCUSSION: While the ‘aging effect’ and ‘more learning experience with age’ can be associated with memory performance, the authors have discussed majorly on the difference in recall between the modalities. The discussion revolves around specific concepts of modality difference. DISCUSSION: The findings are not discussed adequately in the context of the existing literature. I suggest addressing these issues before the article is accepted for indexing. DISCUSSION: The findings are related to the concepts of existing literature. The review of the existing literature has not been added to the manuscript as the paper was designed considering a word limit of 1000 words. However, we have tried to discuss in light of major contributing studies."
}
]
}
] | 1
|
https://f1000research.com/articles/9-380
|
https://f1000research.com/articles/10-7/v1
|
07 Jan 21
|
{
"type": "Software Tool Article",
"title": "rANOMALY: AmplicoN wOrkflow for Microbial community AnaLYsis",
"authors": [
"Sebastien Theil",
"Etienne Rifa"
],
"abstract": "Bioinformatic tools for marker gene sequencing data analysis are continuously and rapidly evolving, thus integrating most recent techniques and tools is challenging. We present an R package for data analysis of 16S and ITS amplicons based sequencing. This workflow is based on several R functions and performs automatic treatments from fastq sequence files to diversity and differential analysis with statistical validation. The main purpose of this package is to automate bioinformatic analysis, ensure reproducibility between projects, and to be flexible enough to quickly integrate new bioinformatic tools or statistical methods. rANOMALY is an easy to install and customizable R package, that uses amplicon sequence variants (ASV) level for microbial community characterization. It integrates all assets of the latest bioinformatics methods, such as better sequence tracking, decontamination from control samples, use of multiple reference databases for taxonomic annotation, all main ecological analysis for which we propose advanced statistical tests, and a cross-validated differential analysis by four different methods. Our package produces ready to publish figures, and all of its outputs are made to be integrated in Rmarkdown code to produce automated reports.",
"keywords": [
"16S",
"ITS",
"amplicon sequencing",
"metagenomics",
"microbial community",
"R package"
],
"content": "Introduction\n\nStudies of microbial communities tends to become a daily routine analysis for lots of laboratories and the main method to explore microbial diversity is metabarcoding, which is an amplicon targeted sequencing method (16S for bacteria and ITS for fungi). Metabarcoding generates a large amount of data and a lot of applications already exist for their processing (FROGS1, qiime2). Methods and software are continuously evolving and the main challenge for bioinformaticians is to implement the most recent and effective ones in their analysis. Here we present rANOMALY, a scalable and lightweight R package which is able to handle every step of a metabarcoding analysis, from read cleaning, contaminant filtering, taxonomic assignment, to advanced statistical analysis. rANOMALY is fully implemented in R language in which each step correspond to one function, allowing to easy implementation of new features or tools while being easy to use and maintain. The package allows the workflow to be executed on any R environment. rANOMALY only needs a CSV table describing the metadata for each sample, and a folder containing the corresponding fastq files as input. It can produce high quality figures for Rmarkdown reports along with statistical tests ready for publication.\n\nThe workflow is illustrated in Figure 1.\n\n\nImplementation\n\nrANOMALY is an R package depending on other CRAN, Bioconductor, and git R packages. It is easy to install via devtools::install_git function.\n\nSamples must have been previously demultiplexed into one file per sample with the file name following this syntax: {sampleid}_R[12].fastq. The denoising process is handled using the dada2 R package3 which produces amplicon sequence variants (ASV) as a taxonomic unit. This improves resolution of the potential presence of microbial organisms by using a prediction model to correct sequencing errors before aggregating similar sequences. rANOMALY handles processing any region of 16S (V1 to V9) and ITS amplicon (ITS1, ITS2) sequences, in which an additional step of cutadapt4 removes ITS probes left in some short sequences. For 16S amplicon, primers are trimmed based on the primer sequence length. ASV identifiers are sequences translated into MD5 hashes which are unique identifiers based on the DNA sequences offering the possibility to be compared between projects. This step results in an object with representing sequences, a raw ASV counts table and a text file containing statistics from the denoising process.\n\nTaxonomic assignments of ASVs are carried out by IDTAXA (part of the DECIPHER package), an algorithm based on a machine learning method5. We implemented a functionality to compute assignment with two reference databases. For example, 16S amplicon sequences can be assigned with an environment specific database (DAIRYdb6, HITdb7, MIDAS8) and a general database like SILVA9 or GreenGenes10. The assignment with the best confidence and the lowest rank is kept, thus increasing assignment depth and accuracy. As taxonomy can differ between reference databases, we implemented a taxonomy validation step in rANOMALY to unravel taxonomy inconsistency like taxa with multiple ancestors or empty ranks. As supplementary features, rANOMALY functions allow users to create their own IDTAXA formatted specific reference database. The first step consists of filling the taxonomic table empty fields with the last known rank, and checking for taxonomy incongruencies as in the assignment function. A taxid file as used with RDP classifier11 is constructed and then a last function takes as input, the corrected taxonomy table, the fasta file and the taxid file to generate the IDTAXA formatted database.\n\nPhylogenetic tree is generated in three steps. First, sequences are aligned with AlignSeqs function from DECIPHER package12 by the guided-tree method. Then, distance matrix is calculated with the dist.ml function from the phangorn package. And finally, neighbour joining and pml function computes the likelihood of the phylogenetic tree.\n\nThe abundance, metadata, taxonomic table, reference sequences and phylogenetic tree are merged into a phyloseq object13.\n\nASVs have the advantage of enabling the distinction between contaminants and the real community. We have integrated the decontam package14 into the workflow. Indeed using OTU based clustering methods can agglomerate contaminant sequences with real sample sequences, the whole cluster could hence be considered as contaminant by mistake. Working with ASVs allows the use of R package decontam which will sensitively exclude contaminant ASVs. It integrates two main methods, one based on the prevalence of the contaminant in the control samples, and another one based on the DNA concentration of the samples. Moreover, our decontamination step allows users to apply various filters such as low ASVs frequency, low ASVs prevalence in real samples, and the minimum number of reads per sample.\n\nStatistical analyses are key features of rANOMALY workflow. Main descriptive analyses are integrated thanks to phyloseq functionalities. In addition, we automatized graphical representations and advanced statistical tests for alpha, beta diversity and composition plots. Above all, we have included the four most up-to-date differential analyses to assess differentially abundant features.\n\nrANOMALY allows users to explore microbial community composition with three different types of plots : classical interactive bar plot15 of raw and relative taxa abundances, rarefaction curves to check sampling effort, and Krona interactive pie charts16.\n\nAlpha diversity indices representing the specific richness are calculated (Richness, Simpson, Shannon...) with the vegan R package17. We added statistical tests such as multi-factors analysis of variance and pairwise Wilcoxon tests to assess significant differences between tested categories. We included repeated measures ANOVA to handle within-subjects variation. For example, it can be used when there are measures taken on the same individuals at different time points. This step outputs graphical representations and tables with results of statistical tests which are both saved in files and returned as list objects for markdown reports.\n\nBeta diversity analysis allows users to estimate the community differences between two samples, it is also based on the vegan package17. rANOMALY can calculate all different distances such as the one based on ASVs abundance (BrayCurtis), rank based Jaccard indices, and phylogenetic distances as UniFrac, weighted Unifrac. Graphical representations PCoA, NMDS and more are available. This analysis can be processed at different taxonomic levels and categorical factors chosen by the user. The additional statistical test of PERMANOVA uses the distance matrix to determine if microbial communities of sample groups are significantly different from each other. We added a pairwise PERMANOVA18 test to confirm significant differences between specific group of samples.\n\nDifferential analyses are meant to assess potential differentially abundant taxas between tested conditions chosen by users. rANOMALY wraps three methods:\n\nDESeq219 uses negative binomial generalized linear model with a variance stabilizing transformation on abundances.\n\nmetagenomeSeq20 uses a zero-inflated log-normal model with cumulative sum scaling normalization.\n\nmetacoder21 applies a total sample sum normalisation and uses a non-parametric Wilcoxon Rank Sum test to compare the log ratio of mean proportions.\n\nThe use of multiple methods for differential analysis allows the user to investigate which feature can be considered as differentially abundant between conditions. rANOMALY function using DESeq2 and metagenomeSeq outputs tables and plots with significant features. Metacoder outputs heat tree plots allowing users to infer differentially abundant features at each taxonomy rank and their position on the phylogenetic tree. With basic data management, and according to the identification of all significant differentially features (i.e. ASVs, genus, species) from the three methods, we generate a single table to find significant features in one or more methods and ease the interpretation. Additional information are added to the final table, like mean relative abundance for each feature and condition. A column in which condition features are significantly more abundant, features taxonomies and related sequences are added to the aggregated table. To complete differential analysis, we have included the well recognized PLS-DA from the mixOmics package22. It is a supervised classification method allowing users to identify features discriminating the sample groups.\n\nFunctions and procedures are available to help the user to generate additional figures or to export the data to third party software. For instance, shared taxa between conditions are useful to explore. We use a function that can generate Venn diagrams, or for more complex visualisation, we can produce files readable by Cytoscape23 to produce shared taxa networks. Krona diagrams can be displayed to explore sample microbial composition, where samples can be merged by a specified factor. rANOMALY allows users to generate inputs for STAMP24, which is a graphical software that provides statistical hypothesis tests and exploratory plots for analysing taxonomic profiles.\n\nrANOMALY requires R 3.6.3 or upper and can be run on any operating system with common specifications (1Go disk space, 4Go RAM, multicore CPU is recommended).\n\n\nUse case\n\nFor this example we are using a dataset from Fretin et al. study25 in which samples are from four different environments: cow milk, cow cheese (rind and core) and cow teat skin. This dataset and metadata are available on NCBI-SRA website: BioProject accession PRJNA421256. To ease access to this dataset, fastq files along with pre-formatted metadata are available on this repository.\n\nUp to date code is hosted by the INRAE gitlab, users can simply download and install the package in R console with following command lines.\n\n\n\nASV definition with DADA2. The first step is to define ASVs thanks to the dada2 package. In rANOMALY, only one function is needed to compute all the different steps require from this package. Here sample names are extracted from the file name, thus be sure that files name match samples name in the metadata file.\n\n\n\nMain outputs of this function are:\n\nread_tracking.csv summarizes the read number after each filtering step (Table 1).\n\nraw_otu-table.csv the raw ASV table.\n\nrep-seqs.fna fasta file with all representative sequences for each ASV.\n\nrobjects.Rdata with saved dada_res list containing raw ASV table and representative sequences in objects otu.table, seqtab.export, seqtab.nochim.\n\n\n\nTaxonomic assignment. assign_taxo_fun function uses IDTAXA function from DECIPHER package, and allows to use two different databases. It keeps the best assignment on two criteria, resolution (depth in taxonomy assignment) and confidence (value givenby IDTAXA). The final taxonomy is validated by checking for multiple ancestors (i.e. same species assigned to different genus) and incongruities (i.e. empty fields or incomplete lineage) correction step.\n\nWe share the latest databases we use in the IDTAXA format in this link. Users can also generate your own IDTAXA formatted database following those instructions and scripts we provide at this page.\n\n\n\nMain file outputs:\n\nrobjects.Rdata with taxonomy in phyloseq format in tax.table object.\n\nfinal_tax_table.csv the final assignation table (tax.table) outputed in CSV format.\n\nallDB_tax_table.csv raw assignations from the two databases, mainly for debugging.\n\nPhylogenetic tree. The phylogenetic tree from the representative sequences is generated using phangorn and DECIPHER packages.\n\n\n\nPhyloseq object. To create a phyloseq object, we need to merge four objects and one file:\n\nthe ASV table otu.table and the representative sequences seqtab.nochim in dada_res variable.\n\na taxonomy table (tax.table).\n\nthe phylogenetic tree (tree).\n\nmetadata from from csv file.\n\n\n\nDecontamination. The decontam_fun function uses decontam R package along with control samples (PCR control) to filter out contaminants. The decontam package offers two main methods, frequency and prevalence (users can also combine those methods). For frequency method, it is mandatory to have the DNA concentration of each sample in the phyloseq object (and hence in the metadata.csv). The prevalence method does not need DNA quantification, this method allows to compare presence/absence of ASV between real samples and control samples and then identify contaminants.\n\nTips: sequencing plateforms often quantify the DNA before sequencing, but do not usually give the information. Just ask for it ;).\n\nOur function integrates the basic ASV frequency (Freq=nb_reads_ASVnb_total_reads) and minimum prevalence in overall samples filtering. We have also included an option to filter out ASV based on their taxa names for known laboratory recurrent contaminants.\n\nMain outputs:\n\nrobjects.Rdata with contaminant filtered phyloseq object named data.\n\nExclu_out.csv list of filtered ASVs for each filtering step.\n\nKrona plot before and after filtering.\n\nraw_asv-table.csv and relative_asv-table.csv.\n\nvenndiag_filtering.png venn diagram showing the repartition of filtered ASVs by decontamination methods.\n\nHere we are going to filter out ASVs representing less than 0.1% (freq) of the reads and that are present in more than 4 samples (prev). Moreover, we are excluding \"unassigned\" taxas for this use case. Our sample dataset do not contain control samples, this step will be skipped.\n\n\n\nWe obtain the final phyloseq object used for downstream analysis:\n\n\n\nRarefaction curves. In order to observe the sampling depth of each sample we start by plotting rarefaction curves. Those plots are generated by plotly which makes them interactive. (Figure 2)\n\n\n\nCommunity composition plot. The bars_fun function allows user to generate interactive community composition plot. Figure 3 presents the composition plot with relative abundances for the top 20 genera existing in our samples. The function allows to plot at different taxonomy rank and to modify the number of taxa to show.\n\nHere two arguments controlling the composition plot aesthetics:\n\nOrd1 option order the sample along the X axis.\n\nFact1 option control labels of the X axis. Fact1=\"sample.id\" if user don’t want the sample to be renamed.\n\nsource_location factor shows very different bacterial community between milk, cheese and cow teats environments.\n\n\n\nAlpha diversity. The alpha_diversity_fun function can computes various alpha diversity indexes. It uses the estimate_richness function from phyloseq (Available measures : Observed, Chao1, ACE, Shannon, Simpson, InvSimpson, Fisher). Here we calculate ASV richness and Shannon index and carry out an analysis of variance on the source_location factor. Sequencing depth is automatically taken into account in this test. A pairwise wilcoxon test is added to ANOVA to define which group might be significantly different from others. Figure 4 shows boxplots of diversity indices, cow teats environment has much more ASV than other environments. Shannon index reveals more differences between cheese rinds and cheese cores. Cheese rinds show a higher Shannon index highlighting a more balanced bacterial community.\n\n\n\nResults of the analysis of variance and pairwise wilcoxon test on Shannon index:\n\n\n\nThe alpha_diversity_fun function returns a list which contains:\n\nboxplots comparing conditions with chosen indices. ($plot)\n\na table of indices values. ($alphatable)\n\nAnd for each of the computed indices :\n\nan ANOVA analysis. (${measure}$anova)\n\na pairwise wilcox test result comparing conditions and giving the pvalue of each comparison tested: ${measure}$wilcox_col1: wilcox test results on the first or unique factor, ${measure}$wilcox_col2_fdr: wilcox test results on the second factor, ${measure}$wilcox_col2_collapsed: wilcox test results on collapsed factor 1 and factor 2.\n\na mixture model if your dataset includes repeated measures, ie. column3 option. (${measure}$anovarepeat, ${measure}$mixedeffect)\n\nBeta diversity. The diversity_beta_light function allows to generate specific tests and figures ready to publish in rmarkdown report as in the example below. It is based on the vegan package function vegdist for the distance calculation and phyloseq-extended in addition to ordinate funtion for the ordination plot.\n\nWe include statistical tests to ease the interpretation of results. A permutational ANOVA is carried out on matrix distance to compare groups by testing if centroids and dispersion are equivalent for all groups. User have to inform col argument and optionally cov (covariable) to assess PERMANOVA to determine significant differences between groups. A pairwise-PERMANOVA is processed to determine which condition is significantly different from another (based on p-value).\n\nAs a return, you will get a list that contains:\n\nAn ordination plot ($plot).\n\nThe permANOVA results ($permanova).\n\nThe pairwise permANOVA ($pairwisepermanova)\n\nHere we present results of beta diversity analysis on source_location factor, Figure 5 show ordination plot and it confirms the big differences between community of cheese, milk and cow teats.\n\n\n\nThe permanova tests on BrayCurtis distance shows significant p-value for the source_location factor. Pairwise permanova test is used to define which level of the factor tested is significantly different from others.\n\n\n\n\n\nDifferential analysis. We choose three different methods to process differential analysis which is a key step of the workflow. The main advantage of the use of multiple methods is to cross validate deferentially abundant taxa between tested conditions. For this use case, we choose to focus on milk and cow teat environment to compare community at genus level.\n\nMetacoder Metacoder is the most simple differential analysis tool of the three. Counts are normalized by total sum scaling to minimize the sample sequencing depth effect and it uses a Kruskal-Wallis test to determine significant differences between groups. The metacoder_fun function allows the user to choose the taxonomic rank, which factor to the test (column1), and a specific pairwise comparison (comp) to launch the differential analysis.\n\nIt produces pretty graphical trees, representing taxas present in both groups and coloring branches depending on which group this taxa is more abundant (Figure 6). Two trees are produced, a raw one, where everything is displayed and a filtered one where only significant features are represented (p-value <= 0.05).\n\n\n\nMain output is a list with :\n\nfor each comparison (${comparison}), two heattree, one with all features (${comparison}$raw) and an other one with only significant features (${comparison}$signif).\n\na table with all wilcoxon test results ($table).\n\nDESeq2 DESeq2 is a widely used method, primarily for RNAseq applications, for assessing differentially expressed genes between controlled conditions. Its use for metabarcoding datas is sensibly the same and well documented. The deseq2_fun allows to process differential analysis as metacoder_fun, and users can choose the taxonomic rank, the factor to test and which condition to compare. DESeq2 algorithm uses negative binomial generalized linear models with VST normalization (Variance Stabilizing Transformation).\n\nMain output is a list with:\n\na plot showing Log2FoldChange value of each significant feature (${comparison}$plot).\n\na table with statistics (LogFoldChange, pvalue, adjusted pvalue...) (${comparison}$table).\n\n\n\nMetagenomeSeq MetagenomeSeq uses a normalization method able to control for biases in measurements across taxonomic features and a mixture model that implements a zero-inflated Gaussian distribution to account for varying depths of coverage. As deseq2_fun, metagenomeseq_fun returns a table with statistics and a plot with significant features for each comparison.\n\n\n\nResults aggregation step The aggregate_fun function allows to merge the results from the three differential analyses methods computed previously to obtain one unique table with all informations of significant differentially abundant features. Figure 7 shows the most significant and differential abundant Genera between the two environments.\n\n\n\nThe generated table include the following fields:\n\nseqid: ASV ID.\n\nComparaison: Tested comparison.\n\nDeseq / metagenomeSeq / metacoder: differentially abundant with this method (0 no or 1 yes).\n\nsumMethods: sum of methods in which feature is significant.\n\nDESeqLFC: Log Fold Change value as calculated in DESeq2.\n\nabsDESeqLFC: absolute value of Log Fold Change value as calculated in DESeq2.\n\nMeanRelAbcond1 / MeanRelAbcond2: Means relative abundance in condition 1 and 2.\n\nCondition: in which the mean feature relative abundance is higher.\n\nTaxonomy and representative sequence.\n\nHere is an overview of the aggregate_fun table informing in which methods each feature is significant, their DESeq2 LogFoldChange value, taxonomy and representative sequences:\n\n\n\nHeatmap. User can generate an interactive heatmap with heatmap_fun function to explore relative abundance of top taxa through samples as showed in Figure 8.\n\n\n\nVenn diagram of shared taxa. Figure 9 shows Venn diagram comparing the three environment of this study, this method is useful to determine shared taxa between group of samples. This function uses venn::venn function which handles up to 7 groups comparison.\n\n\n\nASVenn_fun generate a table allowing user to find which taxa are shared between conditions.\n\n\n\nphy2cyto_fun allows to generate input files (SIF format) for Cytoscape which is useful to visualise shared taxa and easily modify each nodes and arrows position and aesthetic.\n\nphy2tsv_fun function allows user to generate tabulated format abundance table at different taxonomic rank with following commnands:\n\n\n\nexport_to_stamp_fun function creates two files that can be imported into the STAMP software.\n\n\n\ncsv2phyloseq_fun function allows user to import data from other bioinformatic pipeline like FROGS, Qiime2. This function needs tabulated ASV, taxonomy, metadata and DNA sequence table as inputs. It can generate phylogenetic tree if missing and output a phyloseq object ready for downstream analyses.\n\n\nConclusions\n\nrANOMALY allows users to handle metagenomic data from raw sequences quality control to final differential analysis with ready to publish results in an easy and reproducible manner. Users have access to all sources of the rANOMALY package that can be deployed on any operating system or server allowing them to analyse anything from a few samples to several thousand. This workflow combines all of the latest developments in the field: the use of high resolution amplicon sequence variant, contaminant filtering, double automatic taxonomic assignation, integrated statistical analyses and four differential analyses with cross validation. rANOMALY help users those who don’t have big programming skills. And since rANOMALY uses phyloseq objects and standards, original functions from the phyloseq package can still be used to split, filter and select specific samples previously to be visualized and/or tested with rANOMALY functions. For exploratory analysis and interactive experience, a shiny application based on this workflow is under active development and will be shared with the community.\n\n\nSoftware availability\n\nUp-to-date source code, and tutorials are available at: https://forgemia.inra.fr/umrf/ranomaly.\n\nPackage documentation is also provided at: https://umrf.pages.mia.inra.fr/ranomaly/use_case.html\n\nArchived source code as at time of publication are available from: https://doi.org/10.5281/zenodo.433883326\n\nLicense: Creative Commons Attribution 4.0 International\n\nHere we show all packages used in this workflow and their version numbers:\n\n",
"appendix": "Acknowledgements\n\nAuthors deeply thank Philippe Ruiz (INRAE, MEDIS) and Marco Meola (Agroscope) for testing and feedback on the rANOMALY package.\n\n\nReferences\n\nEscudie F, Auer L, Bernard M, et al.: FROGS: Find, Rapidly, OTUs with Galaxy Solution. Bioinformatics. 2018; 34(8): 1287–1294. PubMed Abstract | Publisher Full Text\n\nBolyen E, Rideout JR, Dillon MR, et al.: Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol. 2019; 37(8): 852–857. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCallahan BJ, McMurdie PJ, Rosen MJ, et al.: DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods. 2016; 13(7): 581–583. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartin M: Cut adapt removes adapter sequences from high-throughput sequencing reads. EMBnet.Journal. 2011; 17(1): 10–12. Publisher Full Text\n\nMurali A, Bhargava A, Wright ES: IDTAXA: a novel approach for accurate taxonomic classification of microbiome sequences. Microbiome. 2018; 6(1): 140. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMeola M, Rifa E, Shani N, et al.: DAIRYdb: a manually curated reference database for improved taxonomy annotation of 16S rRNA gene sequences from dairy products. BMC Genomics. 2019; 20(1): 560. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRitari J, Salojarvi J, Lahti L, et al.: Improved taxonomic assignment of human intestinal 16S rRNA sequences by a dedicated reference database. BMC Genomics. 2015; 16: 1056. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNierychlo M, Andersen KS, Yujuan XU, et al.: MiDAS 3 : An ecosystem-specific reference database, taxonomy and knowledge platform for activated sludge and anaerobic digesters reveals species-level microbiome composition of activated sludge. Water Res. 2020; 182: 115955. PubMed Abstract | Publisher Full Text\n\nQuast C, Pruesse E, Yilmaz P, et al.: The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res. 2013; 41(Database issue): D590–D596. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDeSantis TZ, Hugenholtz P, Larsen N, et al.: Greengenes, a Chimera-Checked 16S rRNA Gene Database and Workbench Compatible with ARB. Appl Environ Microbiol. 2006; 72(7): 5069–5072. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang Q, Garrity GM, Tiedje JM, et al.: Naïve Bayesian Classifier for Rapid Assignment of rRNA Sequences into the New Bacterial Taxonomy. Appl Environ Microbiol. 2007; 73(16): 5261–5267. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWright ES: DECIPHER: harnessing local sequence context to improve protein multiple sequence alignment. BMC Bioinformatics. 2015; 16: 322. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcMurdie PJ, Holmes S: phyloseq: An R Package for Reproducible Interactive Analysis and Graph-ics of Microbiome Census Data. PLoS One. 2013; 8(4): e61217. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDavis NM, Proctor DM, Holmes SP, et al.: Simple statistical identification and removal of contaminant sequences in marker-gene and metagenomics data. Microbiome. 2018; 6(1): 226. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSievert C: Interactive Web-Based Data Visualization with R, plotly, and shiny. Chapman and Hall/CRC. 2020. Publisher Full Text\n\nOndov BD, Bergman NH, Phillippy AM: Krona: Interactive Metagenomic Visualization in a Web Browser. BMC Bioinformatics. 2011; 12: 385. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOksanen J: vegan: Community Ecology Package. R package version 2.5-5, 2019. Reference Source\n\nArbizu MP, et al.: pairwiseAdonis: Pairwise multilevel comparison using adonis. R package version 0.3. 2019.\n\nLove MI, Huber W, Anders S: Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014; 15(12): 550. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaulson JN, Stine OC, Bravo HC, et al.: Differential abundance analysis for microbial marker-gene surveys. Nat Methods. 2013; 10(12): 1200–1202. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFoster ZSL, Sharpton TJ, Grunwald NJ: Metacoder: An R package for visualization and manipulation of community taxonomic diversity data. PLoS Comput Biol. 2017; 13(2): e1005404. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRohart F, Gautier B, Singh A, et al.: mixOmics: An R package for 'omics feature selection and multiple data integration. PLoS Comput Biol. 2017; 13(11): e1005752. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShannon P, Markiel A, Ozier O, et al.: Cytoscape: a software environment for integrated models of biomolecular interaction net-works. Genome Res. 2003; 13(11): 2498–2504. PubMed Abstract | Publisher Full Text | Free Full Text\n\nParks DH, Tyson GW, Hugenholtz P, et al.: STAMP: statistical analysis of taxonomic and functional profiles. Bioinformatics. 2014; 30(21): 3123–3124. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFretin M, Martin B, Rifa E, et al.: Bacterial community assembly from cow teat skin to ripened cheeses is influenced by grazing systems. Sci Rep. 2018; 8(1): 200. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTheil S, Rifa E: rANOMALY. 2020. http://www.doi.org/10.5281/zenodo.4338833"
}
|
[
{
"id": "76934",
"date": "12 Jan 2021",
"name": "Florent Murat",
"expertise": [
"Reviewer Expertise Genomics",
"Bioinformatics",
"Evolution"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nTheil and Rifa developed here the R package rANOMALY for data analysis of 16S and ITS amplicons based sequencing. They provide a workflow and a toolkit that are very useful for bioinformaticians as well as for beginners who want to perform complete marker gene sequencing data analyses from fastq treatment (contaminant filtering, taxonomic assignment, alpha and beta diversity indexes) to statistical analysis (using three main differential analysis methods, DESeq2, Metacoder, metagenomeSeq) in an automatic, step by step, and reproducible way. Thus, this is a suitable workflow for exploratory and educational purposes. I was pleased to be able to reproduce all analyses and figures following the use case of the paper.\nHence, I only have a few minor issues that should be discussed:\nThe authors mentioned that the developed package is customizable. This is an important aspect given that projects are different from one to the other, and bioinformatic tools evolve tremendously fast so it will be crucial to adapt/update the workflow accordingly. The workflow actually allows this, however, it would be useful (particularly for beginners) to provide a vignette/tutorial on the gitlab page (and/or in a section of the paper) showing how this can be done with simple examples. The use case is based on Illumina sequences, a use case dealing with Ion torrent sequences could be relevant in this context.\n\nThe use case is based on a dataset from Fretin et al. 2018, it would be relevant to interpret the results of the workflow, and to describe to what extent they match or complement the results of the previous analysis. This would be relevant to benchmark the workflow and highlight the novelties brought by this analysis.\n\nThe access to the use case fastq sequences and to the IDTAXA formatted databases is given via links in the text. It would be convenient for users working with remote servers to download it with command lines. Please provide command lines to do so (via wget for instance), either in the text or on the gitlab page.\n\nThere is a typo on page 6, the IDTAXA formatted database that can be downloaded from the link is DAIRYdb_v1.2.4_20200603_IDTAXA.rdata and not DAIRYdb_v1.2.4_20200604_IDTAXA.rdata.\n\nPage 8: In the “bars_fun\" function, additional space has been introduced between “source_” and “location” leading to an error if one runs this piece of code.\n\nPage 9: In the “diversity_alpha_fun” function, the font has been modified for the closing double quotes after “Observed” and “Shannon”. Thus, this leads to an error if one runs this piece of code.\n\nPage 15: In the “phy2tsv\" function, the phyloseq object “data_prune\" was never created upstream. Would it be better to test the function “phy2tsv” with the phyloseq object “data” previously created in the use case?\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "76936",
"date": "22 Jan 2021",
"name": "Simon Roux",
"expertise": [
"Reviewer Expertise metagenomics",
"microbial ecology",
"phage genomics",
"viral ecology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript describes an R pipeline (rANOMALY) that connects together in a single simplified R workflow many of the standard tools now routinely used in metabarcoding data analysis. I have no doubt that such an integrated pipeline will be very useful for many researchers, especially ones not experts in bioinformatics. The modular design is also well suited for future changes and evolution of the pipeline.\nThe manuscript itself is very clear, nicely outlines how the different components of the pipeline are connected, and also clearly distinguish between which functions are provided by this pipeline and which functions are provided by third-party tools or packages. My only major comment is about user input (either user questions or bug reports), for which there is currently no process or web page mentioned in the manuscript. For instance, should users report bugs in the gitlab repo (https://forgemia.inra.fr/umrf/ranomaly/-/issues)? (Incidentally, I did encounter an unexpected error with DeSeq: Error in `assays<-`(`*tmp*`, value = `*vtmp*`) : please use 'assay(x, withDimnames=FALSE)) <- value' or 'assays(x, withDimnames=FALSE)) <- value' when the dimnames on the supplied assay(s) are not identical to the dimnames on DESeqDataSet object 'x' In addition: Warning message: In DESeqDataSet(se, design = design, ignoreRank) : some variables in design formula are characters, converting to factors). If the authors do intend users to submit gitlab issues, this should be specified in the conclusion and/or the software availability sections. If not, the alternative(s) path to provide feedback and report issues should be clearly described.\nOther than that, I only have minor comments on the manuscript which are listed below:\nAbstract: “uses amplicon sequence variants (ASV) level for”. Shouldn’t this read “uses amplicon sequence variants (ASV) for”?\nIn the keywords, I would argue “metagenomics” should be “metabarcoding”\n\np. 3 Introduction: “Studies of microbial communities tends to” should be “Studies of microbial communities tend to”\n“(FROGS 1 , qiime 2 )”: it should be made clear that these are examples of softwares but many more exist, e.g. by changing to “(e.g. FROGS1, qiime2)”, or explicitly stating “such as FROGS1 or qiime2”.\n“allowing to easy implementation of” should be “allowing for easy implementation of” or “allowing to easily implement”.\n\np. 3 Raw sequence processing: “This improves” should be clarified as “This approach improves” or “This tool improves” (to make it obvious that “this” refers to dada2).\n“an additional step of cutadapt” should be “an additional step using cutadapt”.\n\np. 4 Taxonomic assignment: “with representing sequences” should probably be clarified as “with representative sequences for each ASV” (since the concept of “representative sequences” has not been described earlier)\n“IDTAXA formatted” should be “IDTAXA-formatted”\n\np. 4 Phylogenetic tree: “with AlignSeqs function from DECIPHER package” should be “with the AlignSeq function from the DECIPHER package”\n“computes the likelihood of the phylogenetic tree” This may suggest that a tree exists and that this function evaluates the likelihood of this existing tree, but I believe the first function actually builds a tree? So there may be a clause missing here, e.g. “And finally, the neighbour joining function is used to build a phylogenetic tree based on the distance matrix, and the pml function computes the likelihood of the phylogenetic tree.”?\n\np. 5 Install: The authors may want to add a “library(devtools)” line after the install of the package devtools and before installing the package ranomaly, in case some users already have the devtools package installed.\nThe authors should also specify that a user will need to replace “PATHTOWORKINGDIRECTORY” with the path to their working directory on their own computer. (The same is true in the Taxonomic assignment section when users have to specify a path to the taxonomic databases).\n\np. 6 Processing of raw sequences: “files name match samples name” should be “file names match sample names”\n\np. 6 Taxonomic assignment: “givenby” should be “given by”\n“different genus” should be “different genera”\n“correction step” should be removed, or clarified as e.g. “as part of a correction step”\n“your own” should be “their own”\n\np. 8 Rarefaction curves: On my version of R, I have to call “rareplot” after the line “rareplot = rarefaction(data_filtered, \"source_location\", 100)” otherwise no plot appears. If this is the case in most versions (which I believe it is), this line (“rareplot”) should be included in the authors’ instructions so that a user would not be confused to not see any plot after calling the line “rareplot = rarefaction(data_filtered, \"source_location\", 100)” (this is also true for the other interactive plots, e.g. community composition, alpha and beta diversity, etc).\n\np. 8 Community composition plot: \"source_ location\": there seems to be an extra space between “source_” and “location” here\n\np. 10 Beta diversity: “funtion” should be “function”\n“on matrix distance” should be “on the distance matrix”\n“A pairwise-PERMANOVA is processed” should be “A pairwise-PERMANOVA is performed”\n\np. 11 Differential analysis: “of the use of multiple” would be more clear as “of using multiple”\n“cross validate” should be “cross-validate”\n“which factor to the test” should be “which factor to test” or “which factor to perform the test on”\n“which group this taxa is more abundant” should be “which group this taxa is more abundant in”\n\np. 12 DESeq2: “and well documented”: Could the authors add 1 or 2 references here?\n\np. 15 Venn diagram: Shouldn’t “the three environment” be “the four environments”.\n\np. 15 Export: “function allows user” should be “function allows users” (also for csv2phyloseq_fun).\n\np. 15 Conclusion: “metagenomic data” should be “metabarcoding data”\n“help users those” should be “help users”\n“big programming skills” should be “advanced programming skills”\n“previously to be visualized and/or tested” should be “prior to visualization/test” or “before being visualized and/or tested”\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/10-7
|
https://f1000research.com/articles/9-770/v1
|
24 Jul 20
|
{
"type": "Research Article",
"title": "Anti-c-myc cholesterol based lipoplexes as onco-nanotherapeutic agents in vitro",
"authors": [
"Saffiya Habib",
"Aliscia Daniels",
"Mario Ariatti",
"Moganavelli Singh",
"Saffiya Habib",
"Aliscia Daniels",
"Mario Ariatti"
],
"abstract": "Background: Strategies aimed at inhibiting the expression of the c-myc oncogene could provide the basis for alternative cancer treatment. In this regard, silencing c-myc expression using small interfering RNA (siRNA) is an attractive option. However, the development of a clinically viable, siRNA-based, c-myc silencing system is largely dependent upon the design of an appropriate siRNA carrier that can be easily prepared. Nanostructures formed by the electrostatic association of siRNA and cationic lipid vesicles represent uncomplicated siRNA delivery systems. Methods: This study has focused on cationic liposomes prepared with equimolar quantities of the cytofectin, N,N-dimethylaminopropylamido-succinylcholesteryl-formylhydrazide (MS09), and cholesterol (Chol) for the development of a simple, but effective anti-c-myc onco-nanotherapeutic agent. Liposomes formulated with dioleoylphosphatidylethanolamine (DOPE) in place of Chol as the co-lipid were included for comparative purposes. Results: Liposomes successfully bound siRNA forming lipoplexes of less than 200 nm in size, which assumed globular, bilamellar structures. The liposome formulations were well tolerated in the human breast adenocarcinoma (MCF-7) and colon carcinoma (HT-29) cells, which overexpress c-myc. Lipoplexes directed against the c-myc transcript mediated a dramatic reduction in c-myc mRNA and protein levels. Moreover, oncogene knockdown and anti-cancer effects were superior to that of Lipofectamine™ 3000. Conclusion: This anti-c-myc MS09:Chol lipoplex exemplifies a simple anticancer agent with enhanced c-myc gene silencing potential in vitro.",
"keywords": [
"cancer",
"c-myc",
"siRNA",
"gene silencing",
"cationic liposomes"
],
"content": "Introduction\n\nCancer is one of the leading causes of death world-wide. Deaths due to cancer have been reported to outnumber those due to acquired immune deficiency syndrome (AIDS), malaria and tuberculosis combined1. Cancer treatment currently involves surgery, chemotherapy and/or radiation depending on the type and stage of the disease. Despite advances in understanding tumorigenesis and disease progression, these treatments are limited by harsh side-effects, the possibility of recurrences, and are heavily dependent on early detection and diagnosis for success2. With the global cancer burden projected to increase to 21.7 million new cases and 13 million deaths by the year 2030, it is clear that more effective treatment strategies are required1.\n\nThe altered activity of the c-myc proto-oncogene has been identified as an important element in the initiation and maintenance of the cancerous state of a cell. The c-myc gene encodes a nuclear phosphoprotein that is widely recognized for its role as a transcription factor. The c-Myc protein is believed to participate in the regulation of 10-15% of all genes3. These include genes involved in cell-cycle progression, metabolism, cell growth, differentiation, adhesion, and apoptosis4–10. Hence, advances in the design of appropriate c-myc-silencing systems may prove useful in treating a broad range of cancers11.\n\nIn theory, effective c-myc silencing may be achieved using endogenous cellular machinery, provided that the designed small interfering RNA (siRNA) molecule is successfully introduced. Several factors militate against the success of naked siRNA molecules in vivo. Naked siRNA molecules are highly susceptible to serum nucleases and are rapidly cleared by the kidneys11,12, while the size (~14 kDa) and negative charge of the siRNA prevents its passage across biological membranes13. Therefore, an appropriate carrier is required to protect the siRNA from damage and elimination, as well as to disguise its negative charge. This system must have low toxicity, afford stability in serum, avoid recognition by the immune system, avoid renal clearance, and successfully deliver its contents to the RNAi machinery in the diseased cells.\n\nSince nucleic acids can electrostatically associate with positively charged agents, a variety of cationic molecules as potential carrier vehicles have been investigated, cationic lipids receiving much attention, both in laboratory-scale experiments and clinical trials14–16. Cationic liposomes formed by the self-assembly of cationic and neutral or helper lipids, are the earliest cationic lipid-based delivery systems17,18. These form lipoplexes, when associated with siRNA. Their favorable characteristics such as safety, biocompatibility, and amenability to modification, have sustained their interest in cationic liposomal-siRNA delivery19. Although several novel liposomal-siRNA systems have shown promise, none have resulted in a commercially available treatment16. Major barriers to their application include poor stability in the bloodstream and early recognition by the immune system, as these positively charged lipoplexes associate with anionic serum proteins such as albumin and lipoproteins. This results in opsonization by serum components, destabilization of the lipoplex, and damage to the nucleic acid cargo before it reaches the diseased cells. Furthermore, lipoplexes often aggregate forming larger particles that accumulate in the lung and are rapidly cleared by the reticuloendothelial system, reducing dosage and circulation time20–22. Increase of the mechanical strength of the liposome bilayer to render it more resistant to the destabilizing action of serum proteins can be achieved by the incorporation of rigid, membrane-stabilizing lipids such as cholesterol (Chol)23,24.\n\nThis study involved the formulation of two cationic liposomes containing the cationic lipid N,N-dimethylaminopropylamido-succinylcholesteryl-formylhydrazide (MS09) combined with either the neutral helper lipid dioleoylphosphatidylethanolamine (DOPE) or the bilayer-stabilizing lipid, Chol. The transfection efficiency of these cationic liposomes on oncogenic c-myc expression at the mRNA and protein levels in the MCF-7 and HT-29 cell lines were evaluated.\n\n\nMethods\n\nDOPE, Chol, RIPA buffer, bicinchoninic acid (BCA) kit were obtained from Sigma-Aldrich (St. Louis, MO, USA). 2-[-(2-hydroxyethyl) piperazinyl]-ethanesulphonic acid (HEPES), ethidium bromide (10 mg/ml), tris(hydroxymethyl)-aminomethane hydrochloride (Tris HCl), 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), acridine orange (AO), ethylenediaminetetraacetic acid (EDTA), phosphate buffered saline tablets (PBS), sodium carbonate (Na2CO3) and sodium bicarbonate (NaHCO3) were supplied by Merck (Darmstadt, Germany). The siGENOME non-targeting siRNA #1 (D-001210-01-20), ON-TARGETplus SMARTpool Human MYC (4609) siRNA (L-003282-02-0020), 5× siRNA buffer (B-002000-UB-100, 0.3M KCl, 30mM HEPES, 1mM MgCl2, pH 7.5) and molecular grade RNase-free water (B-003000-WB-100) were purchased from Thermo Scientific Dharmacon Products (Lafayette, CO, USA). Ultrapure™ agarose powder and Lipofectamine™ 3000 reagent were procured from Invitrogen (Carlsbad, CA, USA). Sodium dodecylsulphate (SDS), iScript™ genomic DNA (gDNA) clear complementary DNA (cDNA) Synthesis Kit (1725035), SsoAdvanced™ Universal SYBR® Green Supermix (1725272), PrimePCR™ SYBR® Green Assay: MYC, Human (Unique Assay ID: qHsaCID0012921); PrimePCR™ SYBR® Green Assay: ACTB, Human (Unique Assay ID: qHsaCED0036269), PrimePCR™ Template for SYBR® Green Assay: MYC, Human; PrimePCR™ Template for SYBR® Green Assay: ACTB, Hard-Shell® 96-well PCR plates, 0.2 ml PCR tube strips, 10× Tris-buffered saline (TBS), 10% Tween 20, nonionic detergent; and blotting-grade blocker (non-fat dry milk), were supplied by Bio-Rad Laboratories (Richmond, CA). MCF 7 and HT-29 cells were obtained from the American Type Culture Collection (Manassas, VA, USA). Eagle’s Minimum Essential Medium (EMEM) with L-glutamine, Fetal bovine serum (FBS), trypsin-EDTA and penicillin/streptomycin mixture (10 000 U/ml penicillin, 10 000 μg/ml streptomycin) were purchased from Lonza BioWhittaker (Verviers, Belgium). All sterile plasticware were obtained from Corning Inc. (Corning, NY, USA). TRIzol® reagent (15596-026), c-MYC epitope tag antibody 9E11 (AHO0052), goat anti-mouse IgG2A secondary antibody, horseradish peroxidase (HRP) conjugate were obtained from Life Technologies (Carlsbad, CA, USA). β-actin antibody was purchased from Santa Cruz (Santa Cruz, CA, USA). Ultrapure 18 Mohm water was used throughout. All other reagents were of analytical grade.\n\nMS09 was prepared from cholesterylformylhydrazide hemisuccinate (MS08) via an active ester intermediate according to the method previously published23. Briefly, the N-hydroxysuccinimide ester of cholesterylformylhydrazidehemisuccinate (0.083 mmol) and dimethylpropylamine (0.35 mmol) were dissolved in a H2O:pyridine:DMF (13:7:10, v/v/v) mixture. The product (MS09) was monitored and purified by thin layer chromatography. For liposome formulation, the lipids were combined in quantities as in Table 1 and concentrated to a thin film in vacuo (Büchi Rotavapor R rotary evaporator). The film was rehydrated (4°C, 48h) in sterile HEPES buffered saline (HBS, 20 mM HEPES, 150 mM NaCl, pH 7.5; 0.5 ml), vortexed and, then sonicated in a Transsonic (T460/H) bath-type sonicator (Singen, Germany). Liposome preparations were stored at 4°C and sonicated prior to use.\n\nChol = Cholesterol\n\nThe morphology of liposomes and siRNA lipoplexes were evaluated using cryogenic transmission electron microscopy (cryo-TEM), using a JEOL JEM.1010 transmission electron microscope (JEOL Ltd., Tokyo, Japan). Images were captured using an Olympus MegaView III digital camera in conjunction with SIS iTEM Universal Imaging Platform software (Shinjuku, Japan). Particle size and zeta (ζ) potential were determined using the Nanoparticle tracking analysis (NTA, NanoSight NS500, Malvern Instruments, Worcestershire, UK) fitted with a ZetaSight™ module and NTA 3.0 software.\n\nLipoplexes were prepared (w/w) with varying amounts of the liposomes and a fixed quantity of siGENOME non-targeting siRNA (0.3 μg), as per Table 2. Lipoplexes (10 μl in HBS) were mixed with gel loading buffer (40% sucrose, 0.25% xylene cyanol, 0.25% bromophenol blue; 2.5 μl) and loaded onto 2% agarose gels. Electrophoresis was carried out for 30 min in a Mini-Sub® Cell GT electrophoresis cell (Bio Rad, Richmond, CA) containing tris-phosphate-EDTA (TPE) running buffer (36mM Tris-HCl, 30mM NaH2PO4, 10mM EDTA, pH 7.5) at 50 V. Gels were stained with ethidium bromide (0.1 μg/ml in water) for 30 min, and images were viewed and captured using a Vacutec Syngene G:Box gel documentation system, fitted with GeneSnap software, version 7.05.02 (Syngene, Cambridge, UK). Densitometric analysis was performed with the associated GeneTools software. The fluorescence intensities of unbound siRNA were expressed as a percentage against that of a naked siRNA control. The amount of liposome-associated siRNA at each MS09:siRNA (w/w) ratio was determined as follows:\n\nsiRNA was kept constant at 0.3µg. Chol = Cholesterol\n\nLipoplexes each containing 0.3 μg non-targeting siRNA were assembled at (w/w) ratios of 12:1-32:1. These were incubated with 10% FBS at 37°C for 4h. A control, 0.3 μg naked siRNA was treated similarly. Nuclease activity was terminated using EDTA (10mM), and complexes destabilized with 0.5% (w/v) SDS (55°C, 25 min). This was followed by electrophoresis as described previously.\n\nMCF-7 and HT-29 cells were seeded at densities of 4.0 × 104 cells/well in 48 well plates, and maintained at 37°C for 24h, followed by treatment with lipoplexes containing non-targeting siRNA (14 nM) at ratios used in the nuclease protection assay. Controls with naked siRNA (14 nM) were included. Lipofectamine™ 3000 (LF3K), prepared according to the manufacturer’s protocol was used as a positive control. The LF3K transfecting complex (25 μl) had a final siRNA concentration of 25 nM. At 48h post-transfection, growth medium was aspirated, and cell viability was assessed by the MTT assay. Cells were incubated (37°C, 4h) with 200 μl medium containing 20 μl MTT (5 mg/ml in PBS) per well. Wells were then drained, and formazan crystals dissolved in DMSO (200 μl/well) to produce purple-colored solutions. Absorbance was read at 540 nm in a Mindray microplate reader, MR 96A (Vacutec, Hamburg, Germany), against a DMSO blank. Percentage cell viability was calculated as follows:\n\nIn all ensuing experiments, MS09:Chol and MS09:DOPE lipoplexes (MS09:siRNA w/w = 16:1) contained 12 nM siRNA. MCF-7 and HT-29 cells were seeded in 6-well plates at densities of 3.0 ×105 cells/well and incubated as previously described. Cells were transfected with lipoplexes as previously described. Lipoplexes were assembled with 12 nM of either non-targeting siRNA or ON TARGETplus SMARTpool Human myc-siRNA (anti-c-myc-siRNA). Transfections were carried out in triplicate, followed by cell harvesting for total RNA and protein.\n\nTotal cellular RNA was extracted using TRIzol® Reagent, and cDNA synthesis was carried out using the gDNA Clear cDNA Synthesis Kit, as per manufacturer’s instructions. For a single reaction, 0.8 μg of the RNA sample was used. Reaction mixtures were prepared on ice and the reactions were carried out in a C1000 Touch™ Thermal Cycler (Bio-Rad Laboratories (PTY) Ltd., Richmond, USA). The reaction was performed as follows: DNA digestion (25 °C, 5 min), DNase inactivation (75 °C, 5 min), and samples were maintained at 4 °C for 10 min. The RT supermix (4 μl) was added and cDNA synthesis was carried out as follows: priming (25 °C, 5 min), reverse transcription (46 °C, 20 min), RT inactivation (95 °C, 1 min) and samples were held at 4 °C for 10 min. Two cDNA synthesis reactions per RNA isolate were performed in parallel i.e. one reaction containing the RT supermix and a no RT control in which the no-RT supermix was added instead. cDNA samples were diluted to a final concentration of 25 ng/μl in nuclease-free water and stored at 4 °C, for no more than a week. The product of each cDNA synthesis reaction was subjected to RT-qPCR. A single reaction mixture (20 μl) contained SsoAdvanced™ Universal SYBR® Green Supermix (10 μl); primers (1 μl) specific for either the target gene, c-MYC (PrimePCR™ SYBR® Green Assay: MYC, Human) or reference gene, β-actin (PrimePCR™ SYBR® Green Assay: ACTB, Human); cDNA sample (25 ng, 1μl) and nuclease-free water (8 μl). Reaction mixtures in which DNA templates (either PrimePCR™ Template for SYBR® Green Assay: MYC, Human; or PrimePCR™ Template for SYBR® Green Assay: ACTB, Human; 1 μl) were substituted for cDNA served as positive controls. Reaction mixtures where nuclease-free water (1 μl) was substituted for either primers or cDNA were included as negative controls. All reactions were performed in triplicate. Reaction mixtures were prepared, on ice, in Hard-Shell® 96 well plates, sealed, briefly centrifuged and then loaded in a C1000 Touch™ Thermal Cycler (CFX 96 Touch™ Real-Time PCR Detection System, Bio-Rad Laboratories (PTY) Ltd., Richmond, USA). Data was analyzed with CFX Manager™ Software version 3.0, and c-myc expression was normalized to β-actin using the ΔΔCq comparative quantification algorithm.\n\nMCF-7 and HT-29 cells were seeded in 6-well plates at densities of 3.0 ×105 cells/well and incubated as previously described. Cells were transfected with lipoplexes as previously described. The medium was removed, and cells washed twice with ice-cold PBS (1 ml/well). Cold (4°C) RIPA buffer (100 μl/well) was then added and cells placed on ice for 5 min with gentle shaking to dislodge cells. Cell suspensions were centrifuged (14 000 × g, 4°C, 15 min) and lysates/protein extracts were immobilized onto the wells of a 96-well plate with 50 mM carbonate-bicarbonate coating buffer (pH 9.6, at 4°C), overnight. Three replicates per isolate were performed. Each well received 10 μg protein in 100 μl coating buffer. The coating buffer was removed, and wells were rinsed twice with TBS (20 mM Tris-HCl, pH 7.5, 150 mM NaCl) containing 0.1% Tween 20 (TBST, 100 μl/well). Wells were then treated with 5% non-fat dry milk in TBST (100 μl) with gentle agitation to saturate unoccupied attachment sites. The blocking agent was removed, and wells rinsed twice with TBST (100 μl/well). Either c-myc (1:2000, in TBST) or β actin (1:10 000, in TBST) primary antibodies were added (100 μl/well) and incubated at room temperature for 1h. Primary antibodies were removed and wells washed with TBST (4x, 100 μl/well) for 5 min each, with agitation. The secondary antibody (1:2000, in TBST) was then added and incubated at room temperature for 1h. Wells were drained and washed with TBST, as previously. TMB (100 μl/well) was applied (room temperature, 30 min), and the reaction was terminated by the addition of 2M H2SO4 (100 μl/well). Absorbance was measured at 450 against a TMB (100 μl)/2M H2SO4 (100 μl) blank. Wells containing BSA (10 μg) served as negative controls. Antibody-free and substrate-free controls were included. Expression of c-myc was normalized to β actin and presented relative to untreated cells.\n\nLive, apoptotic and necrotic cells were distinguished by the AO/ethidium bromide (EtBr) dual staining method25. Lipoplexes were introduced to semi-confluent cells (8.0 ×104 cells/well) in 24-well plates as previously described. After 48h, cells were rinsed with PBS (200 μl/well), stained with AO/EtBr solution (100 μg/ml AO and 100 μg/ml EtBr in PBS; 10 μl/well). Excess stain was removed by rinsing with PBS (100 μl/well). Cellular changes associated with apoptosis were observed under an inverted fluorescence microscope (CKX41, Olympus, Japan) at excitation and emission wavelengths of 540 nm and 580 nm, respectively. Images were acquired at 200× magnification using Analysis Five Software (Olympus Soft Imaging Solutions, Olympus, Japan). The % live/apoptotic/early apoptotic/late apoptotic/necrotic cells were calculated as below:\n\nStatistical analyses were performed using one-way analysis of variance (ANOVA), followed by Tukey’s Multiple Comparison Test to compare between groups using GraphPad Prism version 5.04 (GraphPad Software Inc., USA). P values less than 0.05 were considered significant.\n\n\nResults\n\nTEM presented MS09:DOPE and MS09:Chol formulations as round to irregular shaped unilamellar vesicles, respectively (Figure 1a,b). The liposome-siRNA complexes (Figure 1c,d) assumed structures that were different from the vesicles, emphasizing the heterogeneity and assembly of the liposome-siRNA complexes.\n\n(a) MS09:DOPE, (b) MS09:Chol (c) MS09:DOPE:siRNA (d) MS09:Chol:siRNA. Bar = 200 nm\n\nNTA results (Table 3) show that MS09:DOPE and MS09:Chol liposome sizes were below 200 nm in size and may be classified as small unilamellar vesicles. The substitution of DOPE with Chol seems to have no significant effect on size or zeta potential, which remained high. It was observed that Chol is not likely to influence the electrical surface potential of liposomes because it does not bear an ionizable group26. Lipoplexes were less than 150 nm in size (Table 3), which is important for passive targeting of tumour cells via the enhanced permeability and retention effect.\n\nsiRNA was kept constant at 0.3 µg. SD n=3; Chol= cholesterol\n\nZeta potential measurements are based on the interaction of the particle with ions in the medium in which it is dispersed. In this study, liposomes were dispersed in HEPES buffer, which can influence the zeta potential of a bilayer depending on the orientation of the molecule’s ionizable groups relative to the membrane27,28 in the electrical double layer. Hence, the negative zeta potential values do not necessarily imply that the liposomes would be unable to associate with siRNA molecules. Lipoplexes similarly displayed negative zeta potentials (Table 3). Although it is accepted that the net positive charge of lipoplexes allows for binding to anionic membrane-associated proteoglycans to initiate cellular uptake29, it is also possible for siRNA lipoplexes with negative zeta potential to enter cells and successfully facilitate gene silencing.\n\nThe gel retardation assay or band shift assay is widely documented as the first step in assessing the siRNA-binding ability of cationic carriers30–32. This assay is based on the premise that the migration of siRNA is retarded in an electric field when bound to a carrier due to the formation of electroneutral complexes that are unable to permeate the gel matrix. MS09:DOPE was capable of fully preventing the migration of siRNA (Figure 2a), whereas Figure 2b shows that although there was a decrease in unbound siRNA with increasing liposome content for the MS09:Chol formulation, unbound siRNA was evident at all MS09:siRNA (w/w) ratios. Densitometric analysis (Figure 3) confirmed that the binding of siRNA by these liposome formulations increased until a point, whereupon free siRNA in the gel persisted despite addition of liposome. This MS09:siRNA (w/w) ratio was taken as the end-point/optimum binding ratio of these liposomes.\n\nIncubation mixtures (10 μl) contained siRNA (0.3 μg) and varying amounts of liposome corresponding with increasing amounts of cytofectin. a) MS09:DOPE and b) MS09:Chol.\n\nData is presented as the mean ± SD (n = 3). The point at which each formulation best bound siRNA is indicated by an arrow. ●●●P < 0.001 vs. DOPE-containing counterpart at the respective MS09:siRNA (w/w) ratio at which maximum siRNA was bound.\n\nAttempts at lipid-mediated siRNA delivery are often frustrated by adverse interactions with serum. Figure 4 shows that, while naked siRNA was entirely degraded under experimental conditions (lanes 2), intact siRNA bands, less intense than the untreated control, were clearly visible in all instances. This shows that the liposomal formulations partially protected siRNA at the respective MS09:siRNA (w/w) ratios. Densitometric analysis of gels (Figure 5) provided further insight into the siRNA-protecting capabilities of the liposomes. Maximum intact siRNA of 73% afforded by MS09:Chol liposomes at the MS09:siRNA (w/w) ratio of 24:1 was significantly less than that of MS09:DOPE. 25% of siRNA was likely to be surface associated, as it was so loosely bound that it was coaxed off during electrophoresis. Such siRNA is readily detached from the liposomal bilayer upon exposure to serum nucleases33.\n\nIn each gel, lane 1 contained undigested siRNA, lane 2, serum-digested siRNA and lanes 3–8, serum-exposed lipoplexes at varying MS09:siRNA (12:1 – 32:1 w/w) ratios.\n\nIntact siRNA associated with lipoplexes was quantified and expressed as a percentage of untreated siRNA. Data is presented as the mean ± SD (n = 3). ●●P < 0.01, ●●●P < 0.001 vs. DOPE-containing counterpart.\n\nIt was important that any growth inhibitory effects in the cancer cells be attributed solely to the delivered siRNA, and not due to any intrinsic harmful effect of the liposomal carrier. This assay is based on the principle that enzymes of the mitochondria of living cells reduce soluble MTT, a tetrazolium salt, to formazan crystals34. The intensity of the resultant purple solution correlates with the extent of MTT reduction, and the number of viable cells35. No significant reduction in viability was detected upon treatment of all cells with naked siRNA at concentrations contained in lipoplexes (Figure 6). Cells retained viability of at least 88% with exposure to LF3K-siRNA complexes. In general, cell survival after exposure to the liposomal formulations were greater than 85% with no severe cytotoxicity evident.\n\na) MCF 7 and b) HT-29 cells. Each column represents the mean ± SD (n = 3). P > 0.05 vs. untreated cells and DOPE containing counterparts. LF3K denotes Lipofectamine™ 3000.\n\nGiven that the initial RNAi effect is exerted at the mRNA level, the effect of transfection with anti-c-myc lipoplexes on c-myc transcripts in cancer cells was studied using RT-qPCR. Figure 7 shows a decrease in c-myc-mRNA in instances where a transfecting agent was used to deliver anti-c-myc-siRNA sequences. Quantification of cellular c-myc protein by ELISA (Figure 8) showed that a decrease in c-myc mRNA levels was, in all cases, accompanied by a concomitant reduction in protein expression. Complexes assembled with non-targeting siRNA were without effect. This confirmed that the observed reduction in c-myc-mRNA contributed to the RNAi effect of the delivered anti-c-myc-siRNA. The fact that naked anti-c-myc-siRNA did not influence c-myc expression in any way, highlights the need for a delivery vehicle.\n\nc-myc expression was quantified by RT-qPCR and normalized to the β-actin reference gene. Each column represents the mean ± SD (n = 3). *P < 0.05, ***P < 0.001 vs. naked siRNA; ♦P < 0.05, ♦♦♦P < 0.001 vs. non-targeting siRNA; ##P < 0.01, ###P < 0.001 vs. anti c-myc LF3K. P > 0.05, with respect to anti c-myc MS09:Chol vs. anti-c-myc MS09:DOPE. LF3K= Lipofectamine™ 3000\n\nc-myc expression was quantified by ELISA, and normalized to the internal control, β-actin. Each column represents the mean ± SD (n = 3). *P < 0.05, ***P < 0.001 vs. naked siRNA; ♦P < 0.05, ♦♦♦P < 0.001 vs. non-targeting siRNA; #P <0.05, ##P < 0.01, ###P < 0.001 vs. anti c-myc LF3K. P > 0.05, with respect to anti c-myc MS09:Chol vs. anti-c-myc MS09:DOPE. LF3K= Lipofectamine™ 3000\n\nMS09:Chol and MS09:DOPE lipoplexes produced significant gene silencing compared to LF3K in both cell lines. In the MCF-7 cell line, MS09 lipoplexes achieved 8- and 3.5-fold greater knockdown of c-myc than LF3K at the mRNA and protein levels, respectively. In HT-29 cells, the decrease in c-myc-mRNA and protein was 5- and 2.8-fold more for MS09 lipoplexes. The superior performance of MS09 lipoplexes is highlighted by the fact that these contained at half the final siRNA concentration as LF3K.\n\nA desirable feature is for cancer treatment to induce cancer cell death without causing harm to surrounding healthy tissue. Hence, several anticancer approaches exploit the natural mechanisms of cell death such as apoptosis, that normally eliminates damaged and/or harmful cells in a regulated fashion36. The AO/EtBr method is based on the principle that AO enters cells with intact plasma membranes and binds to DNA to emit green fluorescence, while EtBr enters cells with defective membrane integrity and fluoresces red-orange when bound to DNA. Differentiation between normal, early apoptotic, late apoptotic and necrotic cells was made based on observations of nuclear morphology (Figure 9 and Figure 10). Live cells were characterized by a bright green nucleus in the center of the cell. The nuclei of early apoptotic cells, with undamaged membranes, also stained green, but appeared to be fragmented or condensed. In contrast, the nuclei of late apoptotic cells, with compromised membrane integrity, stained orange with evidence of fragmentation or condensation. Finally, necrotic cells were characterized by an intact bright orange nucleus37.\n\na) Control, no treatment, (b) naked anti-c-myc-siRNA, (c) LF3K:siRNA, (d) MS09:DOPE:siRNA and e) MS09:Chol:siRNA. Scale Bar = 100 μm. AO = Acridine Orange, EtBr = Ethidium Bromide, LF3K= Lipofectamine™ 3000, L= live, EA = early apoptotic, LA = late apoptotic, N = necrotic.\n\na) Control, no treatment, (b) naked anti-c-myc-siRNA, (c) LF3K:siRNA, (d) MS09:DOPE:siRNA and e) MS09:Chol:siRNA. Scale Bar = 100 μm. AO = Acridine Orange, EtBr = Ethidium Bromide, LF3K= Lipofectamine™ 3000, L= live, EA = early apoptotic, LA = late apoptotic, N = necrotic.\n\nThe major mechanism of cell death observed with these anti-c-myc lipoplexes was apoptosis, which is similar to other studies demonstrating that inhibition of c-myc in cancer cells leads to apoptosis38–40. Importantly, necrosis, a non-specific form of cell death that is associated with an inflammatory response41, was negligible in all instances, accounting for less than 3% of total cells per sample. Hence, MS09:Chol and MS09:DOPE-mediated anti-c-myc-siRNA delivery is capable of destroying cancer cells without damaging healthy tissue. The application of anti-c-myc-siRNA on its own did not result in any anticancer activity. Hence we can conclude that the anticancer effects can be ascribed to the RNAi activity of liposome bound anti-c-myc-siRNA.\n\n\nDiscussion\n\nMS09, a cationic lipid comprising three structural domains viz. a hydrophobic cholesteryl anchor and a polar dimethylammonium head group, separated by a 15Å spacer arm, was originally co-formulated with DOPE for the delivery of DNA23 and siRNA42 into mammalian cells. DOPE is a commonly used helper lipid in cationic liposome formulations43,44, but may not be suitable for intravenous administration. Early studies showed that the incorporation of Chol with phospholipids at 30 mol% or more resulted in the formation of a phase-separated region in the lipid bilayer45. This property of Chol liposomes, in the absence of other helper lipids, becomes more pronounced and prevents adverse liposome-protein interactions, aggregation, improves mechanical strength and stability24,45,46, and extending circulation time in vivo47. It was reported recently that folate receptor-targeted cholesterol-rich liposomes readily form lipoplexes with Bmi1 siRNA that inhibit tumor growth both in vitro and in vivo48. Furthermore, DNA in cholesterol-rich plasmid DNA lipoplexes was afforded full protection from enzymatic degradation and showed good efficacy in a CRISPR-Cas9 application in HEK293 cells49.\n\nTEM revealed that the liposomes were unilamellar, spherical vesicles that were capable of forming lipoplexes with the siRNA. Both the liposomes and lipoplexes were less than 150 nm in size, favoring targeting as nanoparticles of a suitable size will not pass through the tight junctions of normal blood vessels but can access tumor cells by passing through their more permeable vasculature and are retained because of reduced lymphatic drainage50. These properties are valuable as determinants of lipoplex performance because they impact on the circulation time of lipoplexes in the body, accumulation at target sites, interaction with cells, the efficacy of cellular uptake and, gene silencing activity51. The negative zeta potential of the liposomes and lipoplexes did not affect their cellular uptake. It was reported that siRNA lipoplexes with negative zeta potentials were internalized by breast cancer cells via endocytosis and, that cellular uptake was dependent upon the activity of microtubules and actin52. Negatively charged lipid-based siRNA nanocomplexes may be useful as they can avoid aggregation through interaction with erythrocytes and anionic proteins in biological fluids and have also been associated with lower toxicities than complexes carrying a net positive charge53,54.\n\nMS09:DOPE and MS09:Chol liposomes showed good siRNA binding, but replacing DOPE with Chol at the same molar ratio in MS09 formulations weakened the siRNA interaction with MS09:Chol. It is possible that Chol may have induced arrangement of cytofectin molecules during vesicle formation such that a greater number of cationic centers were positioned inwards rather than on the surface of the bilayer. A further explanation may arise from the fact that Chol is a more rigid lipid than DOPE and results in liposomal bilayers that are less malleable. Therefore, during the process of lipoplex formation, the Chol-containing liposomes may not change conformation as easily as their DOPE-containing counterparts to completely encompass siRNA molecules. An early study with DOTAP/Chol liposomes, also demonstrated that Chol widened the interlamellar spaces of the resultant lipoplexes55.\n\nBoth liposomes were capable of protecting the siRNA against nuclease degradation and were well tolerated by the MCF-7 and HT-29 cell lines. Cytotoxicity testing was performed under the same conditions as those employed in gene silencing experiments, except that lipoplexes were assembled with non-targeting siRNA so as to rule out the possibility of cell death due to silencing of any functional genes. In general, cell survival after exposure to the MS09:Chol formulation was better than its DOPE-containing counterpart. This could be attributed to Chol as an endogenous lipid, and that DOPE is pH-sensitive while Chol is not. This property of DOPE, which is regarded as important for transfection, has also been associated with toxicity, because it causes destabilization of lysosomes and release of debris within the cell56.\n\nGene silencing experiments revealed that both MS09 liposomal formulations exhibited enhanced c-myc gene silencing in the MCF-7 and HT-29 cell lines, superior to that of LF3K. It was observed that c-myc inhibition at both levels of expression by all transfecting agents was more pronounced in the MCF-7 cells than in the HT-29 cells (P<0.05). This is underscored by the fact that HT-29 cells are considerably more difficult to transfect than other human cell lines57,58. This was also evident for the LF3K reagent with a markedly lower oncogene knockdown (1.4-fold reduction). In HT-29 cells cellular uptake of MS09:Chol and MS09:DOPE lipoplexes was comparable (P>0.05). The similar reduction in c-myc-mRNA and protein implied that the MS09 lipoplexes facilitated RISC-engagement of intact anti-c-myc-siRNA molecules with near-equal efficiency. However, in the MCF-7 cells, the MS09:Chol lipoplex, at 12 nM siRNA, facilitated effective siRNA delivery compared to DOPE-containing lipoplexes (P<0.05), but without a more pronounced gene silencing effect. This could be attributed to the possibility of saturation of the RNAi machinery, especially since this complex possibly releases siRNA directly into the cytoplasm. RISC saturation within a broad siRNA concentration range of 5-100 nM has been reported and is dependent upon the potency of the siRNA molecules involved42,59. Given the catalytic nature of siRNA activity, the results suggest that the MS09:Chol lipoplex could provide effective gene silencing at final siRNA concentrations below 12 nM in MCF-7 cells. For comparative evaluation as anti-c-myc agents, both MS09:Chol and MS09:DOPE lipoplexes were tested at the same MS09:siRNA (w/w) ratio, final lipid and siRNA concentration. Although the siRNA binding and protecting capability of MS09:Chol was shown to be weaker than its DOPE-containing counterpart at the MS09:siRNA (w/w) ratio of 16:1, the MS09:Chol lipoplex was proven to be as effective an siRNA carrier and c-myc-silencing agent. This could be attributed to the role of cholesterol nanodomains in transfection60, and Chol-mediated fusion with the cell membrane as a mode of delivering intact siRNA directly to the RNAi apparatus in the cytoplasm61.\n\nA significant decrease in c-myc expression was correlated with anticancer effects following apoptosis analysis, which included inhibition of cancer cell migration, loss of cell viability and elimination of cancer cells through apoptosis, with the exception of anti-c-myc-LF3K in HT-29 cells. Here c-myc inhibition was too low to induce significant apoptosis and reduce cancer cell numbers. Overall, both anti-c-myc lipoplexes produced better anticancer activity than LF3K, in a given cell line. Furthermore, the comparable gene silencing activity mediated by MS09:Chol and MS09:DOPE lipoplexes was coupled with anticancer effects of near-equal potency. Although c-myc inhibition with MS09 lipoplexes was more pronounced (3.8- and 2.8-fold differences at the mRNA and protein levels, respectively) in MCF-7 cells than in HT-29 cells, their associated impact on cell migration, growth and apoptosis induction were similar. This could be due to the level of c-myc inhibition required to elicit anticancer activity differs among cell lines, or that more potent oncogene knockdown does not necessarily correspond with enhanced anticancer activity.\n\nOverall, our study has shown that effective c-myc gene silencing was achieved with both MS09 lipoplexes. Gene silencing experiments with both liposomal formulations showed superior gene silencing compared to Lipofectamine™ 3000 in vitro.\n\n\nConclusion\n\nThe findings from this study show the potential of MS09-based siRNA cholesterol-rich lipoplexes for the effective silencing of the c-myc oncogene in the MCF-7 and HT-29 cells. This co-formulation of MS09 and Chol for siRNA delivery and its application in c-myc gene silencing was not previously explored but was shown to be an effective c-myc silencing agent with results comparable to its DOPE-containing counterpart. On a positive note, the fact that both MS09 lipoplexes have performed better in a recalcitrant cell line than the standard transfection reagent, confirms their applicability as oncogene silencing agents in difficult-to-transfect cancer cells, adding credence to their potential as broad-range anti-c-myc agents. Overall, these novel MS09:Chol liposomes are endowed with physicochemical properties that may render them more suitable for in vivo siRNA delivery than their MS09:DOPE counterparts, and that their anti-c-myc-siRNA lipoplexes should be developed further as a posttranscriptional intervention treatment modality for mono- and polygenic human diseases.\n\n\nData availability\n\nZenodo: Anti-c-myc cholesterol based lipoplexes as onco-nanotherapeutic agents in vitro, http://doi.org/10.5281/zenodo.394664062.\n\nThis project contains the following underlying data:\n\nUncropped/unedited electron microscopy images showing morphology of liposomes and siRNA lipoplexes\n\nRaw data for particle size and zeta potential of liposomes and siRNA lipoplexes\n\nUncropped/unedited gel retardation assay electrophoresis images; and raw data for densitometric analysis of the gel retardation assays\n\nUncropped/unedited nuclease protection assay electrophoresis images; and raw data for densitometric analysis of the nuclease protection assays\n\nRaw data for absorbance readings for MTT assays for MCF-7 and HT-29 cells\n\nRaw data for RT-qPCR quantifications for expression of c-myc and β-actin\n\nRaw data for absorbance readings for c-myc protein expression and b-actin using ELISA\n\nUncropped/unedited apoptosis images; and raw data for percentage of live/apoptotic/early apoptotic/late apoptotic/necrotic cells calculated\n\nZenodo: Anti-c-myc cholesterol based lipoplexes as onco-nanotherapeutic agents in vitro, http://doi.org/10.5281/zenodo.394042662.\n\nThis project contains the following extended data:\n\nNumber of lipid molecules that constitute liposomal vesicles\n\nSize and size distribution of liposomes and lipoplexes by NTA summarized data\n\nZeta potential and zeta potential distribution of liposomes and lipoplexes by Z-NTA summarized data\n\nEstimated number of liposomal vesicles and siRNA molecules per liposome-siRNA nanocomplex\n\nFlow profiles for liposome suspensions and lipoplexes\n\nZeta potential and size vs. concentration graphs for liposome suspensions and lipolexes\n\nMiscellaneous calculations: estimation of the average number of lipid molecules per vesicle; N/P (+/) charge ratio; final siRNA concentration; final cytofectin and lipid concentration; estimation of average number of vesicles/nanocomplex; estimation of average number of siRNA molecules/nanocomplex\n\nSet-up for gel retardation and nuclease digestion assays\n\nAdditional cytotoxicity data\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Acknowledgements\n\nThe authors acknowledge colleagues of the Nano-Gene and Drug Delivery group for technical support.\n\n\nReferences\n\nGlobal Cancer Facts and Figures. 3rd ed. Atlanta, Georgia: American Cancer Society. 2015. Reference Source\n\nDesantis CE, Lin CC, Mariotto AB, et al.: Cancer treatment and survivorship statistics. CA Cancer J Clin. 2014; 64(4): 252–71. PubMed Abstract | Publisher Full Text\n\nGarcía-Gutièrrez L, Delgado MD, Lèon J: MYC oncogene contributes to release of cell cycle brakes. Genes. 2019; 10(3): 244. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHermeking H: The MYC oncogene as a cancer drug target. Curr Cancer Drug Targets. 2003; 3(3): 163–75. PubMed Abstract | Publisher Full Text\n\nMorrish F, Giedt C, Hockenbery D: C-myc apoptotic function is mediated by NRF-1 target genes. Genes Dev. 2003; 17(2): 240–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWilson A, Murphy MJ, Oskarsson T, et al.: C-myc controls the balance between hematopoietic stem cell self-renewal and differentiation. Genes Dev. 2004; 18(22): 2747–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim JW, Zeller KI, Wang Y, et al.: Evaluation of myc e-box phylogenetic footprints in glycolytic genes by chromatin immunoprecipitation assays. Mol Cell Biol. 2004; 24(13): 5923–36. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGebhardt A, Frye M, Herold S, et al.: Myc regulates keratinocyte adhesion and differentiation via complex formation with miz1. J Cell Biol. 2006; 172(1): 139–49. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVan Riggelen J, Yetil A, Felsher DW: Myc as a regulator of ribosome biogenesis and protein synthesis. Nat Rev Cancer. 2010; 10(4): 301–9. PubMed Abstract | Publisher Full Text\n\nDaniels AN, Singh M: Sterically stabilized siRNA:gold nanocomplexes enhance c-MYC silencing in a breast cancer model. Nanomedicine (Lond). 2019; 14(11): 1387–401. PubMed Abstract | Publisher Full Text\n\nSingh A, Trivedi P, Jain NK: Advances in siRNA delivery in cancer therapy. Art Cells Nanomed Biotech. 2018; 46(2): 274–83. PubMed Abstract | Publisher Full Text\n\nHuang Y, Hong J, Zheng S, et al.: Elimination pathways of systemically delivered siRNA. Mol Ther. 2011; 19(2): 381–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAkhtar S, Benter I: Toxicogenomics of non-viral drug delivery systems for RNAi: Potential impact on siRNA-mediated gene silencing activity and specificity. Adv Drug Deliv Rev. 2007; 59(2–3): 164–82. PubMed Abstract | Publisher Full Text\n\nHope MJ: Enhancing siRNA delivery by employing lipid nanoparticles. Ther Deliv. 2014; 5(6): 663–73. PubMed Abstract | Publisher Full Text\n\nLeung A, Tam Y, Cullis PR: Lipid nanoparticles for short interfering RNA delivery. Adv Genet. 2014; 88: 71–110. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSingh Y, Tomar S, Khan S, et al.: Bridging small interfering RNA with giant therapeutic outcomes using nanometric liposomes. J Control Release. 2015; 220(Pt A): 368–87. PubMed Abstract | Publisher Full Text\n\nFelgner PL, Gadek TR, Holm M, et al.: Lipofection: A highly efficient, lipid-mediated DNA-transfection procedure. Proc Natl Acad Sci U S A. 1987; 84(21): 7413–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChatin B, Mèvel M, Devallieŕe J, et al.: Liposome- based formulation for intracellular delivery of functional proteins. Mol Ther Nucl Acids. 2015; 4: E244. PubMed Abstract | Publisher Full Text\n\nShim G, Kim MG, Park JY, et al.: Application of cationic liposomes for delivery of nucleic acids. Asian J Pharm Sci. 2013; 8(2): 72–80. Publisher Full Text\n\nLi S, Tseng WC, Stolz DB, et al.: Dynamic changes in the characteristics of cationic lipidic vectors after exposure to mouse serum: implications for intravenous lipofection. Gene Ther. 1999; 6(4): 585–94. Publisher Full Text\n\nBozzuto G, Molinari A: Liposomes as nanomedical devices. Int J Nanomed. 2015; 10: 975–99. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLambardo D, Kiselev MA, Caccamo MT: Smart nanoparticles for drug delivery application: development of versatile nanocarrier platforms in biotechnology and nanomedicine. J Nanomater. 2019; 2019: 3702518. Publisher Full Text\n\nSingh M, Ariatti M: A cationic cytofectin with long spacer mediates favourable transfection in transformed human epithelial cells. Int J Pharm. 2006; 309(1–2): 189–98. PubMed Abstract | Publisher Full Text\n\nMontiero N, Martins A, Reis RL, et al.: Liposomes in tissue engineering and regenerative medicine. J R Soc Interface. 2014; 11(101): 20140459. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaiyo FC, Moodley R, Singh M: Cytotoxicity, Antioxidant and Apoptosis Studies of Quercetin-3-O Glucoside and 4-(β-D-Glucopyranosyl-1→4-α-L-Rhamnopyranosyloxy)-Benzyl Isothiocyanate from Moringa oleifera. Anticancer Agents Med Chem. 2016; 16(5): 648–56. PubMed Abstract | Publisher Full Text\n\nYang SY, Zheng Y, Chen JY, et al.: Comprehensive study of cationic liposomes composed of DC-Chol and cholesterol with different mole ratios for gene transfection. Colloids Surf B Biointerfaces. 2013; 101: 6–13. PubMed Abstract | Publisher Full Text\n\nMislick KA, Baldeschwieler JD: Evidence for the role of proteoglycans in cation-mediated gene transfer. Proc Natl Acad Sci U S A. 1996; 93(22): 12349–54. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKoerner MM, Palacio LA, Wright JW, et al.: Electrodynamics of lipid membrane interactions in the presence of zwitterionic buffers. Biophys J. 2011; 101(2): 362–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCeballos C, Khiati S, Prata CA, et al.: Cationic nucleoside lipids derived from universal bases: a rational approach for siRNA transfection. Bioconjug Chem. 2010; 21(6): 1062–9. PubMed Abstract | Publisher Full Text\n\nHan SE, Kang H, Shim G, et al.: Novel cationic cholesterol derivative-based liposomes for serum-enhanced delivery of siRNA. Int J Pharm. 2008; 353(1–2): 260–9. PubMed Abstract | Publisher Full Text\n\nSuh MS, Shim G, Lee HY, et al.: Anionic amino acid-derived cationic lipid for siRNA delivery. J Control Release. 2009; 140(3): 268–76. PubMed Abstract | Publisher Full Text\n\nDorasamy S, Narainpersad N, Singh M, et al.: Novel targeted liposomes deliver siRNA to hepatocellular carcinoma cells in vitro. Chem Biol Drug Design. 2012; 80(5): 647–56. PubMed Abstract | Publisher Full Text\n\nBuyens K, Lucas B, Raemdonck K, et al.: A fast and sensitive method for measuring the integrity of siRNA-carrier complexes in full human serum. J Control Release. 2008; 126(1): 67–76. PubMed Abstract | Publisher Full Text\n\nStockert JC, Blázquez-Castro A, Cañete M, et al.: MTT assay for cell viability: Intracellular localization of the formazan product is in lipid droplets. Acta Histochem. 2012; 114(8): 785–96. PubMed Abstract | Publisher Full Text\n\nFotakis G, Timbrell JA: In vitro cytotoxicity assays: comparison of LDH, neutral red, MTT and protein assay in hepatoma cell lines following exposure to cadmium chloride. Toxicol Lett. 2006; 160(2): 171–7. PubMed Abstract | Publisher Full Text\n\nBaig S, Seevasant I, Mohamad J, et al.: Potential of apoptotic pathway-targeted cancer therapeutic research: where do we stand? Cell Death Dis. 2016; 7(1): E2058. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRibble D, Goldstein NB, Norris D, et al.: A simple technique for quantifying apoptosis in 96-well plates. BMC Biotechnol. 2005; 5: 12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang Y, Liu S, Zhang G, et al.: Knockdown of c-myc expression by RNAi inhibits MCF-7 breast tumor cells growth in vitro and in vivo. Breast Cancer Res. 2005; 7(2): R220–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang X, Ge YL, Tian RH: The knockdown of c-myc expression by RNAi inhibits cell proliferation in human colon cancer HT-29 cells in vitro and in vivo. Cell Mol Biol Lett. 2009; 14(2): 305–18. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen Y, Wu JJ, Huang L: Nanoparticles targeted with NGR motif deliver c-Myc siRNA and doxorubicin for anticancer therapy. Mol Ther. 2010; 18(4): 828–34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKasibhatla S, Tseng B: Why target apoptosis in cancer treatment? Mol Cancer Ther. 2003; 2(6): 573–80. PubMed Abstract\n\nDaniels A, Singh M, Ariatti M: Pegylated and non-pegylated siRNA lipoplexes formulated with cholesteryl cytofectins promote efficient luciferase knockdown in Hela tat luc cells. Nucleosides Nucleotides Nucleic Acids. 2013; 32(4): 206–20. PubMed Abstract | Publisher Full Text\n\nMochizuki S, Kanegae N, Nishina K, et al.: The role of the helper lipid dioleoylphosphatidylethanolamine (DOPE) for DNA transfection cooperating with a cationic lipid bearing ethylenediamine. Biochim Biophys Acta. 2013; 1828(2): 412–8. PubMed Abstract | Publisher Full Text\n\nHuang CH, Sipe JP, Chow ST, et al.: Differential interaction of cholesterol with phosphatidylcholine on the inner and outer surfaces of lipid bilayer vesicles. Proc Natl Acad Sci U S A. 1974; 71(2): 359–62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaliwal SR, Paliwal R, Vyas SP: A review of mechanistic insight and application of pH-sensitive liposomes in drug delivery. Drug Deliv. 2015; 22(3): 231–42. PubMed Abstract | Publisher Full Text\n\nEpand RM, Hughes DW, Sayer BG, et al.: Novel properties of cholesterol-dioleoylphosphatidylcholine mixtures. Biochim Biophys Acta. 2003; 1616(2): 196–208. PubMed Abstract | Publisher Full Text\n\nSemple SC, Chonn A, Cullis PR: Influence of cholesterol on the association of plasma proteins with liposomes. Biochemistry. 1996; 35(8): 2521–5. PubMed Abstract | Publisher Full Text\n\nLi W, Yan R, Liu Y, et al.: Co-delivery of Bmi1 small interfering RNA with ursolic acid by folate receptor-targeted cationic liposomes enhances anti-tumor activity of ursolic acid in vitro and in vivo. Drug Deliv. 2019; 26(1): 794–802. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSadat-Hosseini E, Nikkhah M, Hosseinkhani S: Cholesterol-rich lipid-mediated nanoparticles boost of transfection efficiency, utilized for gene editing by CRISPR-Cas9. Int J Nanomedicine. 2019; 14: 4353–66. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGreish K: Enhanced permeability and retention (EPR) effect for anticancer nanomedicine drug targeting. Methods Mol Biol. 2010; 624: 25–37. PubMed Abstract | Publisher Full Text\n\nSchroeder A, Levins CG, Cortez C, et al.: Lipid-based nanotherapeutics for siRNA delivery. J Int Med. 2010; 267(1): 9–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKapoor M, Burgess DJ: Cellular uptake mechanisms of novel anionic siRNA lipoplexes. Pharm Res. 2013; 30(4): 1161–75. PubMed Abstract | Publisher Full Text\n\nHattori Y, Nakamura A, Arai S, et al.: In vivo siRNA delivery system for targeting to the liver by poly-l-glutamic acid-coated lipoplex. Results Pharma Sci. 2014; 4: 1–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTagalakis AD, Do Hyang DL, Bienemann AS, et al.: Multifunctional, self-assembling anionic peptide-lipid nanocomplexes for targeted siRNA delivery. Biomater. 2014; 35(29): 8406–15. PubMed Abstract | Publisher Full Text\n\nWeisman S, Hirsch-Lerner D, Barenholz Y, et al.: Nanostructure of cationic lipid-oligonucleotide complexes. Biophys J. 2004; 87(1): 609–14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFilion MC, Phillips NC: Toxicity and immunomodulatory activity of liposomal vectors formulated with cationic lipids toward immune effector cells. Biochim Biophys Acta. 1997; 1329(2): 345–56. PubMed Abstract | Publisher Full Text\n\nAlameh M, Jean M, Dejesus D, et al.: Chitosanase-based method for RNA isolation from cells transfected with chitosan/siRNA nanocomplexes for real-time RT-PCR in gene silencing. Int J Nanomed. 2010; 5: 473. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCerda MB, Batalla M, Anton M, et al.: Enhancement of nucleic acid delivery to hard-to-transfect human colorectal cancer cells by magnetofection at laminin coated substrates and promotion of the endosomal/lysosomal escape. RSC Adv. 2015; 5: 58345–54. Publisher Full Text\n\nMaiyo F, Singh M: Polymerized Selenium Nanoparticles for Folate-Receptor-Targeted Delivery of Anti-Luc-siRNA: Potential for Gene Silencing. Biomedicines. 2020; 8(4): 76. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXu L, Anchordoquy TJ: Effect of cholesterol nanodomains on the targeting of lipid-based gene delivery in cultured cells. Mol Pharm. 2010; 7(4): 1311–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPozzi D, Marchini C, Cardarelli F, et al.: Transfection efficiency boost of cholesterol-containing lipoplexes. Biochim Biophys Acta Biomembr. 2012; 1818(9): 2335–43. PubMed Abstract | Publisher Full Text\n\nSaffiya H, Aliscia D, Mario A, et al.: Anti-c-myc cholesterol based lipoplexes as onco-nanotherapeutic agents in vitro. Zenodo. 2020. http://www.doi.org/10.5281/zenodo.3946640"
}
|
[
{
"id": "67956",
"date": "06 Aug 2020",
"name": "Patrick Arbuthnot",
"expertise": [
"Reviewer Expertise Gene therapy."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript entitled ‘Anti-c-myc cholesterol based lipoplexes as onco-nanotherapeutic agents in vitro’ describes an analysis of non-viral lipid nanoparticles for the delivery of anti-c-myc siRNAs to cultured breast cells. The formulations comprising a cholesterol derivative termed MS09 in combination with DOPE or cholesterol performed significantly better than a commercially available transfection reagent. Useful properties of the complexes are described and extensive characterization is provided. The manuscript is written well, the study is neat and should be of interest to researchers in the field. Two points that should be addressed before indexing are the following:\nEmphasis is placed on targeting cancer cells for delivery of an oncogene-disabling siRNA. The authors should provide information on how the vectors will/could be targeted to malignant cells specifically.\n\nUltimate utility of the vectors will be dependent on efficient action in vivo. Considerations for clinical translation of the technology should be provided.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6238",
"date": "05 Jan 2021",
"name": "Moganavelli Singh",
"role": "Author Response",
"response": "Emphasis is placed on targeting cancer cells for delivery of an oncogene-disabling siRNA. The authors should provide information on how the vectors will/could be targeted to malignant cells specifically. The issue of tumour cell targeting has been included in the discussion: “ Others have reported ligand modification … … … more easily and economically produced” Ultimate utility of the vectors will be dependent on efficient action in vivo. Considerations for clinical translation of the technology should be provided. The following has been included in the discussion: “It is worth mentioning that … … … a clearer indication of its clinical applicability.”"
}
]
},
{
"id": "74490",
"date": "23 Nov 2020",
"name": "Kaushik Pal",
"expertise": [
"Reviewer Expertise Chemistry",
"Chemical biology",
"biophysics",
"fluorescence microscopy"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this article, Saffiya Habib et al. presented the synthesis of Cholesterol, MS09 and DOPE-based composite liposome and applied that as the oncogene (c-myc) siRNA carrier. After synthesis, they have characterized the liposome with TEM, gel retardation assay, and checked efficacy with nuclease protection study, cell viability assay, gene silencing, and apoptosis assay. However, I feel this article can be only accepted after major revision. My specific comments are as follows:\nThe application of the liposome as the siRNA carrier is not new even from this very group this thing has been presented, so what is the advancement here? Is it the same platform with just a change in the siRNA? It needs to be clarified. Here my suggestion will be comparing the usefulness of the present and previously reported result for the same siRNA. The Introduction of this article needs to be thoroughly revised. The research work done here and the introduction does not complement each other. The introduction is mostly talking about the in vivo challenges but the work is done in vitro. In the Introduction, the authors claim that the main difficulty of this kind of carrier is the stability of itself and in the blood serum but there is no mechanistic clarification on how these liposomes are stable in solution and under serum spiked condition. TEM image of the liposome is not clear and even in a small field of view the heterogeneity is not at all close to the state of the art1. They should provide TEM images of the large field, size distribution (with the sampling number), etc. The liposome after complexing with siRNA (lipoplex) looks like an aggregate structure rather than a single structure, this must be rectified. In the Abstract author specify the liposome as bilamellar but in the Results section that stated/appears to be unilamellar. Also, the size said in the abstract is less than 200 nm which far away from the result. The author should be very objective in the abstract itself. The comparison of the liposome is done with Lipofectamine™ 3000, which is not the commercial benchmark for the siRNA transfection. They should provide efficacy compare to the Lipofectamine™ RNAiMAX in all the cases. Most of the figure captions are inadequate, they must be more informative. The subfigures are not labeled, like Figure 2, it is very difficult to understand which well is loaded with what.\nOverall, the whole manuscript is needed to be revised in order to make the Abstract, Introduction, Result, and Discussion to make completely to each other.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "6239",
"date": "05 Jan 2021",
"name": "Moganavelli Singh",
"role": "Author Response",
"response": "The application of the liposome as the siRNA carrier is not new even from this very group this thing has been presented, so what is the advancement here? Is it the same platform with just a change in the siRNA? It needs to be clarified. Here my suggestion will be comparing the usefulness of the present and previously reported result for the same siRNA. The use of liposomes as siRNA carriers has been previously reported. The novelty here lies in the formulation and its application. Firstly, the cytofectin MS09 has not previously been investigated in coformulation with cholesterol. Secondly, MS09 siRNA lipoplexes were previously investigated in serum-deficient conditions. Thirdly, the MS09:Chol:anti-c-myc siRNA system is introduced as an uncomplicated, easy-to-prepare c-myc-silencing, anti-cancer agent. The Introduction of this article needs to be thoroughly revised. The research work done here and the introduction does not complement each other. The introduction is mostly talking about the in vivo challenges but the work is done in vitro. Although this study is exclusively in vitro, the in vitro experiments were performed as a first step towards preparing and developing a simple c-myc silencing lipid-based nanosystem that will ultimately be applicable in vivo. For this reason, considerations towards designing a clinically viable lipoplex were extensively outlined in the introduction as these provided a framework for the reported study. In the Introduction, the authors claim that the main difficulty of this kind of carrier is the stability of itself and in the blood serum but there is no mechanistic clarification on how these liposomes are stable in solution and under serum spiked condition. All in vitro experiments were conducted under normal cell culture conditions, containing 10 % v/v serum. Generally, in vitro experiments are conducted in serum-free media which does not even come close to approximating in vivo conditions. Moreover, preliminary results obtained with cellular uptake experiments of MS09:Chol lipoplexes in MCF-7 and HT-29 cells in 50 % v/v serum (physiological serum concentration) are encouraging. TEM image of the liposome is not clear and even in a small field of view the heterogeneity is not at all close to the state of the art1. They should provide TEM images of the large field, size distribution (with the sampling number), etc. The liposome after complexing with siRNA (lipoplex) looks like an aggregate structure rather than a single structure, this must be rectified. This has been corrected in the abstract. In the Abstract author specify the liposome as bilamellar but in the Results section that stated/appears to be unilamellar. Also, the size said in the abstract is less than 200 nm which far away from the result. The author should be very objective in the abstract itself. The liposomes were unilamellar. However, the complexes of liposomes and siRNA (lipoplexes) to which reference is made in the abstract were bilamellar The comparison of the liposome is done with Lipofectamine™ 3000, which is not the commercial benchmark for the siRNA transfection. They should provide efficacy compare to the Lipofectamine™ RNAiMAX in all the cases. Most of the figure captions are inadequate, they must be more informative. Additions to figure legends for Figure 1, 2, 4, 5, 6, 9 and 10 have been made. The subfigures are not labeled, like Figure 2, it is very difficult to understand which well is loaded with what. Subfigures in Figure 2, 4 and 9 have been labelled."
}
]
},
{
"id": "67957",
"date": "16 Dec 2020",
"name": "Basiru Olaitan Ajiboye",
"expertise": [
"Reviewer Expertise Metabolic Diseases",
"Drug discovery",
"and Phytomedicine."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript successfully described how anti-c-myc cholesterol can be used as an onco-nanotherapeutic (in vitro). This research work will serve as an eye opener for possible management of cancer. As we all known that oncology is a branch of medicine that deals with the prevention, diagnosis, and treatment of cancer, hence anti-c-myc cholesterol can be an example.\nTherefore, I want to use this medium to congratulate the authors and suggesting that the authors should carry out this experiment in vivo in future (if possible) so as to re-validate the in vitro results.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6240",
"date": "05 Jan 2021",
"name": "Moganavelli Singh",
"role": "Author Response",
"response": "We thank the reviewer for these kind and encouraging words"
}
]
}
] | 1
|
https://f1000research.com/articles/9-770
|
https://f1000research.com/articles/9-1176/v1
|
28 Sep 20
|
{
"type": "Correspondence",
"title": "Adverse events following measles-mumps-rubella-varicella vaccine: an independent perspective on Italian pharmacovigilance data",
"authors": [
"Paolo Bellavite",
"Alberto Donzelli",
"Alberto Donzelli"
],
"abstract": "Vaccine surveillance programs are crucial for the analysis of the vaccine’s safety profile and the guidance of health policies. The Epidemiological Observatory of the Italian Apulia Region carried out an active surveillance program of adverse effects following immunization (AEFI) after the first dose of the measles-mumps-rubella-varicella (MMRV) vaccine, finding 462 AEFIs per 1000 doses, with 11% rated serious. Applying the World Health Organization (WHO) causality assessment algorithm, 38 serious AEFIs/1000 enrolled were classified as ‘consistent causal associations’ with MMRV immunization. Severe hyperpyrexia, neurological symptoms and gastrointestinal diseases occurred in 38, 20 and 15 cases/1000 enrolled, respectively. A projection of such AEFIs in an Italian birth cohort would give tens of thousands of serious AEFIs. These incidence data are much greater than the incidence of serious AEFIs reported by the Italian Medicines Agency (AIFA) for years 2017 and 2018, mainly based on passive (or mixed) pharmacovigilance. In a previous epidemiological study in the same Italian Region, during an eight year passive surveillance, the reporting rate of serious AEFI was 0.06/1000 doses, and no cases of febrile seizures were detected applying the WHO algorithm. Taken together, the data suggest that passive pharmacovigilance is utterly inadequate to document the real incidence of serious AEFIs and that current methods of assessing causality may be questioned. Active surveillance programs are required in representative population samples, with results presented separately from those of spontaneous reporting, and causality assessment should be performed carefully and using a correct technique for AEFIs presenting as complex and multifactorial diseases, like those with serious neurologic disorders.",
"keywords": [
"Measles-mumps-rubella-varicella vaccine",
"MMRV",
"Pharmacovigilance",
"Active surveillance",
"WHO Guidelines for causality assessment",
"Vaccine safety",
"Adverse events following immunisation",
"AEFI."
],
"content": "Introduction\n\nAdverse events following immunization (AEFIs) are normally detected by various methods, starting from the preclinical and clinical studies required for product registration and then extending to passive (spontaneous) or active post-marketing surveillance on samples of the population. In fact, the different methods often give different results and a clear view of how many are such events, particularly the most serious ones, is still quite uncertain, especially for the more recently introduced vaccinations. Moreover, a causality assessment is conducted on the adverse events recorded, mostly using the algorithm proposed by World Health Organization (WHO), to discriminate whether an event is related to vaccination or independent. This procedure normally leads to the exclusion of vaccine responsibility in several cases of adverse reactions, attributing the cause to other factors. Recently such procedure, justified to exclude false attributions in the presence of alternative causes, has been criticized because of its uncertainty of application in the more complex and multifactorial diseases1,2.\n\nVaccines has published a paper by researchers from the University of Bari and the Apulia Region3 reporting and updating the main results of the 2018 report of the Apulia Region official Epidemiological Observatory on surveillance after the administration of measles-mumps-rubella-varicella (MMRV) vaccine4. Subsequently, Human Vaccines and Immunotherapeutics reported a retrospective study of AEFIs following MMRV vaccine during eight years of passive pharmacovigilance5. This vaccination has been made mandatory in Italy for all newborns from 2017 onwards. Since 2018, the MMRV vaccine was adopted by six states in Europe, two states in America, and by Australia6.\n\nDespite the results showing a high incidence of serious AEFIs, the authors concluded reassuringly on the safety profile of the MMRV vaccine5 and stated that “the active surveillance program confirmed and reinforced the safety profile of the vaccine\"3. A specific aspect concerns the incidence of febrile seizures after vaccination with MMRV. According to the research in Apulia would be 0.5 cases per 1000 follow-ups, much lower than what was reported so far by randomized studies and meta-analyses7–11. However, this calculation was obtained after discarding 3 out of 4 cases of febrile seizures by applying the WHO causality assessment algorithm. Furthermore, the data of this paper3 are in partial conflict with a retrospective epidemiological research carried out in the same Region with passive methods of AEFI reporting5.\n\nOur article challenges some conclusions of the cited studies3–5, by analyzing the data and comparing these to those reported by the Italian pharmacovigilance system and by the international literature on the MMRV vaccine. A search was made in PubMed using the key words \"MMRV and adverse\" (83 papers) or \"MMRV and safety\" (71 papers) and in the recent Cochrane review on the same topic12 and its bibliography. The data of the incidence of AEFIs reported by the Italian pharmacovigilance system were taken from the reports for years 2017 and 2018, by the Italian Medicines Agency (AIFA), a public body operating under the direction of the Ministry of Health and supervised by the Ministry of Economy. Every year AIFA publishes a report on the vaccine surveillance activities of the previous year. In the reports concerning the vaccinations made in 201713 and in 201814, the data of the MMR and MMRV vaccine are aggregated, even if most are related to the MMRV vaccine. Very few cases of AEFIs are reported with the monovalent varicella vaccine. The surveillance system adopted in Italy, and in most other countries, is mainly of the \"passive\" type, but the data of some \"active\" studies are also entered.\n\n\nIncidence of adverse events\n\nThe Apulia active vigilance study3 enrolled children in the second year of life, for whom there is an active and free vaccination offer of the first dose of MMRV vaccine, after acceptation by the parents, who received diaries for recording any AEFI occurring in the three post-vaccination weeks. The parents were interviewed by telephone 25 days after vaccination, asking them to detail the AEFIs noted in the diaries; data about hospitalizations were also collected. If the AEFIs had not yet resolved, a further follow-up was scheduled one month later. During the years 2017 and 2018, 2,540 children were enrolled, and post-vaccination follow-up was completed for 2,149 subjects (84.6%)3. Among them, 992 AEFIs were detected over the three weeks of monitoring, with a reporting rate of 462/1000 enrolled children.\n\nAmong the AEFIs reported, 883 (89.0%) were not serious, while 109 (11.0%) were rated as serious according to the WHO Guidelines15. Events are classified as serious when they result in death, are life-threatening, require in-patient hospitalization or prolongation of existing hospitalization, cause persistent or significant disability/incapacity, are a congenital anomaly/birth defect, or require intervention to prevent permanent impairment or damage. Moreover, in 2016, the AIFA published a list of specific health conditions which, if they occur after a vaccination, must be considered as serious AEFIs (for example, hypotonia-hyporesponsiveness, vasculitis and thrombocytopenia)16.\n\nFor serious AEFIs, the authors applied the WHO causality assessment algorithm15, as suggested by AIFA16, to classify AEFIs as \"consistent causal association\", \"inconsistent causal association\", \"indeterminate\", \"not-classifiable\". The causality assessment was carried out separately by two public health physicians with expertise in vaccinology, and a third physician was consulted in case of disagreement. A causal association consistent with MMRV immunization was classified in 82/109 serious AEFIs (most frequently fever/hyperpyrexia, followed by neurological symptoms and gastrointestinal diseases), with a rate of 38/1000. The authors report that most of the 82 serious events had resolved within 25 days of vaccination, while 10 children (12.2% of the serious AEFIs) had longer-lasting consequences, which had resolved within the following month.\n\nTable 1 displays the main numbers of the Apulia research relating to the serious AEFIs classified as consistent with vaccination. The fourth column shows an extrapolation to an annual birth cohort of 430,000 children of the Italian population. The calculation is indicative and approximate, because it is not certain that the Apulia research sample is representative of an Italian birth cohort. Furthermore, as the authors point out, a sample of just over 2000 subjects cannot reliably detect any rare (>1/10,000 but <1/1000) or very rare (<1/10,000) AEFI. The right column of Table 1 shows the incidence of some serious AEFIs in Italy during 2017, according to the AIFA report13.\n\n* clonus/febrile seizures were detected in 4/109 serious AEFIs, but only 1 was considered as associated to MMRV vaccination: 2 cases were \"not consistently” associated, because of the presence of a not described “alternative cause” of adverse events; 1 case was \"indeterminate\", because time from vaccination was compatible, but another cause - viral pharyngotonsillitis - was supposed during hospitalization3. N.r.=not reported by AIFA.\n\n** Ataxia/balance disorder.\n\n*** Thrombocytopenia.\n\nAIFA, in its 2018 Vaccine Report14, described 0.127 serious events out of 1000 doses of MMRV or MMR vaccine, a figure that is difficult to reconcile with that of the Apulian report3, which found 38 serious AEFIs out of 1000 enrolled – a number almost 300 folds greater than that in the AIFA Report.\n\nIn the AIFA Vaccines Report for year 201713, the occurrence of the different clinical reactions to the vaccine is also described (Table 1 right column). The serious AEFI more often correlated to vaccination was hyperpyrexia, with a reporting rate of 0.108 events per 1000 doses administered. Reporting rates for other MMRV or MMR vaccination-related serious AEFIs of major interest (rate per 1000 doses) are also described: generalized skin reactions 0.02; morbilliform/varicelliform rash: 0.013; convulsions: 0.005; thrombocytopenia: 0.007; ataxia/balance disorder: 0.002. These values are clearly very different (and much lower!) from those reported in the research of the Apulia Region.\n\nThe results of the Apulian studies are also surprising with regard to serious neurological symptoms and seizures. On the one side, an incidence of serious but unspecified neurological symptoms of 20/1000 is a new and unexpected finding, on the other a rather low incidence (0.5/1000) of febrile seizures is reported. In the eight year retrospective study, not a single case of febrile seizures was documented. In previous active surveillance programs the incidence of febrile seizures following the first dose of MMRV was 2.6 per 1000 (8/3019)7 or 1.7/100010, higher than the reported risk of 0/10005 or 0.5/10003. In a recent Cochrane review, the attributable risk to vaccine-induced febrile seizures is estimated around 1/1700 – 1/1150 administered doses12, but this value includes also MMR without varicella immunization. Stefanizzi et al. attribute their described low incidence both to the low reporting in the spontaneous pharmacovigilance study5, and to the fact that in three of four cases they discarded the responsibility of the vaccine by applying the WHO causality assessment algorithm.\n\n\nDiscussion\n\nThe data collection of the Epidemiological Observatory of the Apulia Region3–5 provided a great deal of thought-provoking results. Given the importance of the study, the only one in Italy with active surveillance combined with the WHO causality assessment, we believe it is important to discuss the authors’ interpretations of the data.\n\nThese issues are of general interest, because the different methods of detecting AEFI are under debate and because the adverse effects of MMRV vaccine are still under scrutiny7,8,12,17,18. Moreover, the question of the MMRV combination vaccine versus separate administration of MMR and varicella components is open to investigation10, also considering that the real need of vaccination for varicella can be discussed separately from other immunizations19.\n\nIn general, the reported epidemiological data contrast with some reassuring conclusions about the safety of the vaccine and with other statements in the cited papers3–5, about the methods of pharmacovigilance of vaccines. We aim to highlight some inconsistencies in the interpretation of the observed data and to present another point of view, concluding with some proposals about resuming active vigilance programs.\n\n\nEmerging signals\n\nThe abstract of the paper of active surveillance3 ends with an encouraging statement: \"Because no emerging signals were detected, our data from the active surveillance program confirmed the safety profile of the MMRV vaccine.\" Furthermore, in the Introduction, the authors state that following vaccinations \"serious AEFIs are absolutely rare\", precisely citing an Expert Opinion on Drug Safety20 and the aforementioned WHO document15. The Discussion reiterates that \"the active surveillance program confirmed and reinforced the safety profile of the vaccine\". The data in the article3 however, are different: many readers may understand \"absolutely rare\" as \"very rare\", but internationally very rare events are defined as those with a frequency <1/10,000, while in the report the causally related serious AEFIs are 38/1000. This frequency should classify them as “common” AEFIs (<1/10 but > 1/100)21.\n\nThe results of the Apulia report should be compared with what is already known by the reports of the national health authorities. The data of the active surveillance show a number of serious AEFIs related to the MMRV vaccine hundreds of times higher than expected, based on spontaneous surveillance and AIFA reports. It is surprising that 38/1000 serious AEFIs, instead of 0.127/1000 declared in the AIFA Report for the same vaccine, are not considered an \"emerging signal\". Although a high incidence of febrile reactions and skin rash in the first 10 days after MMRV vaccination was expected and here confirmed (but in 38 x 1000 cases they were classified as ‘serious’), serious neurological symptoms and serious gastrointestinal diseases are not described with such frequency in the literature or by current surveillance systems.\n\nThis marked discrepancy is almost certainly due to the differences between active reporting and the passive (or mixed) reporting adopted by the Italian health authorities. It is implausible that such a large discrepancy could be due to local factors, such as the use of different vaccines or different sensitivity to adverse events of the population. It cannot be excluded that an incidence of AEFIs as high as that reported3 may be, in part, due to the concomitant injection of a Hepatitis A vaccine, but data to explore this possibility are not provided. Whatever the reasons, the report of the Apulia Region offers an unexpected and worrying picture of the incidence of serious AEFIs and cannot be taken as a confirmation of what already known.\n\nStefanizzi et al.3 maintain that these numbers are consistent with a similar active surveillance report with telephone interviews by Huang et al.22, in which AEFIs were 480/1000 at follow-up. However, this paper concerned an adjuvanted vaccine against H5N1 influenza and adult participants. A previous randomized controlled trial of active surveillance for the MMRV vaccine (ProQuad)11 reported an AEFI rate of 65.6%, of which 37.7% were judged to be related. However, only 0.7% of the described AEFIs were classified as serious and, among them, 0.3% were judged to be related.\n\nThe data are not reassuring, and require reconsideration of the validity of the AIFA annual reports, essentially based on passive surveillance.\n\n\nUnder-reporting\n\nIn the Introduction of their paper, Stefanizzi et al.3 state that \"passive post-marketing surveillance is affected by under-reporting, especially for non-serious AEFIs\". Their Discussion asserts that “The proportion of non-serious adverse events resulted higher than the Italian estimate that indicated for 2017 in the last AIFA report (88.7% vs. 80.0% in the AIFA report) and higher than Apulia data from passive surveillance in the 2013–2017 period (88.7% vs. 75.4%): this findings seems to indicate that the under-report of passive AEFIs surveillance mainly regarded non-serious adverse events”3 – here the authors mention the 2018 report of the Observatory of the Apulia Region4. However, the data show a different picture. While in the Apulia study3 the “proportion” of non-serious AEFIs detected with passive surveillance is close to the 2017 AIFA Report (albeit a little higher: 88.7% vs 80%), the absolute amount of non-serious AEFIs x 1000 doses, which in the Apulia active surveillance report are 883 in 2,149 children, is exceedingly higher than that reported in the AIFA Report for Italy in 201713. Again, in the same report from the Observatory of the Apulia Region previously published on the institutional website (Table 3.4.3.1.)4, the reporting rate of serious AEFIs with active surveillance was 40.69 per 1000 doses, with passive surveillance 0.12 per 1000 doses. The difference between active and passive surveillance was 339 times. If we consider serious AEFIs with consistent causal association with the vaccine, the ratio published on the complete Apulia report4 (paragraph 3.4.3.2.) is even more unbalanced; reporting rate 29.3/1000 doses with active surveillance, 0.03/1000 doses with passive surveillance – a 977 times difference.\n\nIt should also be taken into account that the current schedules contemplate the repetition of the MMRV vaccine at least twice in life (but perhaps more than that, as a consequence of the shorter duration of vaccination immunity compared to that resulting from the corresponding diseases, and of a lower circulation of wild viruses, with the decline of vaccine protection in the absence of natural boosts)19,23.\n\nGiven that the large discrepancies we have described in Table 1 between the AIFA reports and the active surveillance data also concern serious AEFIs, the authors' opinion that \"the under-report of passive AEFIs surveillance mainly regarded non-serious adverse events\"3 is not evidence-based.\n\nIn the Discussion it is repeated that febrile seizures are “the most common adverse event following the MMRV vaccine”3. However, in the data of the Apulia Region, the other causally related serious neurological symptoms were 43 times more common than seizures (Table 1). A better definition of these serious neurological symptoms would have been appropriate.\n\n\nIncidence of clonus/febrile seizures and causality assessment\n\nHyperthermia following MMRV vaccination can be accompanied by febrile convulsions and this serious event is a major concern for the population. In previous active surveillance programs, the incidence of febrile seizures post dose 1 of MMRV was 2.6 per 1000 (8/3019)7 or 1.7/100010, or “high but under 2.95/1000”8, values higher than the reported risk of 0/10005 or 0.5/10003. In the latter paper (by Stefanizzi et al.), clonus/febrile seizures were detected in 4/109 serious AEFIs (reporting rate = 2/1000 follow-up), but only one episode was considered as consistently associated with vaccination, so that the reporting rate dropped to 0.5/1000. Following this kind of evidence, the authors suggested that “the frequency of seizure consistently associated with the MMRV vaccine is lower than those published in the previous studies”. However, the strength of this conclusion is based on a solitary episode after the exclusion of 3/4 cases of seizures, though a causality assessment of these cases was not clearly described.\n\nA role of vaccination in 2/4 clonus/febrile seizure episodes was excluded because of the presence of “alternative cause of adverse events”. 1/4 cases was excluded as indeterminate because “another cause of hyperpyrexia and febrile seizure was supposed during hospitalization (viral pharyngotonsillitis)”. In the Stefanizzi et al. report3, the supposed “alternative causes” in 2 of 4 cases are not even mentioned, thus hindering any independent evaluation of the circumstance.\n\nWe should also consider that causality assessment was carried out by two public health physicians, experts in vaccinology, not by neurologists or pediatricians or pathologists, while to ensure compliance with the rigorous criteria for causality assessment and wider acceptance of the results, the WHO manual15 recommends that the procedure is performed by a multidisciplinary committee comprising experts from pediatrics, neurology, general medicine, forensic medicine, pathology, microbiology, immunology and epidemiology.\n\nThe text deals with “clonus/febrile seizures”, without clearly distinguishing between the two conditions. However, it is known that the most frequent convulsions not related to hyperthermia are of the tonic-clonic type and not only clonic. Moreover, assuming that the \"alternative causes\" of the seizures were epilepsy, it would be important to consider the relationship between epilepsy and specific vaccines, since vaccination might precipitate adverse events in children with familial tendency to seizures or genetic epilepsy syndrome2,24–26. It would not be conceptually and scientifically correct to use the WHO algorithm in such a way as to consider an \"alternative cause\" a genetic tendency to epilepsy, because the vaccine in this case could be the precipitating or triggering factor of the individual's susceptibility. Recognition of this possible occurrence is important, because there is a need to select vaccines that carry lower risk of febrile seizures in these children who are particularly prone to develop this adverse event27.\n\nEspecially for infectious and inflammatory illnesses, it has been noted that the supposed “other cause”, mentioned by WHO algorithm, should be independent from a possible interaction with the perturbation induced by vaccination1,2. In fact, vaccination can act as a synergistic or triggering factor in a person affected by genetic susceptibility or by a concomitant infectious disease. None of these issues is addressed in the cited paper3. Concerning the case of seizures in a child with “viral pharyngotonsillitis”, this diagnosis does not exclude the contribution of vaccination in the development of hyperpyrexia and seizures. A concomitant infectious or inflammatory disease occurring in the time window of the immune reaction to vaccine cannot make the role of vaccine “indeterminate” and so a causality link cannot be excluded2.\n\nBecause of the extremely small sample of the study and of the outlined doubts about the exclusion of the three cases, it seems highly questionable the conclusion3 that the frequency of seizure consistently associated with the MMRV vaccine “is lower than those published in the previous studies that considered all seizure temporary associated with the vaccination, without a standardized causal evaluation”.\n\nTo illustrate how the application of the WHO algorithm is difficult and potentially error-prone, four case studies are presented (Box 1), in which the causality association of a serious neurological AEFI with the vaccine was excluded5. These examples raise some problems and deserve clarification, without which a high risk of misinterpretation exists. The notes of the authors concern: a) whether the alternative \"other cause\" was sufficiently clear and \"strong\" as the only possible cause of the event and b) whether there could be a plausible interaction between other clinical conditions and the biological action of the vaccine.\n\n\n\nCase 1\n\nCase n. 9 cited in the paper of Stefanizzi et al.5: \"The ninth case involved a 12-months-old female. A week after the vaccination, she presented a sudden loss of consciousness with staring eyes, hypertonic for about 10 min, modest hypersalivation. She was hospitalized and, after medical examination, she was discharged with the diagnosis of hyporesponsive episode in patient with vomiting and metabolic acidosis. Applying the Causality Assessment algorithm, cause/effect relationship between vaccination and adverse events is inconsistent, because an alternative cause (gastrointestinal infectious disease) has been recognized.\"\n\nNote: In this case, the adverse effects following immunization (AEFI) took place precisely in the time window in which the greatest number of episodes of febrile seizures normally occur, so there is a high biological plausibility and a correct time window for attributing causality to the vaccine. In the report by Stefanizzi et al3 there is a very high incidence of serious gastrointestinal symptoms with a causality ascertained with the vaccine. It is not possible to understand how vomiting and metabolic acidosis can justify the diagnosis of \"gastrointestinal infectious disease\" as an “alternative cause”, also without a microbiological analysis. Notably, according to the first step of World Health Organization (WHO) algorithm of causality assessment15, when the AEFI occurs in the expected time window and there is biological plausibility, a supposed “other cause” must be “strong” enough to exclude the role of the vaccine in the causality. This criterion does not appear to support the attribution of neurological symptoms to a supposed “gastrointestinal infectious disease” rather than to a vaccine adverse reaction. Furthermore, even if it were really a gastroenteritis, it cannot be excluded that the neurological symptoms were due to the perturbation of the gut-brain axis28, that is, in our case, to the interaction between the induced inflammatory stress from the vaccine and gastrointestinal disorder with alteration of the mucosa, release of endotoxins or other metabolites in the circulation2.\n\nCase 2\n\nCase cited in both reports3,5: \"The 13th case regarded a 15-months-old male. Nine days after vaccination, he reported hyperpyrexia and febrile seizure associated with eyes rolling, limbs twitchings, and loss of consciousness. This episode ended after a few minutes: for these symptoms, he was admitted to the hospital and discharged after 3 days for the complete AEFI resolution. During hospitalization he presented fever but he did not report another episode of febrile seizures. After medical examination, a final diagnosis of febrile seizure caused by viral pharyngotonsillitis was formulated. Applying the Causality assessment algorithm, the cause/effect relationship between vaccination and adverse events is inconsistent for the presence of an alternative disease (viral pharyngotonsillitis).”\n\nNote: In this case the febrile convulsions occurred in the most probable time window and there is also a considerable literature on the fact that the vaccine can cause this phenomenon. The concomitant presence of pharyngotonsillitis cannot be considered an alternative cause strong enough to rule out the role played by the hyperpyrexia caused by the vaccine. In this case, a trivial viral infection could well have occurred in a child whose immune system was very stressed by vaccination with four live viruses and the strong fever due to the two different causes may have triggered the seizures. It is notoriously recommended not to vaccinate a person if he has another infectious disease in progress, but if the vaccination takes place during the period of incubation of the infection, a pathological synergy between the two stimuli may occur. Another possibility that cannot be ruled out, at least in principle, is that the pharyngotonsillitis was caused directly by one of the injected live vaccine viruses. It is known that the measles vaccine virus infects lymphatic tissue29and vaccine-related upper respiratory infections are reported in 12/1000 of children vaccinated with MMRV (ProQuad)30. Incidentally, the causal assessment decision for the same case (viral pharyngotonsillitis and post-MMRV seizures) was judged as \"indeterminate\" in one case3 and \"inconsistent\" in a subsequent publication5, but the two classifications are very different according to the same WHO manual.\n\nCase 3\n\nCase n. 19 cited in the paper of Stefanizzi et al.5: “The case involved a 15-months-old female vaccinated with MMRV and anti-HAV vaccines. Ten days after immunization, she developed fever and hyperpyrexia and strabismus, which was classified as serious and permanent invalidity. Applying the Causality Assessment algorithm, the cause/effect relationship between vaccination and adverse events is not consistent, because of the absence of biological plausibility between strabismus and vaccine administration.”\n\nNote: in this case, it does not seem correct to exclude a causal relationship with vaccination by appealing to the lack of biological plausibility. In fact, strabismus may be caused by oculomotor nerve palsy31 and several cases of third cranial nerve palsy after vaccination (with both live and inactivated vaccines) have been described and reported in the US Vaccine Adverse Event Reporting System (VAERS)32. Although it is not possible to determine causal associations based on VAERS reports, the authors of the review do not deny it either. More importantly, they do not question the plausibility of such an adverse reaction, because cranial nerve palsy may sometimes be the harbinger of encephalomyelitis, which may, although rarely, be caused by vaccinations. Cases of oculomotor nerve palsy occurring after MMR vaccination has already been described in the scientific literature33,34.\n\nCase 4\n\nCase n. 23 cited in the paper of Stefanizzi et al.5: “The case involves a male child aged 30 months: few hours after vaccination, he developed hyperpyrexia with an episode of febrile seizure. He was hospitalized and symptoms persisted for 9 days. Applying the Causality Assessment algorithm, the cause/effect relationship between vaccination and adverse events is classifiable as inconsistent: even the biological plausibility of AEFI, the time window between vaccination and adverse reactions (hyperpyrexia and febrile seizure) is not compatible (too short).”\n\nNote: Although hyperpyrexia caused by MMRV vaccine usually peaks after one week from the first dose in about 10% of subjects, in some subjects it occurs between the first and the 5th day after inoculation. In a randomized study with active surveillance11 it was observed that the rate of fever (temperature > 39.0° C) in the first 5 days after first dose of MMRV was 8 cases every 1000 doses. This data makes it improper to exclude causation in a case of febrile seizures by applying only a weak criterion such as a time window that excludes the first day after the vaccine injection.\n\n\nMMRV or MMR+V?\n\nA related point, not discussed in the cited paper, concerns the formulation of the vaccine used. An increased risk of onset of fever and seizures has been documented after administration of the first dose of tetravalent vaccine (MMRV) compared to separate administration of MMR vaccines and varicella vaccine35–39. Analysis on the Italian national database of AIFA confirmed a more than two-fold risk of febrile seizures after administration of MPRV vaccine compared to vaccination with separate vaccines. This finding particularly applies to younger children and it is mostly observed in the first 5-12 days after the administration of the vaccine. Therefore, the AIFA Paediatric Working Group, similarly to what is suggested in other countries, recommended pediatricians and other health professionals not to use the tetravalent MPRV vaccine as a first dose for immunization against measles, mumps and rubella40. Furthermore, most of the children who participated in the active surveillance research3 were inoculated with MMRV vaccine plus hepatitis A vaccine, but a potential extra effect of simultaneous vaccination in the same session cannot be inferred from the aggregated results. These important issues, not considered in the study, make even more doubtful the interpretation of the relatively low rate of febrile seizures in the sample from Apulia.\n\n\nPerspectives\n\nThe Apulia epidemiological studies3–5 have the merit of clearly raising the issue of methods for detecting adverse reactions, a crucial aspect of vaccinology. We agree with the conclusion of the article3 which states that \"active surveillance programs periodically have to be implemented in order to improve the overall performance of the pharmacovigilance system and validate the data and emerging signals detected by spontaneous reporting activity\"; and that \"active surveillance programs can be considered an effective solution to a real question: the public concerns about risks associated with immunization.\" However, in explaining this legitimate objective, the authors do not notice that the reiterated reassurance about safety problems, in line with the dominant paradigm, is not consistent with the results of their active surveillance. These results are objectively not so reassuring due to the serious AEFIs carefully documented at the individual level and their possible impact projected at the level of the national community.\n\nThe debate on the best methods of surveillance in the field of vaccinology should remain open, in the interest of the entire population1,2,41–43. Besides the WHO algorithm15 other criteria developed by various groups working within the greater field of pharmacovigilance are used for causality assessment, such as the World Health Organization Collaborating Centre for International Drug Monitoring, the Uppsala Monitoring Centre (WHO–UMC)44 and the Naranjo algorithm45,46. Unlike the WHO algorithm, which judges as inconsistent the association of an AEFI with vaccination if there is “another cause”, the other methods use a score or a probability assessment, taking into account the clinical-pharmacological aspects of the case history and the quality of the clinical documentation. This approach seems more suitable for judging complex cases, in which multiple interacting causes (genetic, infectious, toxic) determine the pathology that occurs after vaccination.\n\nThe cited research highlights the inadequacy of passive surveillance to represent the real incidence of even short-term AEFIs, both of mild and serious kind. The distance between the incidences of AEFIs detected3 and the passive reports collected in the same Region5 and even in the Italian regions with the most efficient reporting system13,14 calls into question the extent of investments to improve spontaneous reports. These reports must be maintained, to allow the reporting of rare events that active surveillance would have little chance to intercept. Moreover, investments should be redirected to active surveillance studies (indeed already committed by the Italian Ministry of Health), designed on representative samples of the Italian population, the only ones from which valid and credible inferences can be drawn. In any event, the public reports should present the active and passive surveillance data in a disaggregated way.\n\nThe publication of the Apulia active surveillance experience has also the merit of having applied the WHO algorithm for causality assessment15, but this important step of analysis was applied in a debatable way. In light of these important questions, we believe it is appropriate to open a scientific debate, also taking into account critical and constructive scientific positions, based on the available set of evidence about effectiveness and safety of current immunization schedules, on a continuous reassessment of their validity, and on the frank admission of the persistent areas of uncertainty. We deem also legitimate an open scientific debate about possible different implementation strategies and priority assessments among the current immunization strategies.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.",
"appendix": "Acknowledgments\n\nWe thank Dr. Donatella Sghedoni for her English revision of the manuscript.\n\n\nReferences\n\nPuliyel J, Naik P: Revised World Health Organization (WHO)'s causality assessment of adverse events following immunization-a critique. F1000Res. 2018; 7: 243. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBellavite P: Causality assessment of adverse events following immunization: the problem of multifactorial pathology. F1000Res. 2020; 9: 170. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStefanizzi P, Stella P, Ancona D, et al.: Adverse Events Following Measles-Mumps-Rubella-Varicella Vaccination and the Case of Seizures: A Post Marketing Active Surveillance in Puglia Italian Region, 2017-2018. Vaccines (Basel). 2019; 7(4): 140. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTafuri, S, Ancona D, Chielli A, et al.: Sorveglianza degli eventi avversi a vaccino in Puglia 2013-2017. Puglia-OER. 2018; 20: 1–38. Reference Source\n\nStefanizzi P, De NS, Patano F, et al.: Post-marketing surveillance of adverse events following measles, mumps, rubella and varicella (MMRV) vaccine: retrospecive study in apulia region (ITALY), 2009-2017. Hum Vaccin Immunother. 2020; 16(8): 1875–1883. PubMed Abstract | Publisher Full Text\n\nWHO: Vaccine-preventable diseases: monitoring system. 2020 global summary. WHO. 2020 [accessed on 2020 August 25]. Reference Source\n\nKlopfer SO, Stek JE, Petrecz M, et al.: Analysis of safety data in children after receiving two doses of ProQuad® (MMRV). Vaccine. 2014; 32(52): 7154–7160. PubMed Abstract | Publisher Full Text\n\nMa SJ, Xiong YQ, Jiang LN, et al.: Risk of febrile seizure after measles-mumps-rubella-varicella vaccine: A systematic review and meta-analysis. Vaccine. 2015; 33(31): 3636–3649. PubMed Abstract | Publisher Full Text\n\nLeung JH, Hirai HW, Tsoi KK: Immunogenicity and reactogenicity of tetravalent vaccine for measles, mumps, rubella and varicella (MMRV) in healthy children: a meta-analysis of randomized controlled trials. Expert Rev Vaccines. 2015; 14(8): 1149–1157. PubMed Abstract | Publisher Full Text\n\nCocchio S, Zanoni G, Opri R, et al.: A postmarket safety comparison of 2 vaccination strategies for measles, mumps, rubella and varicella in Italy. Hum Vaccin Immunother. 2016; 12(3): 651–654. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarshall GS, Senders SD, Shepard J, et al.: A double blind, randomized, active controlled study to assess the safety, tolerability and immunogenicity of measles, mumps rubella, and varicella vaccine (MMRV) manufactured using an alternative process. Hum Vaccin Immunother. 2016; 12(8): 2188–2196. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDi Pietrantonj C, Rivetti A, Marchione P, et al.: Vaccines for measles, mumps, rubella, and varicella in children. Cochrane Database Syst Rev. 2020; 4(4): CD004407. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarchione P, Felicetti P, Petronzelli F, et al.: Rapporto vaccini 2017. Sorveglianza post-marketing in Italia. Roma, AIFA, 2018. [accessed on August 25, 2020]. Reference Source\n\nMarchione P, Felicetti P, Petronzelli F, et al.: Rapporto vaccini 2018. La sorveglianza post-marketing in Italia. Roma, AIFA, 2019. [accessed on August 25, 2020]. Reference Source\n\nWHO: Causality assessment of an adverse event following immunization (AEFI): user manual for the revised WHO classification (Second edition). 2nd Edition, Geneva, World Health Organization, 2018. Reference Source\n\nAIFA-Gruppo di lavoro sull'analisi dei segnali dei vaccini: Guida alla valutazione delle reazioni avverse osservabili dopo vaccinazione in AIFA, (ed): Roma, 2016. [Accessed on August 26, 2020]. Reference Source\n\nGvozdenovic E, Vetter V, Willame C, et al.: Impact of history of febrile convulsions on the risk difference of febrile convulsions with the tetravalent measles-mumps-rubella-varicella vaccine: Post-hoc exploratory analysis of results from a matched-cohort study. Vaccine. 2018; 36(39): 5803–5806. PubMed Abstract | Publisher Full Text\n\nCashman P, Moberley S, Durrheim D: MMRV vaccine safety. Vaccine. 2019; 37(30): 3946. PubMed Abstract | Publisher Full Text\n\nDonzelli A, Demicheli V: [Varicella vaccination: scientific reasons for a different strategic approach]. Epidemiol Prev. 2018; 42(1): 65–70. PubMed Abstract | Publisher Full Text\n\nPrincipi N, Esposito S: Adverse events following immunization: real causality and myths. Expert Opin Drug Saf. 2016; 15(6): 825–835. PubMed Abstract | Publisher Full Text\n\nCIOMS (Council for International Organization of Medical Sciences) Working Group III Report: Guidelines for Preparing Core Clinical Safety Information on Drugs (Chapter 5, Good Safety Information Practices). Geneva, WHO, 1995. Reference Source\n\nHuang WT, Chang CH, Peng MC: Telephone monitoring of adverse events during an MF59-adjuvanted H5N1 influenza vaccination campaign in Taiwan. Hum Vaccin Immunother. 2014; 10(1): 100–103. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDonzelli A, Bellavite P: [Immunocompromised children and non-vaccinated classmates: how massive is this problem?] Epidemiol Prev. 2019; 43(2–3): 194–198. PubMed Abstract | Publisher Full Text\n\nVerbeek NE, Van Der Maas NA, Jansen FE, et al.: Prevalence of SCN1A-related dravet syndrome among children reported with seizures following vaccination: a population-based ten-year cohort study. PLoS One. 2013; 8(6): e65758. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVerbeek NE, Jansen FE, Vermeer-de Bondt PE, et al.: Etiologies for seizures around the time of vaccination. Pediatrics. 2014; 134(4): 658–666. PubMed Abstract | Publisher Full Text\n\nFeenstra B, Pasternak B, Geller F, et al.: Common variants associated with general and MMR vaccine-related febrile seizures. Nat Genet. 2014; 46(12): 1274–1282. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi X, Lin Y, Yao G, et al.: The Influence of Vaccine on Febrile Seizure. Curr Neuropharmacol. 2018; 16(1): 59–65. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDinan TG, Cryan JF: The Microbiome-Gut-Brain Axis in Health and Disease. Gastroenterol Clin North Am. 2017; 46(1): 77–89. PubMed Abstract | Publisher Full Text\n\nCondack C, Grivel JC, Devaux P, et al.: Measles virus vaccine attenuation: suboptimal infection of lymphatic tissue and tropism alteration. J Infect Dis. 2007; 196(4): 541–549. PubMed Abstract | Publisher Full Text\n\nMerck: MMRV information sheet (revised 03/2020); Merck, 2020. [Accessed on August 26, 2020]. Reference Source\n\nQuoc EB, Milleret C: Origins of strabismus and loss of binocular vision. Front Integr Neurosci. 2014; 8: 71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWoo EJ, Winiecki SK, Ou AC: Motor palsies of cranial nerves (excluding VII) after vaccination: reports to the US Vaccine Adverse Event Reporting System. Hum Vaccin Immunother. 2014; 10(2): 301–305. PubMed Abstract | Publisher Full Text | Free Full Text\n\nManzotti F, Menozzi C, Porta MR, et al.: Partial third nerve palsy after Measles Mumps Rubella vaccination. Ital J Pediatr. 2010; 36: 59. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCheng DR, Crawford NW, Hayman M, et al.: Recurrent 6th nerve palsy in a child following different live attenuated vaccines: case report. BMC Infect Dis. 2012; 12: 105. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKlein NP, Fireman B, Yih WK, et al.: Measles-mumps-rubella-varicella combination vaccine and the risk of febrile seizures. Pediatrics. 2010; 126(1): e1–e8. PubMed Abstract | Publisher Full Text\n\nSchink T, Holstiege J, Kowalzik F, et al.: Risk of febrile convulsions after MMRV vaccination in comparison to MMR or MMR+V vaccination. Vaccine. 2014; 32(6): 645–650. PubMed Abstract | Publisher Full Text\n\nMacDonald SE, Dover DC, Simmonds KA, et al.: Risk of febrile seizures after first dose of measles-mumps-rubella-varicella vaccine: a population-based cohort study. CMAJ. 2014; 186(11): 824–829. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMa SJ, Li X, Xiong YQ, et al.: Combination Measles-Mumps-Rubella-Varicella Vaccine in Healthy Children: A Systematic Review and Meta-analysis of Immunogenicity and Safety. Medicine (Baltimore). 2015; 94(44): e1721. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBauchau V, Van HL, Cohen C: Modelling Hospitalisation Ratios for Febrile Convulsions and Severe Varicella Under Combined Measles, Mumps, Rubella, and Varicella (MMRV-Priorix-Tetra™) Compared to Separate MMR + V Vaccination. Drug Saf. 2015; 38(11): 1095–1102. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAIFA pediatric Working Group: Raccomandazioni del Working Group Pediatrico dell'AIFA su utilizzo dei vaccini MPRV. Roma, AIFA. 2011. Reference Source\n\nZaki SA: Adverse drug reaction and causality assessment scales. Lung India. 2011; 28(2): 152–153. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTozzi AE, Asturias EJ, Balakrishnan MR, et al.: Assessment of causality of individual adverse events following immunization (AEFI): a WHO tool for global use. Vaccine. 2013; 31(44): 5041–5046. PubMed Abstract | Publisher Full Text\n\nChandler RE: Modernising vaccine surveillance systems to improve detection of rare or poorly defined adverse events. BMJ. 2019; 365: l2268. PubMed Abstract | Publisher Full Text\n\nUppsala Monitoring Centre: The use of the WHO-UMC system for standardised case causality assessment. 2020; [Accessed on August 30, 2020]. Reference Source\n\nNaranjo CA, Busto U, Sellers EM, et al.: A method for estimating the probability of adverse drug reactions. Clin Pharmacol Ther. 1981; 30(2): 239–245. PubMed Abstract | Publisher Full Text\n\nBelhekar MN, Taur SR, Munshi RP: A study of agreement between the Naranjo algorithm and WHO-UMC criteria for causality assessment of adverse drug reactions. Indian J Pharmacol. 2014; 46(1): 117–120. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "75394",
"date": "26 Nov 2020",
"name": "Arthur Edward Brawer",
"expertise": [
"Reviewer Expertise I am a board certified rheumatologist who has published many peer-reviewed manuscripts encompassing vaccination-induced disorders. I am qualified to testify in vaccine court in Washington",
"D.C.",
"where special masters adjudicate claims of vaccine related ailments."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nVaccination-induced disorders are a genuine reality that continue to generate intense controversy. Although the majority of immunization recipients have little or no safety issues, that does not detract from the occurrences of multiple chronic systemic diseases initiated by a wide variety of parenteral vaccine exposures. Over the past four decades case reports of chronic vaccination-induced disorders published in peer-reviewed journals have generally segregated into two main categories: (a) autoimmune and autoinflammatory diseases; and (b) neuro-psychiatric diseases, with or without seizures and/or dyskinesias, and with or without overlapping clinical features resembling the various neurologic fatiguing syndromes (e.g. chronic fatigue syndrome, dysautonomia, small fiber neuropathy, fibromyalgia, and postural orthostatic tachycardia syndrome). Recently some novel ideas have been proposed regarding definitive identification of those at risk for any of the above phenomena in either category.\nPassive surveillance programs to monitor vaccine safety and effectiveness rely on reporting by patients, family members, manufacturers, and health care providers to capture either a serious or minor adverse event that has temporally occurred after immunization. There are many inherent problems with such programs, not the least of which are: (1) you can’t analyze what hasn’t been reported; (2) you can’t analyze what has been frivolously labelled as unrelated; and (3) you can’t analyze something serious that has been labelled trivial. Active surveillance programs utilize health care encounters and electronic medical records to capture data and are generally better in detecting safety signals in real time. However, as pointed out by the authors, such statistical signals can be spurious for many reasons. In addition, mechanisms of disease causation that initiated an acute severe vaccination-related event can subsequently be accompanied weeks later by other latent mechanisms that evolved more slowly before becoming clinically relevant. These secondary amplification loops can circuitously augment and perpetuate an acute severe adverse event and turn it into a chronic disabling process. Months later another physician may erroneously interpret the two processes to be separate events, whereby valid detection and assignment of the initial adverse immunization reaction are nullified. A specific example of this process is the initiation of a serious adverse event by hidden toxic chemicals that routinely accompany the beneficial ingredients in dozens of different parenteral vaccines. Such chemicals are also capable of inducing the latent development of multiple autoantibodies, the latter of which then transform the initial acute event into a chronic disease.\n\nChildren whose parents refuse or delay vaccinations are often dismissed from ongoing care by pediatricians. Likewise, vaccine refusal by adult patients can trigger the same fate from their internist. Authors Bellavite and Donzelli present valid arguments favoring reassessment of the safety data and effectiveness of current immunization schedules. These reassessments can serve as a template for clinicians of all specialties to reexamine this issue and not merely dismiss parental worries as unfounded hysteria. I favor this manuscript being indexed, as it will also likely induce academicians to reassess their denials of vaccination-induced diseases and refocus their attention towards defining the population at risk for serious adverse events.\n\nIs the rationale for commenting on the previous publication clearly described? Yes\n\nAre any opinions stated well-argued, clear and cogent? Yes\n\nAre arguments sufficiently supported by evidence from the published literature or by new data and results? Yes\n\nIs the conclusion balanced and justified on the basis of the presented arguments? Yes",
"responses": [
{
"c_id": "6230",
"date": "06 Jan 2021",
"name": "Paolo Bellavite",
"role": "Author Response",
"response": "We thank the Reviewer for these notes, confirming and deepening some points of our work. In particular, we agree on the emphasis on informed consent, sometimes underestimated by pediatricians, especially because of the laws that oblige parents to have fully vaccinated children undergo all vccinations, under penalty of exclusion from nursery school or sanctions. Therefore, we have included in the discussion a paragraph about this issue, as follows: “In this context, it seems useful to extend our attention to the choice among different vaccines and even to the choice of whether to vaccinate or not, a choice that must be made by the patient or by parents of underage children, properly advised by their doctors. Informed consent and/or the refusal of therapy by patients (or by parents in case of pediatric vaccinations) are debated issues in the medical profession and their importance is underlined by various international statements, such as the Convention on Human Rights and Biomedicine (ETS No 164) signed the 4 April 1997 in Oviedo (Spain). The introduction in the legislation of many countries of an obligation to be vaccinated has significantly altered the patient-doctor and parents-doctor relationships. Regardless of the different vaccines being useful or not, the rare possible single exemptions provided by the law, if there is a legal obligation the doctor’s role risks to switch from an \"evidence-based\" counsellor to a public officer enforcing the government’s decisions. Such issues were further exacerbated by the introduction of multiple component vaccines, such as the tetravalent with live attenuated viruses or the hexavalent with antigens fixed on aluminum nanoparticles. In these cases the doctor was also denied the possibility of recommending the most appropriate choice based on the epidemiological conditions and on the characteristics of the individual patient's susceptibility. The Italian law issued in July 2017 (law 119/2017) mandated that the vaccination obligation for MMRV should be revised after three years, but so far nothing has been done and nothing is expected to be done in the near future. Although these issues are still open, it is agreed that patients consulting their doctors for vaccination advice should be offered correct information of the highest qualitative and quantitative standard. This is why the principle of informed consent does not seem satisfied by a generic statement of a \"safe profile\" or \"lack of worrying signals\" of a vaccine, that indeed has shown serious adverse effects in a significant percentage of subjects. We hope our contribution may be useful to help the medical professionals and especially the pediatricians to evaluate the information to be shared with the public by utilizing authentic scientific evidence.”"
}
]
},
{
"id": "75611",
"date": "03 Dec 2020",
"name": "James Lyons-Weiler",
"expertise": [
"Reviewer Expertise I have conducted research on biomarkers of adverse events from medical treatments in cancer",
"published studies that have determined pediatric dose limits and excess of aluminum in vaccines in the CDC's schedule",
"catch-up schedule",
"and an alternative schedule",
"and have recently conducted a vaccinated vs. unvaccinated study in a pediatric population looking at numerous health outcomes. My training is in biology",
"computational molecular biology",
"and my professional work experience has been academic research in biomedicine focused on differential diagnosis and data-aided prognosis of health outcomes since 2000."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nEditorial Note from F1000Research – 28th January 2021:\nThis report has been updated to include a conflict of interest statement after concerns were raised by a reader about the impartiality of the reviewer. No changes have been made to the text of the report.\nThis article finds evidence of adverse events associated with the MMRV Vaccine in a post-market surveillance study.\nThe authors have applied the WHO criteria for causality and critique a previous report for underreporting, subjective exclusion of Day 1 events, and provides four examples of arbitrary or subjective exclusion of specific cases that under more objective criteria meet the criteria of evidence in support of causality.\nThe authors have found increased incidence of severe hyperpyrexia, neurological and gastrointestinal diseases compared to the prior report, but provide no reference to national or regional baseline rates of these conditions in age-matched children that could not have received the MMRV vaccine. Such data may not exist, and that would be failing of public health data reporting, not the authors. However, if such data do exist, it would serve well to include references to rates of such conditions in age-match (or close-to-age matched) unvaccinated children.\nThere are some new studies that are worth citing that might be considered. To avoid any COI I do not list my own. I recommend for example Hooker and Miller1 and Peter Aaby's group's recent studies (especially the DTaP study in a vaccine-naïve population)2.\nThe authors are correct in their views that \"This approach seems more suitable for judging complex cases, in which multiple interacting causes (genetic, infectious, toxic) determine the pathology that occurs after vaccination.\"\nHowever, they should also consider that the same level of evidence for assessing causality should be applied to alternatively proposed causal factors. The mere mention of a possible alternative causal factor is not strong evidence ruling out vaccines.\nFurther, they might consider outlining the potential reality that causal interactions exist between vaccines and the proposed alternative factors as risk factors for vaccine adverse event should also not be ignored.\nOverall the writing is truly excellent, there are few sections that repeat similar information, they authors might consider combining a couple of paragraphs separated by others.\nI strongly recommend this report being indexed, with minor updates from the author.\n\nIs the rationale for commenting on the previous publication clearly described? Yes\n\nAre any opinions stated well-argued, clear and cogent? Yes\n\nAre arguments sufficiently supported by evidence from the published literature or by new data and results? Yes\n\nIs the conclusion balanced and justified on the basis of the presented arguments? Yes",
"responses": [
{
"c_id": "6231",
"date": "06 Jan 2021",
"name": "Paolo Bellavite",
"role": "Author Response",
"response": "We thank the reviewer for these positive notes and suggestions, covering two aspects that we have implemented in the secnd version. The reviewer points out that in order to correctly assess the relationship between benefits and risks of a vaccination it would be important to be able to compare the baseline health status of vaccinated versus unvaccinated children. Unfortunately, these data are lacking because systematic and adequate studies are not carried out on pediatric populations. These studies which would be feasible without too much difficulty if we had require efficient registry functions, capable of documenting the vaccination and health status informing about vaccinations and the state of health of the population. However, there are some observational studies that certainly deserve to be cited, because they can give general indications, even with in the presence of inevitable biases related to the study design. Therefore, we have briefly cited some research, such as that of Hooker-Miller and of the Aaby's group. We also mentioned Weiler's study, although he did not correctly ask for it, because it is a very relevant contribution in this context: “A contribution to evaluate the benefits and risks of vaccines, including long-term ones, could come from comparative studies between vaccinated and unvaccinated groups of children. This type of study cannot be realized in randomized groups for ethical reasons, but preliminary data can be collected anyway from accurate observational studies comparing normally vaccinated children with those who refused one or more vaccines. Unfortunately, few studies of this type are known. Here we mention those of Aaby's group47-49, who reported very different results according to the types of vaccines and regions of the world where the studies were conducted, and those of Hooker's 50 and Lyons-Weiler 51. In the first study 50 the children were vaccinated before 1 year of age and then evaluated when they reached a minimum age of 3 years. The vaccination was associated with significant higher odds of developmental delays (OR = 2.18), asthma (OR = 4.49) and ear infections (OR = 2.13). However, the study only allowed the computation of unadjusted observational associations. Higher odds ratios for such diseases were observed in quartiles where more vaccine doses were received than in quartile 1. In the second study 51 it was performed a multicenter retrospective analysis covering ten years of pediatric practice and the incidence of office visits, noting the various pathologies motivating them. The ORs indicated a notable increase in outpatient visits in vaccinated compared with unvaccinated children, because of anemia (OR = 6.3), asthma (OR = 3.5), allergic rhinitis (OR = 6.5) and sinusitis (OR = 3.5). Despite the limitations of this type of investigation and the impossibility of causal inferences, the overall results suggest that vaccinated pediatric patients are no healthier than unvaccinated ones. Further studies are needed to follow vaccinated and non-vaccinated cohorts prospectively, to understand the full spectrum of health effects associated with childhood vaccination.” As a new contribution to this important issue, we have added a final paragraph with a suggestion for possible comparative studies between vaccinated and unvaccinated, taking advantage of the known and frequent phenomenon of \"vaccine hesitancy\": “Finally, we want to relaunch a proposal 52 to investigate the possible “non-specific effects” of vaccines. The current pharmacovigilance systems are not apt to identify such effects, the observational studies cannot establish causality, and the few randomized clinical trials (RCTs) do not have sufficient size and follow up to identify and prove causality of rare or uncommon non-specific effects, even more so if they have a sizeable background rate. A solution might take advantage on the vaccine hesitancy, that remains in some individuals even after receiving an extensive and balanced information, based on the state of knowledge 52. These persistently hesitant individuals (for themselves or for their children) could be tens of thousands in a country: they could be considered as a valuable asset to society, offering them the opportunity to participate in properly designed and long-lasting RCTs. The current ethical barriers to such RCTs (the exclusion of the randomized control groups from the vaccine benefits) could be overcome because these subjects are typically evenly dispersed in a country (posing negligible risk for immunocompromised individuals 23), are really and persistently hesitant and may be willing to give their informed consent to participate in such RCTs. So this experimental approach could contribute to an advance in the scientific.” Finally, the reviewer also noted that our sentence \"This approach seems more suitable for judging complex cases, in which multiple interacting causes (genetic, infectious, toxic) determine the pathology that occurs after vaccination.\" it is itself correct, but incomplete. We agree with this opinion and therefore we have modified it by expressing better the concept, as he suggested, in the \"Perspectives\" section: “This approach seems more suitable for judging complex cases, in which multiple interacting causes (genetic, infectious, toxic) determine the pathology that occurs after vaccination. Moreover, the same level of evidence for assessing causality should be applied to the alternatively proposed causal factors. The mere mention of a possible alternative causal factor is not evidence strong enough for ruling out vaccines.”"
}
]
}
] | 1
|
https://f1000research.com/articles/9-1176
|
https://f1000research.com/articles/9-943/v1
|
07 Aug 20
|
{
"type": "Research Article",
"title": "A proposed molecular mechanism for pathogenesis of severe RNA-viral pulmonary infections",
"authors": [
"Peter K. Rogan",
"Eliseos J. Mucaki",
"Ben C. Shirley",
"Eliseos J. Mucaki",
"Ben C. Shirley"
],
"abstract": "Background: Certain riboviruses can cause severe pulmonary complications leading to death in some infected patients. We propose that DNA damage induced-apoptosis accelerates viral release, triggered by depletion of host RNA binding proteins (RBPs) from nuclear RNA bound to replicating viral sequences. Methods: Information theory-based analysis of interactions between RBPs and individual sequences in the Severe Acute Respiratory Syndrome CoronaVirus 2 (SARS-CoV-2), Influenza A (H3N1), HIV-1, and Dengue genomes identifies strong RBP binding sites in these viral genomes. Replication and expression of viral sequences is expected to increasingly sequester RBPs - SRSF1 and RNPS1. Ordinarily, RBPs bound to nascent host transcripts prevents their annealing to complementary DNA. Their depletion induces destabilizing R-loops. Chromosomal breakage occurs when an excess of unresolved R-loops collide with incoming replication forks, overwhelming the DNA repair machinery. We estimated stoichiometry of inhibition of RBPs in host nuclear RNA by counting competing binding sites in replicating viral genomes and host RNA. Results: Host RBP binding sites are frequent and conserved among different strains of RNA viral genomes. Similar binding motifs of SRSF1 and RNPS1 explain why DNA damage resulting from SRSF1 depletion is complemented by expression of RNPS1. Clustering of strong RBP binding sites coincides with the distribution of RNA-DNA hybridization sites across the genome. SARS-CoV-2 replication is estimated to require 32.5-41.8 hours to effectively compete for binding of an equal proportion of SRSF1 binding sites in host encoded nuclear RNAs. Significant changes in expression of transcripts encoding DNA repair and apoptotic proteins were found in an analysis of influenza A and Dengue-infected cells in some individuals. Conclusions: R-loop-induced apoptosis indirectly resulting from viral replication could release significant quantities of membrane-associated virions into neighboring alveoli. These could infect adjacent pneumocytes and other tissues, rapidly compromising lung function, causing multiorgan system failure and other described symptoms.",
"keywords": [
"SARS-CoV-2",
"Influenza A",
"HIV-1",
"Dengue Virus",
"Apoptosis",
"R-loop",
"DNA damage",
"RNA binding protein"
],
"content": "Introduction\n\nRNA viruses have long been known as an important source of zoonotic disease transmission1. In these infections, a key question that needs to be answered is which infected individuals will progress from mild to severe symptoms that require intensive care? While complex underlying conditions increase susceptibility, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and Influenza A can lead to severe or lethal outcomes regardless of the age or health status in certain individuals. The Chinese and the initial US patients with SARS-CoV-2 showed that higher viral replication and multiplicity of infection are evident in severely ill individuals2–4. Textbook depictions of viral release and infection indicate budding from the cell membrane. This explanation might not adequately explain the rapid onset of symptoms and transmissibility seen in some individuals infected with these agents. We suggest that these factors can be explained by a cytopathology of induced lytic events, releasing high titers of virus. Programmed cell death (apoptosis), which has been suggested to occur in RNA viral conditions such as Influenza, is activated through innate immunity, with concomitant inflammatory responses. Viral RNA has been suggested to signal Toll-Like receptors and type I interferon expression, which binds to its receptor, IFNAR, and stimulates induction of PCD genes such as FasL or TRAIL5.\n\nWe propose an alternative mechanism in which infection of RNA virus triggers unrepaired sites of chromosomal breakage, causing apoptosis and consequentially, high-titer viral release (Figure 1). This is precipitated by the binding of RNA binding proteins (RBPs) to viral genomes and transcripts instead of nuclear transcripts, to prevent destabilization of chromosome structure. This study identifies the sequences, locations and abundance of these binding sites and presents evidence for specific expression changes in DNA damage genes in Influenza and Dengue infections and evidence of expression changes consistent with induction of apoptosis. The damage is thought to arise as the result of replication forks colliding with R-loops formed by host transcripts. Ordinarily these structures are mitigated through formation of stable interactions with frequently bound endogenous RBPs7.\n\nNewly synthesized host RNA binding proteins (SRSF1, RNPS1) are required to stabilize nascent transcripts throughout the nucleus. During influenza or other viral infections, these proteins can be bound to viral genomes and transcriptomes. As viral replication and transcription proceeds, these nucleic acids containing strong binding sites for these RBPs in the cytoplasm (SARS-CoV-2) and nucleus (Influenza) that complete with host RNAs and deplete these proteins from the nucleus. This enables nascent transcripts to reanneal with transcription templates, and R-loops are formed. If not removed by RNAse H or other helicases, unresolved R-loops at numerous genomic loci triggers genomic instability. Their frequency and density of unrepaired chromosome damage would be expected to overwhelm DNA repair components (BRCA1/2, FANC complex, and XPC), inducing multiple chromosomal strand breaks in each cell6. These breakage events initiate apoptosis, releasing a high multiplicity of infectious viral particles.\n\nThe SR protein family consists of RNA binding proteins that play significant roles in the regulation of mRNA splicing8. SRSF1 (formerly ASF/SF2) is an exonic splicing enhancer (ESE) that has been shown to interact with the U1 snRNP and recruit the protein to the donor (5’) splice site9,10. However, binding of SRSF1 to nascent transcripts has also been shown to play a significant role in genome stability, first described in reference 6, whereby the presence of SRSF1 bound to pre-mRNA repressed the formation of DNA:RNA hybrids, which led to R-loops, double-stranded breaks, and a hypermutation phenotype. This phenotype could be corrected not only by increasing RNase H expression (to eliminate DNA:RNA hybrids), but with the overexpression of the RNA binding protein RNPS111. RNPS1, part of the apoptosis-and splicing-associated protein (ASAP) complex, can directly interact with SRSF112 and could possibly help recruit SRSF1 to ESE sites13. Other RNA binding proteins have been shown to increase genome instability when depleted, including THOC114, MFAP115, and FIP1L116.\n\nBinding sites for these RBPs are identified using information theory (IT)-based sequence analysis, which has proven both theoretically and in numerous practical examples to be an accurate approach for predicting binding affinities of nucleic aid sequences recognized by particular DNA or RNA binding proteins17. IT can be used to identify binding sites, and to evaluate the impact a sequence variant may have on binding site strength18. IT has been applied in studies which involved mRNA splicing19,20, splicing regulatory factors (SRFs21,22), other RNA binding proteins23 and transcription factor binding sites (TFBS;24,25), and has been used to accurately predicted level of gene expression and identify causative mutations in a wide spectrum of diseases17. IT-based analysis has the distinct advantage to other bioinformatic approaches as the predicted information content (known as Ri; measured in bits) can be quantified as binding site affinity as it is related to thermodynamic entropy26. The binding affinity of a sequence predicted by IT has been shown experimentally to directly relate to the observed binding quantity of said sequence26. IT-based models are generated from a series of annotated binding sites for a particular RBP. The average strength of the sites used to generate said is referred to as its Rsequence. IT-based models can also be derived from high-throughput binding site identification techniques such as ChIP-seq (e.g. the derivation of TFBS models in 24). Information density-based clustering (IDBC) analysis, where groups of closely situated binding sites are evaluated based on their combined strength (their “information density”) and intersite distances, has been applied along with these TFBS models in both the identification of TFBS-dense clusters, and accurate prediction of gene expression patterns25.\n\nWe suggest that the viral genome binds these also and define the locations of likely strong binding sites across the genomes of various RNA viruses. As it replicates, we propose that the viral genome binds these proteins, preventing their reimportation into the nucleus where they are normally needed for essential post-transcriptional activities. We theorize that incremental replication and transcription of viral RNAs in the cytoplasm creates a sink for these proteins, starving the host nucleus, and initiating a series of events that release viral particles into the lumen, enabling rapid infection of neighboring lung epithelial cells (Figure 1). An infographic has been created to provide a detailed step-by-step guide to the proposed mechanism, from the initial viral infection to spread of infection to the lungs and other major organs, leading to lowered blood oxygen levels, and multi-system organ failure27.\n\nRNA viral genomes of Influenza viruses replicate in the nucleus and are processed by host RNA spliceosomes. For example, the M and NS segments of the Influenza genome are processed using the host splicing mechanism28. Viral RNAs, like host transcripts, are capable of sequence specific binding to RBPs. This can conceivably deplete RBPs from host encoded RNAs, where they ordinarily function. These unbound RNAs are capable of hybridizing to the non-template derived strand of the chromosome29. RNA naturally forms a stronger bond to DNA than DNA does to itself, especially rG:dC hybrids6. As a result, mRNAs would replace DNA by hybridizing complimentary bases, resulting in R-loop formation, and can lead to DNA damage.\n\nThe RNA spliceosome regulator SRSF1 acts on exonic splicing enhancer sequences in pre-mRNA and forms RNP complexes with nascent mRNA precursors. Aside from its established role in enhancing exon recognition10, binding of SRSF1 to these transcripts is required to prevent or destabilize the formation of R-loops6. R-loops are derived from RNA transcripts that anneal to the chromosomal strand complimentary to the transcription template stand. If not eliminated, these structures pose a threat to genomic integrity as targets for DNA damage. The structure of R-loops consists of two duplex-single strand junctions which are recognized by nucleases that cleave the DNA29. DNA fragmentation causes a G2 phase cell cycle arrest which can potentially lead to cell death11. R-loops that are not targeted by nucleases are nonetheless still non-functional and thus, inflict damage on the cell6. As RNA viruses enter the cell and replicate, the nucleic acid sequences they encode divert RBPs such as SRSF1 away from binding to nuclear RNA transcripts, thus promoting the creation of R-loops.\n\nRNPS1 is a pre-mRNA splicing activator protein that functions together with SRSF1 to form RNP complexes on nascent transcripts13,30, but also has a role in preventing transcriptional R-loop formation11. RNPS1 also suppresses high molecular weight DNA fragmentation at high expression levels. These two proteins work together but have independent mechanisms as RNPS1 cannot compensate for SRSF1 splicing function in its absence and vice versa11.\n\nIn Dengue virus, the protein called NS5 binds to host spliceosome complexes and modulates endogenous splicing to change mRNA isoform abundance of antiviral factors. By also interacting with U5 snRNP particles, it reduces the efficiency of pre-mRNA processing, hence resulting in a less restrictive environment for viral replication. It has also been shown that NS5 interacts with the host protein, RNPS1, which disrupts normal nuclear RNA binding processes31.\n\nViral infections interfere with post-transcriptional processing of host pre-mRNA including splicing, capping, and translation during viral invasion. Since SRSF1 binds and interacts with pre-mRNA during the earliest stages of splicing, diversion of SRSF1 and other spliceosomes to other RNA sequences depletes the cell’s resources. Normally, cellular mRNA is 7-methylguanosine cap is added to the 5’ end to protect the sequence from degradation. However, Influenza carries proteins that has “cap-snatching” abilities32. Influenza snatches the 5’ cap by cleaving the mRNA 10 to 15 nucleotides away from the guanosine and this cap is used to prime transcription of the virus. Finally, during viral infections, all RNA processing mechanisms are now being shared between two genomes. Ultimately, as transcriptional and translation mechanisms fail to facilitate the mRNA, they will create R-loops with DNA, cause DNA damage, and induce higher expression of DNA repair genes (such as DDB2; see Results).\n\nUnrepaired damaged DNA that encounters a replication fork leads to unresolved double strand breaks, triggering apoptosis. The quantity of virus that escapes into tissues, blood and other conduits (e.g. lymphatic), and other systems would likely dwarf the amount that is released by conventional viral budding from the cell membrane. This viral load will likely overwhelm the immune system in individuals who are already immune deficient and might provoke a systemic inflammatory response (like a cytokine storm). However, the high titer of virus is likely to infect neighboring cells and other tissues. The extent of the apoptotic response may be the distinguishing finding which separates the patients who survive the infection from those who end up in intensive care, develop pulmonary insufficiency and multi-system failure.\n\nThe deficiency in SRSF1 and other RBPs in the nuclei of Influenza, Dengue or SARS-CoV-2 infected cells does not require any specialized mechanism. Assuming that the virus is replicating freely in the cytoplasm (or nucleus, in the case of Influenza), the significant excess of unpackaged, replicated viral RNA acts as a sponge to sequester newly synthesized, folded RBPs. Based on mass action, the quantity of RBPs that would be transported into the nucleus for host mRNA processing would have a much-diminished nuclear stoichiometry in comparison with normal, uninfected cells.\n\n\nResults\n\nCells depleted of SRSF1 has been shown to have unstable genomes which can be corrected by overexpression of RNPS111. In order to investigate the significance of SRSF1 and RNPS1 binding in viral genomes, we first developed information theory-based models for the recognition sequences for each of these proteins using binding site datasets derived from transcriptome-wide RNA binding protein datasets of CLIP sequencing data. We then scanned multiple RNA viral genomes, as well as the human transcriptome, with these derived models to identify and predict the strength of individual binding sites.\n\nAn Information Weight Matrix (IWM) for SRSF1 has been previously derived21, however, it was only based on very small set of manually curated binding sites (N=28). We therefore derived new SRSF1 IWMs using publicly available eCLIP data (two separate replicates from reference 33). Multiple SRSF1 models exhibited very similar binding motifs, however, their differences justified our analyses using the two most divergent IWMs in this study. These models are referred to as SRSF1 “Replicate 1” and “Replicate 2” models, as they are models derived from two separate eCLIP experimental replicates from the same study. SRSF1 “Replicate 1” is derived from a larger number of eCLIP peaks (50,000) compared to 5,000 for “Replicate 2”. Since SRSF1 “Replicate 1” was derived from a greater number of sites, it therefore may be more accurate for detection of weaker SRSF1 binding sites.\n\nA distinct IWM was derived by iCLIP data from transcriptome-wide, protein crosslinking to sequences recognized by RNPS130. It was evident that the RNPS1 IWM and the newly derived SRSF1 models exhibited a similar pattern of nucleotide conservation based on comparison of their respective sequence logos (Table 1). STAMP, a program which analyzes DNA-binding motifs, was used to compare these models34. The SRSF1 “Replicate 1” and “Replicate 2” models were both highly similar (motif alignment e-value < 0.01) to the RNPS1 IWM (Table 1), implying that individual binding sites recognized by these two factors are similar. Indeed, the motif similarity between these two factors has been described13. We suggest that this overlap in their respective binding affinities may account for why RNPS1 overexpression can enable SRSF1-deficient cells to overcome their inherent genomic instability phenotype.\n\n1 RNPS1 model derived from publicly available iCLIP data (E-MTAB-4215; ArrayExpress), while SRSF1 models were derived from eCLIP data (ENCSR456FVU; ENCODE Data Coordination Center); 2 SRSF1 [Rep1] and [Rep2] were derived from eCLIP dataset replicate 1 [50,000 peaks] and replicate 2 [5,000 peaks], respectively; 3 RNA binding motifs were compared using STAMP34 using the Pearson Correlation Coefficient distance metric37; 4 A549 cell line expression from GSE141171 dataset; 5 Primary type II pneumocyte expression from GSE86618 dataset; 6 Influenza A virus H3N2 strain (Ontario/104-25/2012). 7 RNPS1 sites used as denominator for all percentages.\n\nThe newly derived SRSF1 and RNPS1 models (as well as an hnRNP A1 model to act as a positive control [its derivation described in 22], as the RBP has been shown to regulate transcription of beta coronaviral genes35) were used to scan the genomes of multiple RNA viruses: Dengue (Type 3), HIV (Strain B and C), Influenza A (H3N2; two separate strains), and SARS-CoV-2 (NC_045512.2). In coronaviruses, the infectious particle contains the positive strand, but the negative strand copy of the RNA is generated for protein translation36 and may be available to bind RBPs. Therefore, both the positive and negative strands of the viral genomes were scanned for SRSF1, RNPS1 and hnRNP A1 binding, regardless of the replication mechanism of the virus.\n\nThe SARS-CoV-2 genome was determined to contain >600 SRSF1 (with either SRSF1 model) and RNPS1 binding sites (Table 1). However, histograms which illustrate the distribution of the strengths of all SRSF1 and RNPS1 binding sites in SARS-CoV-2 (Figure 2A) reveal that the majority of these are weak sites (where Ri < Rsequence) that may not be used. We therefore focused downstream analysis on strong binding sites (where Ri ≥ Rsequence) of each IWM (Rsequence: 6.7 bits for the SRSF1 “Replicate 1” model; 6.4 bits for the SRSF1 “Replicate 2” model; 7.8 bits for the RNPS1 model; and 4.6 bits for the hnRNP A1 model). There are only 35 RNPS1 and between 31-60 SRSF1 binding sites (depending on SRSF1 model) on the positive strand of the SARS-CoV-2 genome that meet this Rsequence threshold (Table 1). The total number of SRSF1 binding sites within all other viral genomes tested are provided in Table 2, while RNPS1 and hnRNP A1 binding site counts are available within a Zenodo repository for this study (extended data38 Section 1 – Table 1). The hnRNP A1 model consistently predicts more strong binding sites than the SRSF1 and RNPS1 models across all the RNA viral genomes tested, as well as in the human gene controls. This is likely partially due to its relatively low Rsequence threshold compared to the other models used. Interestingly, we observed significantly more SRSF1 and RNPS1 binding sites on the positive strand compared to the negative strand for all tested RNA viral genomes (exception: sites in SARS-CoV-2 predicted by SRSF1 “Replicate 1” model). This phenomenon was observed in both positive-strand and negative-strand RNA viruses (e.g. both Influenza A strains tested). This imbalance was not observed in the human genes tested (Table 2).\n\nHistograms display the distribution of Ri values for SRSF1 [“Replicate 1” and “Replicate 2” models] and RNPS1 binding sites strengths identified in A) the SARS-CoV-2 viral genome, and B) all transcribed regions in the human genome.\n\nColumns labeled as “Both” indicate the number of binding sites or clusters on both strands of the viral genome.\n\nPreviously, tightly organized groups of transcription factor binding sites (TFBS) were identified using information dense clustering25,39. We applied this method to identify regions of the viral genomes with large concentrations of binding sites (extended data38 Section 1 – Table 2). Clusters of weak SRSF1 and RNPS1 sites are common (e.g. there are 5 SRSF1 clusters on the positive strand of SARS-CoV-2; extended data38 Section 1 – Tables 2A and 2B); however, clusters made up exclusively of strong binding sites (Ri ≥ Rsequence) are extremely rare in the viral genomes tested.\n\nWe observed that all strong RNPS1 sites were also predicted to be strong (Ri ≥ Rsequence) by the SRSF1 “Replicate 2” model. This is not surprising, as the two models were found to have significantly similar binding motifs (Table 1). This overlap, as well as the location and strength of all other strong SRSF1 (“Replicate 2” model only) and RNPS1 binding sites, can be observed in Figure 3 where sites were mapped across the SARS-CoV-2 and Influenza A genomes. This was not observed, however, for SRSF1 “Replicate 1” despite its similarity to the RNPS1 model. For this SRSF1 model, nearly half of all strong RNPS1 sites were predicted to be weak (Ri below the Rsequence threshold).\n\nThe viral genomes of A) SARS-CoV-2 (NCBI Reference Sequence: NC_045512.2) and B) Influenza A virus (A/swine/Ontario/104-25/2012[H3N2]) were scanned for strong pre-existing binding sites for the RBP RNPS1 and SRSF1 (newly derived “Replicate 2” model). Custom wiggle tracks which contained those RBP of Ri ≥ Rsequence were generated and visualized by NCBI Nucleotide. Track images were manually adjusted to indicate the strand in which the binding site was identified (blue vertical lines indicate sites on the positive strand, orange on the negative strand). The majority of sites predicted by the RNPS1 model were simultaneously predicted by the SRSF1 model, however the SRSF1 model identifies additional unique binding sites.\n\nDespite its low mutation rate, over 220 SARS-CoV-2 strains have already been identified, with potential mutational hot spots of different geographic origins40. If the proposed mechanism does play a role in the severity of infection, then it is expected that various strains of SARS-CoV-2 would not significantly differ in numbers of binding sites, as no particular strain of SARS-CoV-2 has yet been proven to affect disease recovery (indeed, more transmissible strains have been identified but none more pathogenic41,42). To test this theory, genomes of 8 SARS-CoV-2 strains were downloaded from the Global Initiative on Sharing All Influenza Data (GISAID) database and analyzed using the IWMs for SRSF1, RNPS1 and hnRNP A1 (Table 3 for positive strand analysis; extended data38 Section 1 – Table 3 for analysis of both strands). The particular strains that were selected were those that showed maximum divergence from one other based on analyses by NextStrain (which tracks the genomic epidemiology of SARS-CoV-243). Binding site counts of different strains were within 90% across all strains, except for MT198652.1 (Spain), which contains an undetermined sequence where binding site differences are mapped. A strong consistency between binding site counts and strengths was noted, despite maximizing in the divergence between the selected SARS-CoV-2 strains. For RBPs binding, it was therefore not significant as to which SARS-CoV-2 sequence was selected for the subsequent analyses.\n\na MT188339.1 strain was found to contain an extra SRSF1 binding site (original SRSF1 model). This additional 6.9 bit site is located within the large \"orf1ab\" gene. This site is caused by a T>C substitution (TGGAGGT -> CGGAGGT; 0.7 -> 6.9 bits); b Sequences contains a small stretch of undefined nucleotides, which is likely contributing to the lower number of binding sites found.\n\nThe absence of severe symptoms associated with the SARS-CoV-2 Singaporean strain (which features a deletion in ORF8 [pos. 27,848 to 28,229]44), however, is not related to a significant loss of strong SRSF1 and RNPS1 binding sites. The SRSF1 “Replicate 1” model does not identify any binding sites (≥ Rsequence) in this region. The SRSF1 “Replicate 2” model predicts 2 strong binding sites in this region, as does the RNPS1 model. There are 17 hnRNP A1 binding sites in this region, however there are 1168 sites in total across the coronavirus genome; therefore, the missing hnRNP A1 sites account for only 1.4% of the total detectable hnRNP A1 binding sites.\n\nGiven the high Influenza A mutation rate, we evaluated the variability in RBP site count and affinities between strains, that is, whether these binding sites might be under selection for conservation of RBP binding. Four Influenza A strains (H3N2) from four separate clades (analogous to the SARS-CoV-2 strain selection procedure using NextStrain43; A/Denmark/316/2020; A/England/323/2019; A/Singapore/TT0333/2019; and A/Sydney/1017/2018) along with the two Influenza A strains previously selected (A/swine/Ontario/104-25/2012 and A/Duck/Shanghai/C84/2009) were analyzed and their genomes scanned for the presence of strong RNPS1, SRSF1 and hnRNP A1 binding sites (extended data38 Section 1 – Table 4). Depending on the strain, 13 to 16 RNPS1 and 30 to 35 SRSF1 (“Replicate 2” model) binding sites were identified on the negative strand of Influenza A (a range of 16–23 binding sites for SRSF1 “Replicate 1”, and 221 to 241 strong hnRNP A1 binding sites). Thus, it appears as though the overall number of binding sites remains relatively consistent between each Influenza A strain, despite their divergent genomic sequences.\n\nThe locations of all predicted binding sites and information-dense clusters within the genome of each RNA virus tested has been made available within the extended data archive (Section 238). This data is provided in the form of ‘bedgraph’ genome browser tracks. The location of binding site clusters are also provided as lollipop plots within the archive (Section 3), as are the IWMs used to evaluate each site (Section 4).\n\nEach of these RNA viral genomes contain multiple strong RNA binding sites. The frequency of RBP binding in human transcriptomes was determined to relate the relative abundance of these proteins bound to viral RNAs compared to their normal reservoir in host nuclear RNA of infected cells. Expressed host gene sequences were scanned with IWMs for SRSF1, RNPS1 and hnRNP A1 to locate all potential binding sites throughout transcribed regions of the human genome, then partitioned among these genes based on their abundance in relevant cell types. These were compared with binding sites within 300nt of a known exon, as many of these RBPs have critical functions in exon recognition and maturation of mRNA splice isoforms (provided as bedgraph tracks in the Zenodo archive [Section 2]38). While the majority of these binding sites are considered weak (Ri < Rsequence; Figure 2B), the numbers of strong (binding sites with Ri > Rsequence) residing within transcribed regions are substantial (SRSF1 “Replicate 1” Model: 5,543,429; SRSF1 “Replicate 2” Model: 8,275,472; RNPS1: 4,368,943; hnRNP A1: 44,885,381). The intersite distance (the average distance between binding sites) appears to be inversely related to the overall number of binding sites, as the mean intersite distance between strong hnRNP A1 binding sites was considerably shorter than the distance between strong SRSF1 and RNPS1 binding sites (hnRNP A1: 24±40 nt; RNPS1: 149±248 nt; SRSF1 [“Replicate 1” model]: 105±241 nt; SRSF1 [“Replicate 2” model]: 89±197 nt; analysis using a maximum intersite distance threshold of 1000nt). Regardless of these differences, however, this analysis illustrates that many strong binding sites are separated by < 200nt and highlights how densely arrayed these sites are in the human transcriptome.\n\nThe number of strong SRSF1, RNPS1 and hnRNP A1 binding sites (Ri ≥ Rsequence) were enumerated by gene (extended data38 Section 1 – Table 5 [A–D]; genes without any strong binding sites are not listed). Similar tables were created which count the number of information-dense clusters located within each gene (extended data38 Section 1 - Table 5 [E–H]). In general, there were more hnRNP A1 clusters identified than SRSF1 and RNPS1 clusters (SRSF1 “Replicate 1” Model: 112,955; SRSF1 “Replicate 2” Model: 98,872; RNPS1: 39,285; hnRNP A1: 709,226), which is likely due to the higher frequency of strong hnRNP A1 sites and significantly lower hnRNP A1 intersite distance. Table 5 (from extended data38 Section 1) also provides type II pneumocytes (from single-cell [sc] RNAseq data) and the A549 (human alveolar adenocarcinoma) cell line (RNAseq) expression values for each gene listed (in Transcripts Per Million [TPM]). Genes that are both highly expressed in lung cells and contain a high frequency of SRSF1 and/or RNPS1 information-dense binding site clusters would be considered strong candidate genes for R-loop formation in cells infected by an RNA virus. The gene PTPRN2 has the highest total number of SRSF1 clusters (N=116 to 138 depending on the SRSF1 model used) but has relatively low level expression in pneumocytes (TPM = 0.052). The THSD4 gene, however, has 35-36 high-density SRSF1 clusters (N=2236-3475 individual strong SRSF1 binding sites) and is expressed (≥ 1 TPM) in both lung cell expression data sets tested (Figure 4A; extended data38 Section 1 - Table 5 [E and F]). Overall, there are 1,225 genes with ≥10 SRSF1 and 127 genes with ≥10 RNPS1 information-dense clusters which are also expressed (TPM ≥ 1) in the expression datasets tested.\n\nA) Lollipop plot of information density of clusters annotated by coordinate range and number of sites comprising that cluster using the SRSF1 “Replicate 2” information-based weight models (all Ri ≥ Rsequence) for the NM_024817 mRNA splice form of THSD4 (some clusters counted in S5 Table are found in other THSD4 splice forms which span beyond the range of this particular mRNA). B) Information dense SRSF1 clusters within THSD4 that overlap a DRIPc-seq interval (GSE70189 DRIPc-seq dataset). One additional overlapping cluster is not displayed, as is located immediately upstream of the 5’ untranslated region of the NM_024817.2 splice form. No intervals from the GSE68845 DRIP-seq dataset overlap this gene.\n\nDRIP (DNA-RNA immunoprecipitation) sequencing is a high-throughput method of identifying regions of the genome where R-loops can form. DRIPc sequencing is an improvement which provides higher resolution mapping data in a strand-specific manner45. To determine to what degree these DRIP-seq (GSE68845 [IMR90 cells]) and DRIPc-seq intervals (GSE70189 [NTERA2 cells]) overlapped RNPS1 and SRSF1 binding sites in uninfected cells, we performed an intersection between the two datasets and information dense clusters (extended data38 Section 1 – Table 6 [A and B]) or individual binding sites (extended data38 Section 1 – Table 6 [C and D]). It was uncommon for strong binding site clusters to overlap a DRIP-seq interval (0.4 – 1.7% of all transcriptome-wide clusters overlap a DRIP-seq interval). Despite an additional level of filtering (where the strand of the clusters and DRIPc-seq intervals must match), the frequency of overlap between binding site clusters and DRIPc-seq was much higher compared to the frequency of overlap to the DRIP-seq dataset (~15-17% overlap depending on IWM; extended data38 Section 1 – Table 6A). In all test cases, limiting analysis to only those genes that are expressed in A549 cells (≥1 TPM) increased the percent overlap of clusters and both DRIP- and DRIPc-seq data sets (e.g. we find a 15.3% of RNPS1 clusters/DRIPc-seq overlap among all genes, but 20.2% overlap when considering expressed genes in the A549 cell line only). When this analysis was repeated but limited to only those clusters near an exon (within 300nt), this also showed a significant increase in the fraction of clusters overlapping DRIP-seq intervals (extended data38 Section 1 – Table 6B). These observations remain consistent when considering individual binding sites, rather than binding site clusters (extended data38 Section 1 – Table 6C and 6D). It therefore seems that the vast majority of individual binding sites and information-dense binding site clusters do not overlap these DRIP- and DRIPc-seq regions. For example, only 5 of 36 clusters within THSD4 overlap the DRIPc-seq dataset (Figure 4B; extended data38 Section 1 – Table 5F).\n\nInterestingly, the computed intersite distances for RNPS1, SRSF1 and hnRNP A1 binding sites that overlap DRIPc-seq intervals were shorter compared to the intersite distances of sites across the entire transcriptome (mean intersite distances: hnRNP A1: 22±45nt; RNPS1: 120±228nt; SRSF1 [“Replicate 1” model]: 76±205nt; SRSF1 [“Replicate 2” model]: 69±170nt; maximum intersite distance of 1000nt). The general distributions of intersite distances between these two analyses were also found to be quite similar (extended data38 Section 5). As we are limiting this analysis to sites that are within a few, often short DRIPc-seq intervals, the distances between pairs of sites are likely to be tightly grouped. We also computed the average number of all binding sites and clusters, and only those which overlap the DRIPc-seq dataset, for each individual gene (sites and clusters per 100nt of gene length; extended data38 Section 1 - Table 5). Binding site densities within specific genes are reduced for sites overlapping DRIPc-seq intervals (e.g. THSD4 SRSF1 cluster density reduces from 5.2E-03 to 7.0E-04 clusters per 100nt).\n\nWe have previously described a machine learning (ML) based approach for developing gene signatures for expression various environmental exposures to cells, initially focusing on prediction of chemotherapy effects46. This method was applied to ionizing radiation data, from which accurate gene signatures were derived that could differentiate levels of radiation exposures. In particular, low exposures were distinguished from higher radiation levels that cause Acute Radiation Syndrome (ARS47). ARS is characterized by vomiting, diarrhea, fever, low white blood cell count and fatigue. Physicians might not consider ARS in the differential diagnosis when presented with a patient exhibiting these symptoms, since Influenza and Dengue (viral) infections also present with vomiting, diarrhea, lymphopenia (especially Influenza H1N148) and fatigue, and are more common. Like ARS, these conditions lead to death in some cases. While Influenza A has a worldwide distribution, Dengue is more prevalent in Southeast Asia, the Americas and the Western Pacific where it presents typically with severe manifestations including hemorrhagic fever and shock. We have considered how the life cycle of these viruses might be related to the corresponding cellular responses.\n\nExpression data from irradiated blood samples were used to derive the human radiation gene signatures reported in Zhao et al.47. While it was assumed that these ML models were specific for diagnosing ARS, the models were further tested to determine if they could distinguish ARS from other conditions that share similar clinical presentation (e.g. vomiting, diarrhea). Four human ML radiation signatures from Zhao et al. (assessed by traditional validation; denotated as ML models “M1”, “M2”, “M3” and M4” which are described in extended data38 Section 1 – Table 7) were used to evaluate 11 gene expression studies of patients infected with: Influenza (N=5, includes Influenza A [H3N2], swine flu [H1N1] and Influenza B viral infection data sets), Dengue virus (N=4) and aplastic anemia (N=2). On average, the ML models misclassified 26.4% of Influenza and 22.4% of Dengue patients as irradiated (Section 1 – Table 7). Approximately 15% of aplastic anemia patients were also misclassified. The model “M1” showed the lowest misclassification rate against Influenza patients (9-29% of patients misclassified), models “M2” best classified Dengue-infected patients (7-33% misclassified), while models “M1” and “M3” performed well with patients with aplastic anemia (5-20% misclassified for “M1” and 0-14% misclassified for “M3”). In nearly every instance, the inclusion of normal controls from the Influenza and Dengue studies improved overall accuracy of all four ML models (17.4% and 18.1% average misclassification of Influenza and Dengue-infected patients, respectively). This phenomenon was not observed in the aplastic anemia dataset tested. The observation that normal controls are more often correctly classified indicates that these models are not so much incorrectly classifying infected patients, as they are identifying gene expression differences that may be a response to or caused by the viral infection itself.\n\nThe four radiation gene signatures assessed from Zhao et al.47 consist of 32 unique genes. When performing feature removal analysis (where model accuracy is reassessed after each gene is individually removed from it), 10 genes were identified that greatly contribute to patient misclassification: DDB2, PCNA, GTF3A, PRKCH, CDKN1A, GADD45A, BCL2, MOAP1, TRIM22 and TALDO1 (extended data38 Section 1 – Table 8). DDB2 is a DNA damage binding protein that is present in all four ML models. DDB2 expression levels were elevated in irradiated patients, which is likely due a cellular response to radiation exposure, as this gene participates in nucleotide excision repair (it ubiquitinates histones H3 and H4 to increase accessibility of nucleosomes, exposing DNA and enabling access to XPC [xeroderma pigmentosum group C-complementing protein], which performs NER49,50). DDB2 shared a similar pattern of expression between irradiated samples as well as infected patients that were misdiagnosed as irradiated (elevated DDB2 expression in misclassified Influenza and Dengue patients; Figure 5). The activation of DDB2 would be consistent with the proposed mechanism, whereby high levels of RNA viral genome increase the formation of abnormal, unresolved R-loops which in turn activate a DNA damage response. Expression of DDB2 between those correctly classified and those misclassified as irradiated was deemed significant by the Mann Whitney test (p-value = 0.0001). Other genes with significant differences in expression included GTF3A, PRKCH and PCNA (which also has a role in the DNA damage response; extended data38 Section 1 – Table 8).\n\nDDB2 expression for Model 1 including Influenza patients, controls and radiation patients plotted using GraphPad. Each colour represents a different dataset. The left distribution of the radiation data (shaded grey) represents the expression of the radiated patients and the right distribution represents unirradiated controls. For all Influenza datasets (coloured), the left-most distribution represents the true negatives, the middle distribution represents the false positives, and the right-most distribution represents uninfected controls.\n\nIn the mechanism proposed (Figure 1), the fraction of SRSF1 and RNPS1 bound to host RNA decreases as the fraction of SARS-CoV-2 genome increases as it replicates in the cell, causing RNA:DNA hybrids which result in R-loops. We therefore estimate the quantity of viral genomes and extent of viral replication required for viral binding site counts to approach, match, and exceed the number of host RNA sites available. These are derived from the number of SRSF1 and RNPS1 sites expressed in either a single A549 cell or a type II primary pneumocyte. The overall expression of each host gene was normalized by dividing by total expression of the given dataset, then by multiplying the number of all binding sites within a gene to its normalized gene expression value, and finally by multiplying the sum of all expression-adjusted binding site counts by the expected number of mature RNAs in a cell. We estimate a total of 80,000 RNAs per single cell (as determined by Marinov et al.51), which is comparable with totals determined in other studies (e.g. Xia et al.52 determined that a single osteosarcoma cell contains 92,000 ± 32,000 mature RNAs).\n\nBased on this approach, the total number of expressed binding sites (of any strength) was computed for SRSF1 and RNPS1 (Table 1). However, this estimate includes sites expected to be weakly binding. When taking only strong binding sites into account, we estimate 12.7 to 18.2 million expressed SRSF1 (“Replicate 1” and “Replicate 2” SRSF1 models, respectively) and 9.9 million expressed RNPS1 binding sites in a single A549 cell. In a single primary pneumocyte, we estimated 6.6 to 9.4 million expressed SRSF1 sites (“Replicate 1” and “Replicate 2” models, respectively), as well as 5.2 million expressed RNPS1 binding sites. These estimates are based on expression levels in normal cells and may differ in infected cells. While the dissociation constant for RNPS1 is unknown, the dissociation constant of SRSF1 (KD) bound to the RNA sequence 5′-UGAAGGAC-3′ has been experimentally measured as 0.8 μM53. With the KD, a Scatchard plot for SRSF1 binding was derived where host bind sites are substrates and viral binding sites are considered to be inhibitors of host RNA binding. We assumed no free RNA binding protein, which is bound to either host or viral binding sites. This assumption is reasonable for strong binding sites (where Ri ≥ Rsequence). We use KD to compute the theoretical number of viral genomes required to satisfy various viral genome to host binding site ratios (Figure 6 [Table left]). This calculation is also carried out without reference to KD, by instead computing the number of viral genomes required to achieve binding site ratios in viral to host-bound RBP from a direct analysis of primary pneumocyte and A549 transcriptomes. The number of strong SRSF1 binding sites in a single viral genome multiplied by the level of viral replication is compared with the estimated number of expressed SRSF1 sites in the host nucleus (in a pneumocyte or an A549 cell; Figure 6 [Table right]). The data presented in Figure 6 uses the number of sites predicted by SRSF1 “Replicate 2” model, and only considers the positive strand of SARS-CoV-2. Despite their similarities, the SRSF1 “Replicate 2” model predicts far more binding sites on the positive strand of SARS-CoV-2 compared to the “Replicate 1” model (N=60 and 31, respectively). This leads to small differences in the estimated doubling time, when only the positive strand of the virus (extended data38 Section 1 – Table 9A) is considered. An examination of potential binding sites on both strands of SARS-CoV-2 does not appreciably alter the estimated doubling time for both SRSF1 IWMs (extended data38 Section 1 – Table 9B).\n\nAs the fraction of SARS-CoV-2 genomes increase in the host cell, the fraction of SRSF1 bound to the host transcriptome versus the viral genome decreases, resulting in R-loops. Strong SRSF1 binding sites (“Replicate 2” model) were identified in both SARS-CoV-2 (N=60 on the positive strand) and in the human transcriptome. A Scatchard plot (right) was created and used to determine the theoretical number of viral SRSF1 binding sites expected at different viral genome (inhibitor) to host (substrate) ratios (left).\n\nThe doubling times required for infection initiated by a single virion were computed for varying numbers of viral genomes, as replication increases the overall counts of viral RBP binding sites. The processivity rate of genome replication for SARS-CoV-2 is currently unknown, so a value was estimated based on a polymerization rate of 3.7 nt/s for a different RNA-dependent viral RNA polymerase, that of Vesicular Stomatitis Virus (VSV)54. The doubling time was then adjusted to 2.31 hours per replication event, based on the increased length of the SARS-CoV-2 genome (L=30,899nt) compared to VSV. The doubling time is estimated to be between 32.5 to 44.1 hours to achieve a level of SARS-CoV-2 binding that depletes RBP from an equal number of expressed host nuclear RNA sites (1:1 ratio). However, fewer replication events and shorter doubling times are computed using the published KD of SRSF1 (between 5-14 hours less). The number of replication events required for viral genome binding sites to overtake host RNA binding was less in primary pneumocytes compared to A549 cells (~2.3 hours or 1 doubling of the SARS-CoV-2 genome). This was anticipated, since the total number of expressed SRSF1 (and RNPS1) sites are lower in primary pneumocytes than the immortalized cell line due to lower overall gene expression levels.\n\n\nDiscussion\n\nWe propose a previously undescribed putative mechanism of RNA viral infection-induced apoptosis, supported RNA binding events determined by information theoretic analysis. In the mechanism, viral release is enhanced by viral genome replication, which sequesters RBPs, thereby depleting native binding of RBPs to and stabilization of host-encoded transcripts. This process can occur in either the cytoplasm or the nucleus of the host cell, depending on specific replication requirements of different viral families. In SARS-CoV-2, this is expected to substantially reduce import of RBPs into the nucleus. Reduced availability of nuclear RBPs promotes R-loops through formation of complimentary duplexes between nascent transcripts and chromosomal sequences. High densities of R-loops at a late stage of infection would be expected to overwhelm cellular DNA repair mechanisms that ordinarily remove these structures and eliminate DNA breakage. DNA damage markers DDB2 and PCNA are increased in both Influenza and Dengue infections, respectively. Unrepaired, persistent chromosome double strand breaks are unstable and induce apoptosis, which would be expected to release high viral titers.\n\nWe utilized a well-established information theory-based approach to demonstrate the validity of this proposed mechanism17–24. IT-based models of RBP binding sites was used to scan viral RNA genomes (Influenza, SARS-CoV-2 and Dengue) and host transcriptomes. IT models derived from thousands of validated RBP binding sites delineated numerous strong SRSF1, RNPS1 (and hnRNP A1) RNA binding sites within these viral genomes. The derived SRSF1 and RNPS1 binding motifs were shown to be highly similar, consistent with previous published studies demonstrating that RNPS1 could partially complement genomic instability due to SRSF1 deficiency. Indeed, both models detected many of the same RNA binding sites in the host transcriptome and all strong RNPS1 binding sites detected in the SARS-CoV-2 genome were simultaneously detected by at least one SRSF1 information model. In divergent strains of both SARS-CoV-2 and Influenza A (H3N2), the frequencies and strengths of these binding sites are highly consistent. Finally, we estimate that the quantity of replicated viral genomes necessary to meet or exceed the number of binding sites expressed within a lung can exceed the site counts in the host genome, and the doubling time required to deplete these RBPs which is consistent with the observed time course of severe infections.\n\nFunctional analyses will be needed to prove that this mechanism plays a role in viral pathogenicity. Such studies should further investigate how infections of SARS-CoV-2 (and other RNA viruses) cause increased DNA damage. RNAseq and protein expression analysis of DDB2, RAD17, PRKDC, PCNA and other ATR pathway markers of infected cells accompanied by time course studies of nascent double stranded chromosomal breaks (i.e. H2AX antibody staining due to viral infection) would provide such evidence. Increased R-loop formation upon infection will be required, with particular attention to host encoded transcripts enriched in SRSF1 and RNPS1 binding site clusters. Although the genomic coordinates where R-loops form can be anticipated from information dense clustering, the strand- and gene specific techniques used to detect these, i.e. DRIPc-Seq, cannot measure RNA-DNA hybrids of lengths shorter than 70bp45. Sequence-based chromatin immunoprecipitation with antibodies to H2AX, 53BP1 or other markers of DNA damage should be consistent with the sites of R-loop formation. Changes in the expression of apoptotic markers (e.g. BCL2, BCL2L2, BAX, and TNFRSF10B) would also be expected in infected cells with high levels of replication. Direct interaction between RBPs and viral genomes must also be demonstrated, possibly by immunoprecipitation or copackaging in viral capsids. It should also be possible to evaluate the possibility that inhibitors of viral replication, such as remdesivir (and any other nucleoside analogs), can reduce DNA damage, R-loop formation, and apoptosis of infected cells.\n\nSARS-CoV-2 efficiently infects multiple species of mammals55, and possesses an RNA polymerase with proofreading capability, which enables it to faithfully and accurately replicate and transcribe its genome. In this study, we suggest that effects of SARS-CoV-2 infection are mild in most individuals because most of us mount robust immune responses and eventually clear the virus. The mechanism that we propose (Figure 1), which may be a contributing factor of a variety of different RNA viruses, has the potential to overwhelm that response through jackpot replication coupled to apoptotic events caused by loss of chromosome integrity stemming from depletion of essential RBPs. This results in high multiplicities of infection of cells in the most vulnerable cells. This could cause a rapid onset of loss of viable pneumocytes, and compromising oxygen transport, to a point where it is insufficient to maintain blood pO2 levels to support organ functions. Systemic inhibition of viral replication and transcription of viral proteins will be essential to prevent or mitigate this pathological mechanism.\n\nOther coronaviruses such as MERS and SARS have been shown to induce apoptosis56. The polyphenol Resveratrol has been shown to downregulate apoptosis in vitro56,57, possibly by overexpressing sirtuins (a family of signalling proteins). However, this is ultimately not a practical solution to infection, as the drug will only delay an eventual high multiplicity infection event. In order to inhibit the viral mechanism proposed in this study, a drug must inhibit the viral machinery that sequesters spliceosomal components, leading to R-loops and DNA damage. This may explain, in part, why remdesivir (Gilead) improves the recovery of patients with severe respiratory symptoms. The drug, which was originally developed for treatment of Ebola virus by inhibiting its RNA dependent RNA polymerase, also inhibits viral replication of SARS-CoV-2. Other potential therapies include those targeting expression of genes encoded by the viral genome, which use a common 5’ leader sequence of all transcripts. The promoter sequence for these genes binds to the host encoded hnRNP A1, which regulates transcription of beta coronaviral genes (of which SARS is a member of that family). While hnRNP A1 could be a potential drug target for therapy (there are small molecules that have been shown to inhibit hnRNP A1 RNA splicing activity58), there would be concerns that this may cause inadvertent side effects due to its impact on normal mRNA splicing.\n\nRegardless of whether apoptosis releases large quantities of mature infectious virus, the proposed mechanism will still likely impact pneumocyte function. Should high multiplicity of infection arise from apoptotic release of infectious particles from jackpot viral replication and RBP depletion is expected to severely damage both the original cell and neighboring pneumocytes. The severe symptoms might be the result of rapid lysis of cells responsible for oxygen transport, rather than a cytokine storm. Autopsies of infected individuals from Wuhan China have shown evidence of inflammation, but not necessarily macrophage invasion and pulmonary edema59. Furthermore, apoptosis has been demonstrated in lung epithelial cells in Macaques infected with Influenza virus60. This could explain why physicians and other health professionals in repeated contact with multiple infected patients do not seem to have time to develop immunity to the virus, regardless of their age. Type II pneumocytes which produce surfactant, required at high levels in newborns, decrease with age61 and are particularly diminished in individuals with respiratory disease like COPD (Chronic obstructive pulmonary disease) and ARDS (Acute respiratory distress syndrome). If the multiplicity of infection (MOI) of virus damages this population of cells, then individuals with fewer cells might be more susceptible to exhibiting insufficient pulmonary function due to the high MOI released by the mechanism proposed. These patients would be at greater risk for severe complications requiring assisted ventilation. It is also possible that the deficiency of functional pneumocytes in such individuals cannot be compensated for by extracorporeal membrane oxygenation to rescue multiple organ failure.\n\nHumans have the highest number of type II cells at birth, because the first breath requires significant levels of surfactant to expand lung volume. Synthetic surfactant is an essential treatment for premature birth, since type II pneumocytes mature late in gestation. Age-related loss of these cells has been measured and the mechanism leading to it was described61. Loss of functional pneumocytes is particularly evident in individuals with ARDS, who exhibit significant lung fibrosis, which is also seen in patients with SARS-CoV-2 infections. Older individuals (or those with pre-existing respiratory conditions) are more susceptible to the loss of the remaining cells by apoptosis or autophagia. Decreased pneumocyte counts affect O2 transport efficiency, which lowers pO2 in the blood, tissues and organs. The proposed mechanism implies that jackpot viral replication events, regardless of age of the infected individual, enhance viral release through apoptosis and infection of adjacent cells. Jackpot replication events are more likely in coronaviruses like SARS-CoV-2, as they are capable of repressing the innate immune response, i.e. induction of interferon response to viral double stranded RNA (unlike Influenza)62–64. Repression of innate immunity enables the virus to replicate unabated, which would be expected to delay recognition of these cells by regulatory T cells and killing by macrophages.\n\nThe immune system appears to be a witness, rather than a direct participant in the process of killing infected pneumocytes in many SARS-CoV-2 patients. Approximately a third of infected patients do not raise antibodies after exposure to SARS-CoV-265, and lack proinflammatory cytokines66. Indeed, other coronaviruses have been shown to counter the innate immune response36. One plausible explanation for this is that coronaviruses can evade the host immune response62,63, specifically response involving antiviral type I and type III Interferon proteins (IFNs)64. While many viruses can suppress immune response, IFN response is significantly delayed in coronavirus-infected host cells. There is very little IFN detected in the early stages of MERS-CoV infection67,68. In MERS-CoV, innate immunity is suppressed by NS4B. This raises questions about how, or even whether, extrinsic (IFN receptor-mediated) apoptotic response occurs. However, while the suppression of interferon response is clearly a necessity for SARS-CoV-2 progression, it is not sufficient to explain the severity of the molecular pathogenesis since unrestricted replication alone cannot account for the release of high titer of mature virus that would be necessary for widespread, rapid infection. We therefore suggest that the mechanism proposed could explain how SARS-CoV-2 causes severe lung damage without requiring a hyperinflammatory reaction resulting from a cytokine storm initiated by activating an innate immunity response.\n\nViral infections significantly alter the transcriptional profiles of host genes in infected cells. Recent studies of Zika virus (an RNA virus) have revealed that infection not only impacts transcription, but affects alternative mRNA splicing as well69. Both RNA and DNA viral infections encode factors that directly70 or indirectly69 alter host RNA processing, resembling alternative mRNA isoforms. We suggest that the mRNA splicing changes observed subsequent to infection of an RNA virus could be a consequence of replicated viral genome binding to RBPs, thus changing the nuclear stoichiometry of splicing proteins (such as SRSF1). This would effectively reduce the concentration of available splicing factors, which could be responsible for the observed alternative splicing events of other splicing factors (such as SRSF2 and SRSF3) reported by Bonenfant et al.69. Thus, the mechanism proposed in this study may not only impact genome stability by the introduction of R-loops, but may simultaneously alter the global alternative splicing landscape in infected host cells.\n\nRNA-based vaccines based upon synthetic SARS-CoV-2 transcripts containing modified nucleosides that have been dephosphorylated to escape innate immunity are being tested71. These candidates exploit host protein synthesis machinery to transiently express viral antigens that activate B and T-cell immunity. However, these synthetic RNAs would also be available for RBP binding. A transcript encoding the SARS-CoV-2 spike glycoprotein ‘S’ gene, for example, would contain 7 strong RNPS1 and between 6 to 8 strong SRSF1 binding sites (depending on SRSF1 model). If the levels of expression produced from these transcription templates cannot be carefully controlled, excess production of these RNAs could potentially elicit undesirable side effects through sequestration of critical host RNA binding proteins required to inhibit R-loop formation.\n\nLocalization of viral replication to the cytoplasm does not obviate the fact that there is still a competition between the host and viral genomes for these RNA binding proteins. While the binding site stoichiometry calculations are unchanged, compartmentalization of the viral and host genomes does have implications for preventing R-loops during host transcription. Since coronavirus replicates in the cytoplasm, binding of newly synthesized RBPs occurs there. This makes less protein available to be imported into the nucleus for binding to nascent transcripts to prevent R-loops from forming. The viral genome may have an advantage in this competition for binding to RBPs relative to nuclear transcripts, due to the proximity of the viral genome to nascent RBPs in the cytoplasm, which may limit transport and impede their import into the nucleus. RBPs are often highly expressed, including SRSF1 and RNPS1, and are abundant in the lung (where SARS-CoV-2 infection is most prominent). Thus, the cytoplasmic concentration of viral genome necessary to prevent the localization of RBPs into the nucleus is likely to vary between different tissues.\n\nThe proposed mechanism of RNA virally-induced apoptosis is supported by extensive bioinformatic analyses indicating that strong RNA binding sites of host RBPs are common in RNA viral genomes, and that the frequencies of such binding sites are relatively consistent between divergent strains in both Influenza A and SARS-CoV-2. Future efforts should elucidate details of the mechanism with functional analysis of infected cells, including demonstration of increased R-loop formation, induction of relevant apoptotic or DNA repair responses, and direct interaction between viral genomes and host RBPs. This would justify further investigations into binding of specific RBPs to viral sequences in infected patients. The potentially prognostic significance of such data could be useful in differentiating among drug therapies that target RNA viral genome replication and/or expression.\n\n\nMethods\n\nThe IWMs for the RBPs investigated in this study (SRSF1, RNPS1 and hnRNP A1) were either obtained for previously published analyses or derived in this study. The hnRNP A1 IWM used in this study was previously derived in Peterlongo et al. (using PoWeMaGen software [v1]22) using an hnRNP A1 CLIP-seq dataset72. The functionality provided by PoWeMaGen is also available in Delila software, which is open source. IWMs can also be derived with the ‘Ri’ program, and RBP binding sites can be localized with the ‘Scan’ program of the Delila package. Individual binding site strengths (Ri values) of these IWMs can also be determined using the ‘Scan’ program.\n\nA previously described IWM for SRSF121 was based on only 28 manually curated and validated and aligned binding sites. To update this IWM, we derived new SRSF1 models from high-throughput eCLIP datasets containing thousands of validated binding sites of 150 different RBPs33. Narrow peak files from two separate SRSF1 eCLIP replicates (ENCFF179SCM and ENCFF184TBM) were retrieved from the ENCODE Data Coordination Center (ID: ENCSR456FVU)33. The new SRSF1 IWMs were generated using Maskminent v1.0.2 (24; https://doi.org/10.5281/zenodo.49234). Both PoWeMaGen and Maskminent utilize the Bipad algorithm to align binding sites73. Similarly, RNPS1 was derived from publicly available iCLIP data (30; E-MTAB-4215). However, this iCLIP dataset was only available in FASTQ file format, which required further processing to identify CLIPseq peaks. Thus, the available RNPS1 iCLIP data was first aligned to the human genome (GRCh37) with TopHat v2.1.1, and then converted to peaks using Piranha v.1.2.1 (a CLIP- and RIP-seq peak caller) under default settings.\n\nIWMs for SRSF1 and RNPS1 were derived from eCLIP and iCLIP-seq datasets (respectively) using Maskminent under varying model length conditions (6-10nt long; 1000 Monte Carlo cycles). As experimental noise has been found to contribute to non-specific IWMs24, we limited model derivation to only the to the 5,000 or 50,000 iCLIP peaks with the highest signal value (SRSF1) or the lowest p-values (RNPS1; computed by Piranha). In practice, the derived models remained similar regardless of the size of peak subset used. As many intervals from the SRSF1 and RNPS1 datasets were short (<20nt), peak lengths were extended on either direction by the sequence length (e.g. a 10nt interval becomes 30nt long). We found that both RNPS1 and SRSF1 models derived at lengths of 6nt to be most informative with similar Ri densities, although they differed slightly (Table 1). Both the RNPS1 model and the SRSF1 model derived from the second replicate (SRSF1 “Replicate 2”) selected was generated from 5000 CLIP-seq peaks, while the SRSF1 “Replicate 1” model was derived from 50,000 peaks.\n\nThe derived RNPS1 and SRSF1 models had visually similar RNA binding motifs. To evaluate the similarity between them, the RNPS1 and SRSF1 IWMs were compared using the STAMP web server34, which performs a pairwise alignment between each motif (ungapped Smith-Waterman alignment method) and compared using a Pearson correlation coefficient distance metric.\n\nThe human reference genome (GRCh37; Genbank Acc. GCA_000001405.1; downloaded from UCSC [https://hgdownload.soe.ucsc.edu/goldenPath/hg19/bigZips/]) and viral genomes (Dengue virus 3 [GenBank accession: NC_001475.2]; human immunodeficiency virus type 1 [HIV-1] HXB2 [Genbank: K03455.1] and subtype C [Genbank: U46016.1]; Influenza A H3N2 strains [Ontario/104-25/2012; Genbank (segments 1-8): KJ413878.1, KJ413896.1, KF840477.1, KJ413897.1, KJ413880.1, KJ413864.1, KJ413915.1, KJ413925.1] and [Shanghai/C84/2009 Genbank (segments 1-8): JX286598.1, JX286597.1, JX286596.1, JX308801.1, JX286594.1, JX286593.1, JX286592.1, JX286595.1]; and SARS-CoV-2 [Genbank: NC_045512.2]) were scanned with each IWM (SRSF1 [two separate models], RNPS1 and hnRNP A1). Human genome scans were then filtered so that only those predicted binding sites found in transcribed regions (using the Ensemble Genes database [release 99]) would be considered. Only sites exceeding Rsequence, the average information content of the binding site model, were retained in subsequent analyses as these consist of mean binding affinity or higher and are likely to more effectively compete for binding to these proteins24,25. The Rsequence values of each model are: 6.7 bits (SRSF1 “Replicate 1” model), 6.4 bits (SRSF1 “Replicate 2” model), 7.8 bits (RNPS1 model), and 4.6 bits (hnRNP A1 model).\n\nBesides those previously indicated, viral genomes of multiple other SARS-CoV-2 and Influenza A (H3N2) strains were scanned using the IWMs for SRSF1, RNPS1 and hnRNP A1 to evaluate whether divergent strains of these viruses carry significantly different strong binding sites counts. NextStrain (which provides real-time tracking of the SARS-CoV-2 and Influenza A) was utilized to choose divergent strains of either virus by selecting strains from separate clades, i.e. different monophyletic groups. The viral genome sequences of selected SARS-CoV-2 strains (Genbank accessions: MT007544.1 [Australia], MT066176.1 [Taiwan], MT121215.1 [China], MT163718.1 [USA], MT188339.1 [USA],MT198652.1 [Spain], and MT198653.1 [Spain]) and Influenza A H3N2 (GISAID accessions: EPI1676017-EPI1676024 [Denmark], EPI1635542-EPI1635549 [England], EPI1594883-EPI1594890 [Singapore], EPI1614613-EPI1614620 [Sydney]) were downloaded from the GISAID database. Each of these genome sequences were evaluated for strong SRSF1, RNPS1 and hnRNP A1 binding sites. All binding sites (with Ri ≥Rsequence) are provided in extended data38, Section 1 – Tables 3 and 4.\n\nPublicly-available expression datasets were downloaded from the Gene Expression Omnibus for A549 cell lines (GSE141171; RNAseq) and primary type II pneumocytes (GSE86618; scRNAseq). Normal expression for each cell type was computed by taking the average of all control samples from each dataset (N=3 control samples in GSE141171; N=215 control samples in GSE86618). We then use this information to estimate the total number of binding sites present in a single pneumocyte or A549 cell. First, the program “ScanDataSummaryProgram.pl” (available within underlying data38 Section 6) was used to compute the total number of binding sites (≥Rsequence) in each cell type for each expressed gene (TPM >0; underlying data37 Section 1 - Table 5). The overall expression of each gene was then normalized using the program “TotalBindingSitePerCellCalculator.pl” (underlying data38 Section 6), which divides expression by the sum of all TPM values in the cell, multiplied by the estimated number of mature RNAs in a cell at any given timepoint (80,000 RNAs per lymphoblastoid cell51). It then multiplies this normalized gene expression value with its binding site total to determine the overall contribution of binding sites from that gene in a single cell. The sum of this value across all expressed genes gives the total number of RNA binding sites expected to be available in a cell at any given time (Figure 6).\n\nInformation dense clustering has previously been applied to the human genome to identify clusters of organized TFBSs25,39. The clustering software (v1; described in reference 25; software provided in a Zenodo archive - https://doi.org/10.5281/zenodo.1707423) was used in this study to identify clusters of low-affinity (Ri > 0 bits), moderate-affinity (≥12Rsequence) and high-affinity (≥Rsequence) RBP sites in both the viral genomes investigated in this study, and across the entire human transcriptome. To be considered a cluster, each set of component sites was required to occur ≤25nt from one other, and the total information of all sites within the cluster equalled or exceeded ≥50 bits. In its original design, the clustering algorithm considered binding sites on both strands in forming clusters. To maintain strand specificity, we separated input by strand. Due to the high memory demands of the clustering algorithm, transcriptome scan input was separated into segments of ~200,000 sites per run, which was then subsequently combined. To avoid the inadvertent separation of a binding site cluster, input was split only when two sequential binding sites were >1000nt apart.\n\nAll binding sites and information-dense clusters identified in the human genome were intersected with DRIP-seq and DRIPc-seq intervals, which indicate where there is evidence of R-loop formation in the human genome (performed by “ClusterToDRIPseqAnalysisProgram.pl”; underlying data38 Section 6). The DRIP-seq dataset (GSE68845; IMR90 cells) is not strand specific, thus binding sites and clusters from either strand are considered when intersected against these intervals. DRIPc-seq data (GSE70189; NTERA2 cells), however, is strand specific which has been taken into account (e.g. positive strand clusters found in positive strand DRIPc-seq intervals reported). We then computed the gene density of sites and clusters that are found within these intervals (underlying data38 Section 1 - Table 5) using the script “ClusterToDRIPseqAnalysisProgram.GeneDensityFinder.pl” (underlying data38 Section 6) to determine if there is a correlation between the presence of binding sites and R-loop formation.\n\nLollipop plots which indicate the location of information-dense clusters for all viral genomes described in this study and for all genes in the human transcriptome (with ≥1 cluster) were generated in R (version 3.6.3) using the Bioconductor package “trackViewer” (v.1.20.374). The lollipop plots presenting human genes contain intron and exon boundary information which was generated using the RefSeq database (release 60). Multiple lollipop plots were generated for multi-segmented viral genomes (one image per segment). The height of each “lollipop” corresponds to the information density of a cluster, and its location in the genome is indicated (GRCh37) along with the number of sites which comprise the cluster.\n\nHistograms which illustrate the distribution of binding site Ri values and the frequency of the distance between RBPs (“intersite distances”) were generated using the R package ‘ggplot2’ (v3.1.175). Intersite distance frequency was determined by first grouping all RBP by gene, followed by determining the distance between each site in sequential order. Distance thresholds of 500nt or 1000nt were assigned for all intersite distance histograms. Rare instances of distances greater than these thresholds were excluded from the histogram, as their inclusion led to plots too wide to be informative.\n\nGene expressions for individuals with the diseases above were collected from Gene Expression Omnibus (GEO), which consisted of 5 Influenza studies (GSE29385, GSE82050, GSE50628, GSE61821, GSE27131), 4 Dengue studies (GSE97861, GSE97862, GSE51808, GSE58278) and 2 studies involving Aplastic Anemia patients (GSE16334, GSE33812). We also collected expression data from two studies with radiation-exposed samples (GSE6874 and GSE10640). The best performing human signatures (assessed by traditional validation; described in Table 7 [underlying data38 Section 1]) from Zhao et al.47 were then used to test the gene expression datasets in order to determine if these models would misclassify infected patients as irradiated (with and without control patients). Models were tested using the MatLab script used to perform “traditional validation” in the Zhao et al. study (“regularValidation_multiclassSVM.m”, https://zenodo.org/record/1170572), which first normalizes gene expression values by quantile normalization before applying the radiation model to the infected patient data to predict outcome. The script then compares prediction of radiation exposure to the clinical data provided. MatLab scripts are compatible with GNU Octave.\n\nTo better understand why the radiation models are predicting certain Influenza- and Dengue-infected patients as irradiated, violin plots were generated using GraphPad Prism v8 to visually illustrate differences in gene expression between infected individuals correctly classified and those misclassified by each radiation model (Figure 5). When inspecting violin plots of the 32 genes which make up the 4 radiation models tested, 10 genes were identified to have contributed towards false positives predictions as they shared a similar pattern of expression in those that were radiated in two gene expression datasets of irradiated individuals (GSE6874 and GSE10640). The 10 genes are: DDB2, PCNA, GTF3A, PRKCH, CDKN1A, GADD45A, BCL2, MOAP1, TRIM22 and TALDO1. Mann Whitney tests were used to compare the expression of these genes in false negative and true positive patients. Four genes (DDB2, PCNA, GTF3A and PRKCH) were consistently found significant in most of the studies tested.\n\nThe dissociation constant of SRSF1 RRM2 domain bound to the RNA sequence 5′-UGAAGGAC-3′ was experimentally determined to be 0.8 μM53. This information allowed for the derivation of a theoretical Scatchard plot for SRSF1 binding by varying the relative proportions of viral to host binding sites bound (where viral binding sites are considered inhibitors, and host binding sites as substrate). We can compute the theoretical number of viral genomes necessary to reach these relative proportions according to:\n\n\n\nWhere Kd is the SRSF1 dissociation constant, n is the number of sites an SRSF1 protein can bind at one time (n=1), [L] is the concentration of free SRSF1, and v is the amount of SRSF1 bound to the viral genome relative to host. With this derivation, we assume there is no free RNA binding protein. These proportions were converted to numbers of viral genomes per infected host cell (determined using the above formula in an MS- Excel spreadsheet), adjusted for the computed number of viral genomes per cell by the number of SRSF1 binding sites in a single viral genome (described earlier). We also computed the number of viral genomes necessary to reach these proportions by taking A549 or pneumocyte host cell binding site expression (computed previously) into account. We then used the known processivity rate of 3.7 nucleotides/sec for VSV RNA dependent RNA polymerase54 to estimate the doubling time required.\n\n\nStatistical analysis\n\nThe average distances between adjacent binding sites of SRSF1, RNPS1 and hnRNP A1 were determined within both expressed human genes and RNA viral genomes (Dengue, HIV-1 strains B and C, Influenza A and SARS-CoV-2). A program script “calculateIntersiteDistance.pl” (underlying data38 Section 6) takes a set of binding site coordinates and their associated genes as input and determines the pairwise distances between all consecutive binding sites in the same gene. Subsequently, “removeOutliersHigherThanN.pl” is used to discard extreme outlier distances exceeding a specified threshold (thresholds of 500nt and 1000nt were evaluated). Finally, “getStatisticsOnCol.pl” evaluates a given set of intersite distances and computes the count, geometric mean, median, arithmetic mean and their standard deviation. The program was used to evaluate intersite distances at multiple Ri thresholds (low- [Ri > 0 bits], moderate- [≥12Rsequence] and high-affinity [≥Rsequence] binding sites). We also examined binding sites which intersect DRIPc-seq intervals in the human genome using this procedure. Output from this analysis are provided as histograms in extended data38 Section 5, as described earlier.\n\n\nData availability\n\nA data repository titled “Characteristics of human and viral RNA binding sites and site clusters recognized by SRSF1 and RNPS1” has been deposited as a Zenodo archive (DOI: 10.5281/zenodo.373708938). The archive contains the following underlying and extended data, organized across 6 sections. Section 1 primarily consists of extended data, and Sections 2-6 contains the underlying data presented in the paper.\n\nZenodo: Characteristics of human and viral RNA binding sites and site clusters recognized by SRSF1 and RNPS1. http://doi.org/10.5281/zenodo.373708938\n\nThis project contains the following extended data:\n\nSection 1 – The nine additional tables described in this study (“Section 1 - Tables 1–9”), which provide SRSF1, RNPS1 and hnRNP A1 binding site and information-dense cluster counts across various RNA viral genomes [including multiple SARS-CoV-2 and Influenza strains] and the human transcriptome, the estimated SARS-CoV-2 doubling time necessary for viral genome SRSF1 binding site availability to exceed sites within the host transcriptome, and an analysis of Influenza, Dengue, and aplastic anemia patients misdiagnosed as irradiated by established radiation gene signatures.\n\nZenodo: Characteristics of human and viral RNA binding sites and site clusters recognized by SRSF1 and RNPS1. http://doi.org/10.5281/zenodo.373708938\n\nSection 2. All SRSF1, RNPS1 and hnRNP A1 binding site genome browser tracks for human and all viral genomes analyzed in this study (GRCh37).\n\nSection 3. The full set of lollipop plots (indicating the location of SRSF1, RNPS1 and hnRNP A1 information-dense clusters) in all human genes and in each of the viral genomes analyzed.\n\nSection 4. The Ri(b,l) matrices or IWMs for all RBPs analyzed (SRSF1, hnRNP A1 and RNPS1).\n\nSection 5. The full set of histograms which display the distribution of Ri strength and intersite distance between the binding sites for each RBP [across all transcribed regions or within known DRIPc-seq intervals.\n\nSection 6. A set of 7 Perl scripts created specifically for this study, with instructions for their use: A) “ClusterToDRIPseqAnalysisProgram.pl” – reports which information-dense clusters are located within DRIPc- and/or DRIP-seq intervals (individually and by gene); B) “ClusterToDRIPseqAnalysisProgram.GeneDensityFinder.pl” – uses the output from script “A” to determine the number and the density of information-dense clusters within a gene (total clusters within the gene and those within DRIPc-seq intervals); C) “calculateIntersiteDistance.pl” – determines the distance between all binding sites in the same gene from a list of genomic coordinates; D) “removeOutliersHigherThanN.pl” – discards intersite distances computed by script “C” that are greater than a specified threshold; E) “getStatisticsOnCol.pl” – calculates the count, geometric mean, median, arithmetic mean, and standard deviation of values from script “D”; F) “ScanDataSummaryProgram.pl” – determines the number of binding sites (above a specified Ri threshold) found within known genes (the program also reports the total expression of those genes using external A549 and pneumocyte expression datasets) from binding site coordinate data; G) “TotalBindingSitePerCellCalculator.pl” – estimates the number of binding sites expressed in a single A549 or pneumocyte cell at any given time.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "References\n\nCarrasco-Hernandez R, Jácome R, López Vidal Y, et al.: Are RNA Viruses Candidate Agents for the Next Global Pandemic? A Review. ILAR J. 2017; 58(3): 343–58. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu Y, Yan LM, Wan L, et al.: Viral dynamics in mild and severe cases of COVID-19. Lancet Infect Dis. 2020; 20(6): 656–657. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZheng S, Fan J, Yu F, et al.: Viral load dynamics and disease severity in patients infected with SARS-CoV-2 in Zhejiang province, China, January-March 2020: retrospective cohort study. BMJ. 2020; 369: m1443. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHolshue ML, DeBolt C, Lindquist S, et al.: First Case of 2019 Novel Coronavirus in the United States. N Engl J Med. 2020; 382(10): 929–36. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFujikura D, Miyazaki T: Programmed Cell Death in the Pathogenesis of Influenza. Int J Mol Sci. 2018; 19(7): 2065. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi X, Manley JL: Inactivation of the SR protein splicing factor ASF/SF2 results in genomic instability. Cell. 2005; 122(3): 365–78. PubMed Abstract | Publisher Full Text\n\nAguilera A, García-Muse T: R loops: from transcription byproducts to threats to genome stability. Mol Cell. 2012; 46(2): 115–24. PubMed Abstract | Publisher Full Text\n\nManley JL, Tacke R: SR proteins and splicing control. Genes Dev. 1996; 10(13): 1569–79. PubMed Abstract | Publisher Full Text\n\nKrainer AR, Conway GC, Kozak D: The essential pre-mRNA splicing factor SF2 influences 5’ splice site selection by activating proximal sites. Cell. 1990; 62(1): 35–42. PubMed Abstract | Publisher Full Text\n\nBlack DL: Mechanisms of alternative pre-messenger RNA splicing. Annu Rev Biochem. 2003; 72: 291–336. PubMed Abstract | Publisher Full Text\n\nLi X, Niu T, Manley JL: The RNA binding protein RNPS1 alleviates ASF/SF2 depletion-induced genomic instability. RNA. 2007; 13(12): 2108–15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEwing RM, Chu P, Elisma F, et al.: Large-scale mapping of human protein-protein interactions by mass spectrometry. Mol Syst Biol. 2007; 3: 89. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDeka B, Singh KK: Multifaceted Regulation of Gene Expression by the Apoptosis- and Splicing-Associated Protein Complex and Its Components. Int J Biol Sci. 2017; 13(5): 545–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGómez-González B, García-Rubio M, Bermejo R, et al.: Genome-wide function of THO/TREX in active genes prevents R-loop-dependent replication obstacles. EMBO J. 2011; 30(15): 3106–19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSalas-Armenteros I, Barroso SI, Rondón AG, et al.: Depletion of the MFAP1/SPP381 Splicing Factor Causes R-Loop-Independent Genome Instability. Cell Rep. 2019; 28(6): 1551–1563.e7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStirling PC, Chan YA, Minaker SW, et al.: R-loop-mediated genome instability in mRNA cleavage and polyadenylation mutants. Genes Dev. 2012; 26(2): 163–75. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCaminsky N, Mucaki EJ, Rogan PK: Interpretation of mRNA splicing mutations in genetic disease: review of the literature and guidelines for information-theoretical analysis [version 1; peer review: 2 approved]. F1000Res. 2014; 3: 282. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRogan PK, Schneider TD: Using information content and base frequencies to distinguish mutations from genetic polymorphisms in splice junction recognition sites. Hum Mutat. 1995; 6(1): 74–6. PubMed Abstract | Publisher Full Text\n\nRogan PK, Faux BM, Schneider TD: Information analysis of human splice site mutations. Hum Mutat. 1998; 12(3): 153–71. PubMed Abstract | Publisher Full Text\n\nMucaki EJ, Ainsworth P, Rogan PK: Comprehensive prediction of mRNA splicing effects of BRCA1 and BRCA2 variants. Hum Mutat. 2011; 32(7): 735–42. PubMed Abstract | Publisher Full Text\n\nMucaki EJ, Shirley BC, Rogan PK: Prediction of mutant mRNA splice isoforms by information theory-based exon definition. Hum Mutat. 2013; 34(4): 557–65. PubMed Abstract | Publisher Full Text\n\nPeterlongo P, Catucci I, Colombo M, et al.: FANCM c.5791C>T nonsense mutation (rs144567652) induces exon skipping, affects DNA repair activity and is a familial breast cancer risk factor. Hum Mol Genet. 2015; 24(18): 5345–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMucaki EJ, Caminsky NG, Perri AM, et al.: A unified analytic framework for prioritization of non-coding variants of uncertain significance in heritable breast and ovarian cancer. BMC Med Genomics. 2016; 9: 19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLu R, Mucaki EJ, Rogan PK: Discovery and validation of information theory-based transcription factor and cofactor binding site motifs. Nucleic Acids Res. 2017; 45(5): e27. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLu R, Rogan PK: Transcription factor binding site clusters identify target genes with similar tissue-wide expression and buffer against mutations [version 2; peer review: 2 approved]. F1000Res. 2018; 7: 1933. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchneider TD: Information content of individual genetic sequences. J Theor Biol. 1997; 189(4): 427–41. PubMed Abstract | Publisher Full Text\n\nRogan PK, Klesc R, Mucaki EJ, et al.: Proposed mechanism of SARS-CoV-2 severe infection. Figshare. 2020. Publisher Full Text\n\nDubois J, Terrier O, Rosa-Calatrava M: Influenza viruses and mRNA splicing: doing more with less. mBio. 2014; 5(3): e00070–00014. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi X, Manley JL: Cotranscriptional processes and their influence on genome stability. Genes Dev. 2006; 20(14): 1838–47. PubMed Abstract | Publisher Full Text\n\nHauer C, Sieber J, Schwarzl T, et al.: Exon Junction Complexes Show a Distributional Bias toward Alternatively Spliced mRNAs and against mRNAs Coding for Ribosomal Proteins. Cell Rep. 2016; 16(6): 1588–603. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDe Maio FA, Risso G, Iglesias NG, et al.: The Dengue Virus NS5 Protein Intrudes in the Cellular Spliceosome and Modulates Splicing. PLoS Pathog. 2016; 12(8): e1005841. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSamji T: Influenza A: understanding the viral life cycle. Yale J Biol Med. 2009; 82(4): 153–9. PubMed Abstract | Free Full Text\n\nNostrand ELV, Freese P, Pratt GA, et al.: A Large-Scale Binding and Functional Map of Human RNA Binding Proteins. bioRxiv. 2018; 179648. Publisher Full Text\n\nMahony S, Benos PV: STAMP: a web tool for exploring DNA-binding motif similarities. Nucleic Acids Res. 2007; 35(Web Server issue): W253–258. PubMed Abstract | Publisher Full Text | Free Full Text\n\nde Wilde AH, Snijder EJ, Kikkert M, et al.: Host Factors in Coronavirus Replication. Roles Host Gene Non-Coding RNA Expr Virus Infect. 2017; 419: 142. Publisher Full Text\n\nPerlman S, Netland J: Coronaviruses post-SARS: update on replication and pathogenesis. Nat Rev Microbiol. 2009; 7(6): 439–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPietrokovski S: Searching databases of conserved sequence regions by aligning protein multiple-alignments. Nucleic Acids Res. 1996; 24(19): 3836–45. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRogan PK, Mucaki EJ, Shirley BC: Characteristics of human and viral RNA binding sites and site clusters recognized by SRSF1 and RNPS1. 2020. Zenodo.http://www.doi.org/10.5281/zenodo.3956522\n\nDinakarpandian D, Raheja V, Mehta S, et al.: Tandem machine learning for the identification of genes regulated by transcription factors. BMC Bioinformatics. 2005; 6: 204. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPachetti M, Marini B, Benedetti F, et al.: Emerging SARS-CoV-2 mutation hot spots include a novel RNA-dependent-RNA polymerase variant. J Transl Med. 2020; 18(1): 179. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTang X, Wu C, Li X, et al.: On the origin and continuing evolution of SARS-CoV-2. Natl Sci Rev. 2020; [cited 2020 May 22]. Publisher Full Text\n\nKorber B, Fischer WM, Gnanakaran S, et al.: Spike mutation pipeline reveals the emergence of a more transmissible form of SARS-CoV-2. bioRxiv. 2020; 2020.04.29.069054. Publisher Full Text\n\nHadfield J, Megill C, Bell SM, et al.: Nextstrain: real-time tracking of pathogen evolution. Bioinformatics. 2018; 34(23): 4121–3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSu YC, Anderson DE, Young BE, et al.: Discovery of a 382-nt deletion during the early evolution of SARS-CoV-2. bioRxiv. 2020; 2020.03.11.987222. Publisher Full Text\n\nSanz LA, Chédin F: High-resolution, strand-specific R-loop mapping via S9.6-based DNA-RNA immunoprecipitation and high-throughput sequencing. Nat Protoc. 2019; 14(6): 1734–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRogan PK: Multigene signatures of responses to chemotherapy derived by biochemically-inspired machine learning. Mol Genet Metab. 2019; 128(1–2): 45–52. PubMed Abstract | Publisher Full Text\n\nZhao JZL, Mucaki EJ, Rogan PK: Predicting ionizing radiation exposure using biochemically-inspired genomic machine learning [version 2; peer review: 3 approved]. F1000Res. 2018; 7: 233. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMerekoulias G, Alexopoulos EC, Belezos T, et al.: Lymphocyte to monocyte ratio as a screening tool for influenza. PLoS Curr. 2010; 2: RRN1154. PubMed Abstract | Publisher Full Text | Free Full Text\n\nScrima A, Konícková R, Czyzewski BK, et al.: Structural basis of UV DNA-damage recognition by the DDB1-DDB2 complex. Cell. 2008; 135(7): 1213–23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang H, Zhai L, Xu J, et al.: Histone H3 and H4 Ubiquitylation by the CUL4-DDB-ROC1 Ubiquitin Ligase Facilitates Cellular Response to DNA Damage. Mol Cell. 2006; 22(3): 383–94. PubMed Abstract | Publisher Full Text\n\nMarinov GK, Williams BA, McCue K, et al.: From single-cell to cell-pool transcriptomes: stochasticity in gene expression and RNA splicing. Genome Res. 2014; 24(3): 496–510. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXia C, Fan J, Emanuel G, et al.: Spatial transcriptome profiling by MERFISH reveals subcellular RNA compartmentalization and cell cycle-dependent gene expression. Proc Natl Acad Sci U S A. 2019; 116(39): 19490–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCléry A, Sinha R, Anczuków O, et al.: Isolated pseudo-RNA-recognition motifs of SR proteins can regulate splicing using a noncanonical mode of RNA recognition. Proc Natl Acad Sci U S A. 2013; 110(30): E2802–2811. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTimm C, Gupta A, Yin J: Robust kinetics of an RNA virus: Transcription rates are set by genome levels. Biotechnol Bioeng. 2015; 112(8): 1655–62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShi J, Wen Z, Zhong G, et al.: Susceptibility of ferrets, cats, dogs, and other domesticated animals to SARS-coronavirus 2. Science. 2020; 368(6494): 1016–1020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLin SC, Ho CT, Chuo WH, et al.: Effective inhibition of MERS-CoV infection by resveratrol. BMC Infect Dis. 2017; 17(1): 144. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang L, Guo X, Xie W, et al.: Resveratrol exerts an anti-apoptotic effect on human bronchial epithelial cells undergoing cigarette smoke exposure. Mol Med Rep. 2015; 11(3): 1752–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCarabet LA, Leblanc E, Lallous N, et al.: Computer-Aided Discovery of Small Molecules Targeting the RNA Splicing Activity of hnRNP A1 in Castration-Resistant Prostate Cancer. Mol Basel Switz. 2019; 24(4): 763. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhou Y, Fu B, Zheng X, et al.: Aberrant pathogenic GM-CSF+ T cells and inflammatory CD14+CD16+ monocytes in severe pulmonary syndrome patients of a new coronavirus. bioRxiv. 2020; 2020.02.12.945576. Publisher Full Text\n\nShinya K, Gao Y, Cilloniz C, et al.: Integrated Clinical, Pathologic, Virologic, and Transcriptomic Analysis of H5N1 Influenza Virus-Induced Viral Pneumonia in the Rhesus Macaque. J Virol. 2012; 86(11): 6055–66. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLehmann M, Hu Q, Hu Y, et al.: Chronic WNT/β-catenin signaling induces cellular senescence in lung epithelial cells. Cell Signal. 2020; 70: 109588. PubMed Abstract | Publisher Full Text\n\nKonno Y, Kimura I, Uriu K, et al.: SARS-CoV-2 ORF3b is a potent interferon antagonist whose activity is further increased by a naturally occurring elongation variant. bioRxiv. 2020; 2020.05.11.088179. Publisher Full Text\n\nGoldstein SA, Weiss SR: Origins and pathogenesis of Middle East respiratory syndrome-associated coronavirus: recent advances [version 1; peer review: 3 approved]. F1000Res. 2017; 6: 1628. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBlanco-Melo D, Nilsson-Payant BE, Liu WC, et al.: Imbalanced Host Response to SARS-CoV-2 Drives Development of COVID-19. Cell. 2020; 181(5): 1036–1045.e9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWu F, Wang A, Liu M, et al.: Neutralizing antibody responses to SARS-CoV-2 in a COVID-19 recovered patient cohort and their implications. medRxiv. 2020; 2020.03.30.20047365. Publisher Full Text\n\nWilk AJ, Rustagi A, Zhao NQ, et al.: A single-cell atlas of the peripheral immune response to severe COVID-19. medRxiv. 2020; 2020.04.17.20069930. Publisher Full Text\n\nKindler E, Jónsdóttir HR, Muth D, et al.: Efficient replication of the novel human betacoronavirus EMC on primary human epithelium highlights its zoonotic potential. mBio. 2013; 4(1): e00611–00612. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChan RWY, Chan MCW, Agnihothram S, et al.: Tropism of and innate immune responses to the novel human betacoronavirus lineage C virus in human ex vivo respiratory organ cultures. J Virol. 2013; 87(12): 6604–14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBonenfant G, Meng R, Shotwell C, et al.: Asian Zika Virus Isolate Significantly Changes the Transcriptional Profile and Alternative RNA Splicing Events in a Neuroblastoma Cell Line. Viruses. 2020; 12(5): 510. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoudreault S, Armero VES, Scott MS, et al.: The Epstein-Barr virus EBNA1 protein modulates the alternative splicing of cellular genes. Virol J. 2019; 16(1): 29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMandal PK, Rossi DJ: Reprogramming human fibroblasts to pluripotency using modified mRNA. Nat Protoc. 2013; 8(3): 568–82. PubMed Abstract | Publisher Full Text\n\nHuelga SC, Vu AQ, Arnold JD, et al.: Integrative genome-wide analysis reveals cooperative regulation of alternative splicing by hnRNP proteins. Cell Rep. 2012; 1(2): 167–78. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBi C, Rogan PK: Bipartite pattern discovery by entropy minimization-based multiple local alignment. Nucleic Acids Res. 2004; 32(17): 4979–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOu J, Zhu LJ: trackViewer: a Bioconductor package for interactive and integrative visualization of multi-omics data. Nat Methods. 2019; 16(6): 453–4. PubMed Abstract | Publisher Full Text\n\nWickham H: ggplot2: Elegant Graphics for Data Analysis. New York: Springer-Verlag; 2009 [cited 2020 May 19]. (Use R!). Reference Source"
}
|
[
{
"id": "73888",
"date": "13 Nov 2020",
"name": "Maurizio Romano",
"expertise": [
"Reviewer Expertise Splicing",
"Neuroscience."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn the present manuscript, starting from the observation that riboviruses can cause fatal pulmonary in some infected patients, the Authors propose the interesting hypothesis that depletion of host RNA binding proteins from nuclear RNA bound to replicating viral sequences might be part of the mechanism that trigger apoptosis and viral release.\nInformation theory-based analysis was used to test interactions between RBPs and individual sequences in different virus, since expression of viral sequences might sequester RBPs (the study is focused on SRSF1 and RNPS1). It is proposed a correlation RBPs depletion / destabilization of R-loops / chromosomal breakage.\nThe stoichiometry of inhibition of RBPs in host nuclear RNA has been estimated by counting competing binding sites in replicating viral genomes and host RNA.\nIt is concluded that the RNA virally-induced apoptosis could lead to release significant quantities of membrane-associated virions and cause the fatal pulmonary.\nAlthough functional analyses might be helpful to strengthen the validity of the proposed mechanism, all the steps and conclusions of the study are sufficiently clear to support the hypothesis.\nThe Discussion might be shortened by taking out aspects that are not directly related to the proposed theory.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6188",
"date": "06 Jan 2021",
"name": "Peter Rogan",
"role": "Author Response",
"response": "Thank you for reviewing our manuscript. We concur that the Discussion section of the manuscript is lengthy. We have therefore removed the paragraph beginning with ““The immune system appears to be a witness, rather than a direct participant …”, as it is only tangentially related from the mechanism being proposed."
}
]
},
{
"id": "73885",
"date": "18 Nov 2020",
"name": "Mansi Srivastava",
"expertise": [
"Reviewer Expertise Genomics and Systems Biology of RNA regulatory processes."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study provides mechanistic insights into the RNA viral infection that triggers unrepaired sites of chromosomal breakage, causing apoptosis and consequentially, high-titer viral release. The hypothesis suggests that the viral genome binds RNA binding proteins of the host thus, preventing their essential post-transcriptional activities.\nIn the result section that describes the human transcriptome analysis of RNA binding sites, the authors evaluate the frequency of RBP binding in human transcriptomes to relate the relative abundance of these proteins bound to viral RNAs compared to their normal reservoir in host nuclear RNA of infected cells. To support their observation, authors should include a discussion on the impact of other RNA binding proteins that may bind/regulate the same site on the viral genome.\n\nAuthors discuss that the viral genome may have an advantage in the competition for binding to RBPs relative to nuclear transcripts, due to the proximity of the viral genome to nascent RBPs in the cytoplasm, which may limit transport and impede their import into the nucleus. However, the authors should also mention RBPs that shuttle between the nucleus and cytoplasm dynamically and thus account for binding to the nascent transcripts at the basal level.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6189",
"date": "06 Jan 2021",
"name": "Peter Rogan",
"role": "Author Response",
"response": "Thank you for reviewing our manuscript. Newly synthesized RBPs imported into the nucleus and bound to nascent transcripts do not necessarily have an impact on R-loop formation. Our study put a strong focus on SRSF1 and RNPS1, which have been documented to be antagonistic to R-loop formation. We have not investigated other RNA binding proteins that are known to stabilize nascent transcripts, such as the THO complex, PCF11, and the exoribonucleases EXOSC3 and EXOSC10 (Santos-Pereira & Aguilera. Nat Rev Genet. 2015. 16: 583-597). The CLIP data required to analyze RNA for binding sites is not currently available for many of these RBPs. Ribosomal RNA (rRNA) constitutes the most abundant RNA in the cytoplasm (indeed, the cell), and would likely be the most likely to interact with SRSF1 and RNPS1. Ribosomal proteins interacting with rRNA-scaffold would likely represent the most abundant competitor to viral RNAs in infected cells. rRNA interactions have a structural basis (e.g. bulged duplexes, hairpin loops) that explains their affinity for ribosomal proteins. This contrasts with the sequence-specific binding by SRSF1 and RNPS1 and other RBPs containing one or more RRM domains (Ciriello et al. BMC Bioinformatics. 2010; 11(Suppl 1): S41). The method we describe does not detect or quantify the type of structural interactions seen in rRNA and ribosomal proteins. The present approach cannot determine whether viral RNAs could bind to ribosomal proteins."
}
]
},
{
"id": "73887",
"date": "24 Nov 2020",
"name": "Gregory Fonseca",
"expertise": [
"Reviewer Expertise Bioinformatics",
"transcriptomics",
"genomics. Lung disease."
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this study, the authors present evidence that R-loops are associated with RNA binding protein binding sites and this may lead to DNA damage associated apoptosis. The authors define IWMs for RNA binding proteins, SRSF1 and RNPS1 based on previously published, high quality data. They then show the occurrence and quality of these IWMs in viral genomes including the relative stability of these IWMs across evolution. The authors then compared the quality and relative quantity of these IWMs in the human transcriptome. They predict the number of viral RNA particles necessary to squelch RNA binding proteins from the human genome.\nOverall, this is an extremely interesting paper with a very exciting hypothesis. The paper is well written and well organized and makes use of available datasets very well.\nA few notes.\nIt should be mentioned whether IWM discovery was compared to background to understand if IWNs are found above random chance.\n\nIf you randomly curate IWMs from 5000 sites of Rep1 or bin 5000 sites do the results compare to Rep2?\n\nWhat would the predicted likelihood of IWMs changing by chance compare to observed?\n\nIs there a correlation of SRSF and RNPS1 sites in the mRNA and gene expression at basal and during infection?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6191",
"date": "06 Jan 2021",
"name": "Peter Rogan",
"role": "Author Response",
"response": "1. It should be mentioned whether IWM discovery was compared to background to understand if IWNs are found above random chance. Response: We derived IWMs from (mock) control binding studies for SRSF1 and RNPS1. Sequences obtained for 3 negative controls were generated in the same dataset that was used to build IWMs for RNPS1 (E-MTAB-4215; control datasets ‘ERR1201436’, ‘ERR1201437’ and ‘ERR1201438’). These were samples containing a GFP-tag (lacking the RNPS1 fusion protein for pulldown). IWMs were also derived from two SRSF1 control datasets (‘ENCFF241ORF’ and ‘ENCFF773PUP’; positive and negative strands from mock input sample). From these datasets, we derived IWMs of length 6nt from the top 5,000 peaks from each dataset with Maskminent using the same parameters (see Methods). The motifs of the newly derived control IWMs were compared to the true SRSF1 and RNPS1 models using STAMP software (described in Methods). The e-value obtained from STAMP is the number of hits expected against a database of the same size (i.e. 5,000 sequences containing random sequences of the same length). The comparison is based on the log likelihood ratio of the e-values of the IWM motifs derived from the RBP bound sequences relative to the sequences obtained from the mock control, which is a modified LOD score. The LOD scores for the RNPS1 model ranged from 3.8 to 6.1 depending on which control IWM was compared, and for SRSF1, these scores ranged from 4.4 to 8.8. We therefore conclude that the motifs obtained for these protein binding sites are significantly more robust (and different from a random set of sequences of the same size and composition generated from control samples). Please note that this analysis has been incorporated into the manuscript (the fourth paragraph of the Results and in the second and fourth paragraphs of the Methods). 2. If you randomly curate IWMs from 5000 sites of Rep1 or bin 5000 sites do the results compare to Rep2? Response: In this study, the two SRSF1 IWMs utilized were derived from two separate replicates from publicly available eCLIP data. The first model “SRSF1 Replicate 1” was based on the 50,000 largest eCLIP peaks from the ‘ENCFF179SCM’ replicate, while “SRSF1 Replicate 2” was based on the top 5,000 peaks in the ‘ENCFF184TBM’ replicate. Despite being derived from a far smaller dataset, the models were computed to be quite similar and non-random by STAMP analysis (e-score: 7.4e-10). While the SRSF1 “Replicate 1” and “Replicate 2” models were those which were selected to be used in the study, IWMs for SRSF1 were derived utilizing a series of different conditions (i.e. number of peaks, number of Monte Carlo cycles, etc.), however discussion of these additional derived SRSF1 models was not included in the final manuscript. Most commonly, models derived under these varying conditions were similar to that of the final models. On occasion, the method used here has been reported to identify binding motifs of other factors whose binding site sequences are in close proximity with the factor being crosslinked, as well as IWMs with a noise motifs that can resemble repetitive sequences (Lu et al. Nucleic Acids Res. 2017 Mar 17;45(5):e27). We carefully evaluated each IWM before it was utilized in any downstream analyses. For example, while the SRSF1 model derived from 10,000 peaks from the ‘replicate 1’ dataset is highly similar to that of the 50,000 replicate 1 peak model (as well as those models derived from replicate 2 data), the 5,000 replicate 1 peak model contained a slight variation of the primary motif, reporting instead an unexpected “G[G/C]AG” sub-motif. The pairwise IWM e-values from comparison of SRSF1 Replicates 1 and 2 are: “SRSF1 Replicate 2” (self comparison): 3.9e-11; “SRSF1 Replicate 1” (top e-CLIP 50,000 peaks): 7.4e-10; “SRSF1 Replicate 1” (top 10,000 peaks): 3.6e-09; and “SRSF1 Replicate 1” (top 5,000 peaks: 5.0e-03. In general, however, modifying the number of binding sites from which models are derived generally leads to IWMs with highly similar binding motifs. 3. What would the predicted likelihood of IWMs changing by chance compare to observed? Response: See response to Question 1. 4. Is there a correlation of SRSF and RNPS1 sites in the mRNA and gene expression at basal and during infection? Response: The expression of basal SRSF1 and RNPS1 may be significantly altered in an infected cell. Blanco-Melo et al. (Ref. 55) performed differential gene expression analysis of A549 immortalized cell lines, comparing infection with influenza A- or SARS-CoV-2 with controls and ranking genes by fold change p-values adjusted for multiple testing. Both SRSF1 and RNPS1 exhibited statistically significant lowered expression in SARS-CoV-2 infected A549 cell lines (73% [p=3.7e-16] and 47% [p=3.5e-89] of controls, respectively). Their expression was also significant decreased in SARS-CoV-2 infected Calu-3 cell lines (74% for SRSF1 and 75% for RNPS1). Expression of SRSF1 and RNPS1 in A549 cells infected with respiratory syncytial virus was also significantly reduced (73% for both). No significant differences were evident in A549 infected with either SARS-CoV-1, MERS nor influenza A. Expression of SRSF1 and RNPS1 is not only altered by viral infection, but the extent of these changes is related to the specific infectious pathogen. We have added a statement to the manuscript acknowledging that the gene expression datasets utilized in this study were from uninfected cells which may differ in infected cells (new text bolded): “These are derived from the number of SRSF1 and RNPS1 sites expressed in either a single A549 cell or a type II pneumocyte (cells were not infected; note that infection would be expected to alter the expression profile, which could affect expressed binding site estimates).”"
}
]
},
{
"id": "73886",
"date": "26 Nov 2020",
"name": "Ian Eperon",
"expertise": [
"Reviewer Expertise RNA splicing."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe purpose of this research is to establish by computational methods whether the genomes of certain RNA viruses could provide enough binding sites for proteins that bind nascent transcripts to reduce their availability in the cell and thereby trigger R-loop formation and ultimately apoptosis. The authors focussed their attentions of SRSF1 and RNPS1, both of which are known to prevent R-loops.\nThe first part of the work involved a derivation of a new information weight matrix from ENCODE eCLIP datasets. I am unable to comment on the methods used, but it is reassuring that the resultant consensus sequences for SRSF1 matched a coalescence of previous results from SELEX, RNA-seq and structural work. The authors then analysed the occurrence and distribution of all motifs in both viral and genomic transcripts that had a higher information content than the mean of the information content for all the binding sites in the model for each protein. It was assumed, but has not yet been tested, that the sites elected are the stronger binding sites. It is not clear whether this has any biological relevance, i.e., how bound lifetimes (affinity) vary across all the sequences in the model and whether the threshold chosen is likely to reflect the real behaviour of the proteins. Nonetheless, in the context of this heuristic work, this is not an unreasonable choice to make.\nThe authors analysed in particular the occurrence of clusters of these sites, i.e., where sites were within 25 nts of each other, and they looked for a correspondence between the locations of these clusters and sites located by high-throughput methods at which it was known that R-loops are likely to form. This correspondence was not strong, although the binding sites for RNPS1, SRSF1 and hnRNP A1 were closer than average in these regions. This is followed by a comparison of gene expression in patients with acute radiation sickness with those infected by the RNA viruses, which concluded that certain DNA damage-related proteins were expressed more highly in both types of patient, consistent with the overall hypothesis.\nThe final part describes an attempt to calculate the effects of viral genomes on the availability of SRSF1 in cells. The authors assume that the Kd is 0.8 µM. The Kd term they use is taken from assays with just the second RRM domain, but the value with both RRM domains is around 0.2 µM (Anczukow et al. (2015))1. The native protein is, of course, affected by its RS domain and therefore the state of phosphorylation. The authors also assume that there is no free protein, but this is not supported by any arguments. Taking the Kd to be 0.2 µM, the cellular concentration of the protein to be 3.6 µM (Hein et al. (2015))2 and the concentration of sites to be as described by the authors, this seems reasonable. However, and this is a significant caveat, the protein is not distributed evenly throughout the cell: it is largely nuclear and, within the nucleus, may be sequestered in speckles. Thus, the concentration in the cytoplasm, where it would encounter the viral RNA, might be much lower and thus affect the authors’ argument. The authors’ model involves competition between the nascent transcripts (nuclear) and the viral RNA (cytoplasmic) for SRSF1 binding, and any difference in the local concentrations and proportions of sites bound would undermine the model. However, none of the values required are known accurately and so, again, for the purpose of developing a model it is reasonable to make these simplifying assumptions.\nOverall, this is an interesting piece of work that makes the best use of the limited data available to support a model that proposes new and plausible routes by which RNA viruses could cause widespread apoptosis. It would be improved by a more rigorous discussion of the assumptions made, as noted above.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6192",
"date": "06 Jan 2021",
"name": "Peter Rogan",
"role": "Author Response",
"response": "It was assumed, but has not yet been tested, that the sites elected are the stronger binding sites. It is not clear whether this has any biological relevance, i.e., how bound lifetimes (affinity) vary across all the sequences in the model and whether the threshold chosen is likely to reflect the real behaviour of the proteins. Response: The IWMs derived in this manuscript may not yet been verified in the laboratory, however we have verified binding sites by such approaches previously (Vyhlidal, Rogan et al. J Biol Chem. 2004. 279:46779-86). The information theory framework used to derive them has been well validated and the relationships between information content and binding affinity have been rigorously proven (Schneider. J Theor Biol. 1997; 189:427-41). While the models used in this study were recently derived, other information theory-based RNA-protein binding site models have been utilized in hundreds of published studies, some involving the verification of phenotypes of mutations that alter splice sites, and others applied to transcription factor binding site recognition (Rogan et al. Hum Mutat. 1998; 12: 153–171; Rogan et al. Pharmacogenetics. 2003; 13: 207–218; Caminsky et al. F1000Res. 2014; 3:282; Lu et al. Nucl. Ac. Res. 45: e27, 2017). Our assumptions regarding binding site strength are well founded both theoretically and experimentally for many proteins. Quantification of the predicted strengths of these binding sites made by these models is reasonable and may likely reflect actual protein binding events. The Kd term they use is taken from assays with just the second RRM domain, but the value with both RRM domains is around 0.2 µM (Anczukow et al. (2015). The authors also assume that there is no free protein, but this is not supported by any arguments. Response: We are grateful to the reviewer for pointing out the updated dissociation constant of SRSF1 (Anczukow et al. 2015; Ref. 54), based on assays that included both RRM domains. We had inadvertently overlooked this study. The Kd value used in the initial version of our paper (0.8µM) was based on an earlier publication from the same group (Cléry et al. Proc Natl Acad Sci U S A. 2013; 110(30):E2802–2811). The new Kd value altered Scatchard analysis presented in Figure 6. The number of doublings for replicated virus was significantly increased, which reduced the discrepancy with the number of viral genomes required to compete with host transcriptome biding sites in A549 cell lines. In this revision, we have recomputed all values based on Kd from Anczukow et al. (2015) and have updated the main text, Figure 6 and Section 1 Tables 9A and 9B (extended data; Ref. 39). We previously did not include a justification for our assumption that [L] = 0 (no free protein) in our derivation of the Scatchard plot. The proposed mechanism relies on the likelihood that these RBPs are largely sequestered by binding to viral sequences, so that their effective concentration in the nucleus is inadequate to prevent R-loops (thus, [L] ≈ 0). SARS-CoV-2 replication is highly efficient and rapid, leading to levels constituting up to 60% of the total cellular RNA (Blanco-Melo et al. 2020; Ref. 55). Viral replication produces an excess of viral binding sites that will perturb the equilibrium between bound and free RBPs (Le Chatelier’s Principle) and drive binding of free RBPs and reduce the pool of free RBPs. The degree of viral replication that depletes nuclear RBP concentrations to a point at which the abundance of these factors becomes insufficient to prevent R-loop formation is not known. To clarify our assumption that all RBPs are bound to viral (or host), we have added the following to the Results: “We assumed no free RNA binding protein (that the vast majority of SRSF1 is bound to either host or viral binding sites) as the concentration of free RBPs is likely to be low due to sequestration of RBPs by the excess of viral sequences present in infected cells (~60% of all RNA 55 )” and to the Methods: “Upon infection and viral replication, it is assumed there is no free RNA binding protein (all RBP is assumed to be bound to either viral or host RNA).” However, and this is a significant caveat, the protein is not distributed evenly throughout the cell: it is largely nuclear and, within the nucleus, may be sequestered in speckles. Thus, the concentration in the cytoplasm, where it would encounter the viral RNA, might be much lower and thus affect the authors’ argument. The authors’ model involves competition between the nascent transcripts (nuclear) and the viral RNA (cytoplasmic) for SRSF1 binding, and any difference in the local concentrations and proportions of sites bound would undermine the model. Response: The reviewer has commented that the difference in local RBP concentrations between the cytoplasm and the nucleus would have an impact on the amount of RBP (such as SRSF1) that could possibly be sequestered in the cytoplasm by viral RNA. We do not state that we assume that these RBPs are uniformly distributed within the cell. We suggest that viral RNA binds to newly synthesized RBPs that have been translated in the cytoplasm, which could result in such an imbalance by limiting their nuclear entry (last paragraph of the “Proposed molecular pathogenetic mechanism of RNA-viral infection” section). It was also illustrated in panels 3 and 4 of the infographic of our proposed mechanism (Ref. 28; (http://doi.org/10.6084/m9.figshare.12718799.v2)."
}
]
}
] | 1
|
https://f1000research.com/articles/9-943
|
https://f1000research.com/articles/9-52/v2
|
31 Jan 20
|
{
"type": "Brief Report",
"title": "Use of the informational spectrum methodology for rapid biological analysis of the novel coronavirus 2019-nCoV: prediction of potential receptor, natural reservoir, tropism and therapeutic/vaccine target",
"authors": [
"Veljko Veljkovic",
"Júlia Vergara-Alert",
"Joaquim Segalés",
"Slobodan Paessler",
"Júlia Vergara-Alert",
"Joaquim Segalés",
"Slobodan Paessler"
],
"abstract": "A novel coronavirus recently identified in Wuhan, China (2019-nCoV) has expanded the number of highly pathogenic coronaviruses affecting humans. The 2019-nCoV represents a potential epidemic or pandemic threat, which requires a quick response for preparedness against this infection. The present report uses the informational spectrum methodology to identify the possible origin and natural host of the new virus, as well as putative therapeutic and vaccine targets. The performed in silico analysis indicates that the newly emerging 2019-nCoV is closely related to severe acute respiratory syndrome (SARS)-CoV and, to a lesser degree, Middle East respiratory syndrome (MERS)-CoV. Moreover, the well-known SARS-CoV receptor (ACE2) might be a putative receptor for the novel virus as well. Actin protein was also suggested as a host factor that participates in cell entry and pathogenesis of 2019-nCoV; therefore, drugs modulating biological activity of this protein (e.g. ibuprofen) were suggested as potential candidates for treatment of this viral infection. Additional results indicated that civets and poultry are potential candidates for the natural reservoir of the 2019-nCoV, and that domain 288-330 of S1 protein from the 2019-nCoV represents promising therapeutic and/or vaccine target.",
"keywords": [
"2019-nCoV",
"Wuhan coronavirus",
"SARS",
"MERS"
],
"content": "Introduction\n\nFears are mounting worldwide over the cross-border spread of the new strain of coronavirus (denoted as 2019-nCoV) originated in Wuhan, the largest city in central China, after its spread to Thailand and Japan. The newly emerging pathogen belongs to the same virus family as the deadly severe acute respiratory syndrome and Middle East respiratory syndrome coronaviruses (SARS-CoV and MERS-CoV, respectively). The World Health Organization (WHO) has recently published surveillance recommendations for a possible “large epidemic or even pandemic” of the novel coronavirus and it has issued guidelines for hospitals across the world. However, many questions about 2019-nCov remain unanswered: (i) what is the origin and/or natural reservoir of the virus? (ii) is it easily transmitted from human to human? and (iii) what are the potential diagnostic, therapeutic and vaccine targets? Currently, only nucleotide sequences of eight human 2019-nCoV isolates are available without any additional information about biological properties of the virus, beyond the morphology confirmation of the virion using electronic microscopy. This is likely not enough information to answer the important abovementioned questions.\n\nThe informational spectrum method (ISM), a virtual spectroscopy method for analysis of proteins, is based on the fundamental electronic properties of amino acids and requires only nucleotide sequence availability to investigate proteins1. For this reason, ISM was previously used for analysis of novel viruses for which little or no information were available2–5. Here, the 2019-nCoV was analyzed with ISM to identify its possible origin and natural host, as well as putative therapeutic and vaccine targets.\n\n\nMethods\n\nThe S1 surface protein sequences from 8 human 2019-nCoV, deposited in the publicly available GISAID database (assessed on January 19, 2020), were analyzed by ISM. The studied sequences were BetaCoV/Wuhan/IVDC-HB-04/2020, BetaCoV/Wuhan/IVDC-HB-01/2019, BetaCoV/Wuhan/IVDC-HB-05/2019, BetaCoV/Wuhan/IPBCAMS-WH-01/2019, BetaCoV/Wuhan/WIV04/2019, BetaCoV/Wuhan-Hu-1/2019, BetaCoV/Nonthaburi/61/2020, and BetaCoV/Nonthaburi/74/2020.\n\nIn the phylogenetic analysis, different amino acid sequences of other coronaviruses were also included: (i) S1 proteins from the following viruses: AVP78042, AVPvp78031, AY304486, AY559093, JX163927, YN2018B, KY417146, used already by other authors in the study of the phylogenetic relationship between 2019-nCoV and nearest bat and SARS-like CoVs (GISAID database); and (ii) S1 proteins from three first isolated human MERS-CoV: AGG22542, AFS88936, AFY13307, deposited in the GISAID database\n\nDetailed description of the sequence analysis based on ISM has been published elsewhere2. According to this approach, sequences (protein or DNA) are transformed into signals by assignment of numerical values of each element (amino acid or nucleotide). These values correspond to electron-ion interaction potential6, determining electronic properties of amino acid/nucleotides, which are essential for their intermolecular interactions. The signal obtained is then decomposed in a periodical function by the Fourier transformation. The result is a series of frequencies and their amplitudes. The obtained frequencies correspond to the distribution of structural motifs (primary structure) with defined physico-chemical characteristics responsible for the biological function of the putative protein corresponding to the analyzed sequence. When comparing proteins that share same biological or biochemical function, the technique allows detection of code/frequency pairs that are specific for their common biological properties. The method is insensitive to the location of the motifs and, therefore, does not require previous alignment of the sequences. In addition, this is the only method that allows immediate functional analysis.\n\nThe phylogenetic tree of S1 proteins from coronaviruses was generated with the ISM-based phylogenetic algorithm ISTREE, previously described in detail elsewhere7. In the presented analysis, we calculated the distance matrix with the amplitude on the frequency F(0.257) as the distance measure between sequences.\n\n\nResults and discussion\n\nIn order to compare informational similarity between 2019-nCoV, SARS-CoV, MERS-CoV and Bat SARS-like CoV, the cross-spectra (CS) of S1 proteins from these viruses were calculated. Figure 1a shows the CS of 2019-nCoV, SARS-CoV and MERS-CoV. These CS contain only one dominant peak corresponding to the frequency F(0.257). Figure 1b displays the CS of S1 proteins from 2019-nCoV and Bat SARS-like CoV. Amplitudes in these latter CS are significantly lower than in those CS presented in Figure 1a. These results show that (i) S1 proteins from 2019-nCoV, SARS-CoV, MERS-CoV and Bat SARS-like CoV encode common information, which is represented with the frequency F(0.257), and (ii) S1 proteins from 2019-nCoV are remarkable more informationally similar with S1 from SARS-CoV and MERS-CoV than with S1 from Bat SARS-like CoV. This suggests that biological properties of 2019-nCoV are apparently more similar to SARS-CoV and MERS-CoV than to Bat SARS-like CoV.\n\n(a) CS of S1 from SARS-CoV, MERS-CoV and 2019-nCoV; (b) CS of Bat SARS-like CoV and 2019-NCov. The abscissa represents the frequencies from the Fourier transform of the sequence of electron-ion interaction potential corresponding to the amino-acid sequence of proteins. The lowest frequency is 0.0 and the highest is 0.5. The ordinate represents the signal-to-noise ratio (the ratio between signal intensity at one particular IS frequency and the main value of the whole spectrum, S/N).\n\nTo confirm this conclusion, the ISM-base phylogenetic tree for S1 proteins was calculated (Figure 2). In this calculation the amplitude on the frequency F(0.257) was used as the distance measure. As observed in Figure 2, all analyzed 2019-nCoV S1 amino acid sequences are grouped with SARS-CoV and MERS-CoV and separated from Bat SARS-like CoV. This indicates that 2019-nCoV are more phylogenetically similar to SARS-CoV and MERS-CoV than to Bat SARS-like CoV. This result differs from those obtained with the homology-based phylogenetic analysis, which showed that 2019-CoV are closely related to Bat SARS-like CoV (https://platform.gisaid.org/epi3/frontend#lightbox1296857287).\n\nThe frequency F(0.257) as the distance measure was used.\n\nIt has been previously shown that the dominant frequency in the informational spectrum of viral envelope proteins corresponds to interaction between the virus and its receptor2,3,7,8. The ISM analysis showed that the frequency component F(0.257) is present in the CS of S1 SARS-CoV and its receptor angiotensin converting enzyme 2 (ACE2)9, but not in the CS of S1 MERS-CoV and its main receptor dipeptidyl peptidase 4 (DPP4)10. Of note is that both receptors ACE2 and DPP4 are expressed in airway epithelia. Presence of F(0.257) in the informational spectrum of MERS-CoV (Figure 1) suggests also possible interaction between this virus and the ACE2. The dominant peak on the frequency F(0.257) in the CS of S1 from SARS-CoV and MERS-CoV and ACE2 supports this possibility (Figure 3), although this has not been formally proved for MERS-CoV11.\n\nThe abscissa and the ordinate are as described in Figure 1.\n\nAs it is shown in Figure 1a, the frequency F(0.257) is also present in the informational spectrum of the 2019-nCoV, suggesting that ACE2 might be the receptor for this novel coronavirus too. Calculation of the CS for S1 protein from the 2019-nCoV and all ACE2 sequences available at the UniProt database revealed that the highest amplitudes on the frequency F(0.257) correspond to ACE2 from civet and chicken. This result indicates that these species can be included as potential candidates for the natural reservoir of the 2019-nCoV. However, it is possible that 2019-nCoV viruses use very different receptors in the natural host(s) and not only the ACE2 as it is the putative case in humans.\n\nFinally, the S1 amino acid sequence from the 2019-nCoV was scanned to look for the domain that gives the highest contribution to the information represented by the frequency F(0.257) (Figure 4a). This analysis revealed domain 266–330 (numbering concerns the maturated protein) is essential for interaction of 2019-nCoV with ACE2. Of note is the striking homology between these domains of S1 proteins from 2019-nCoV and SARS-CoV, but not from MERS-CoV for which ACE2 is not the main receptor (Figure 4b).\n\n(a) Mapping of the domain of S1 protein from 2019-nCoV (BetaCoV/Wuhan/IVDC-HB-01/2019) which gives the dominant contribution to the information represented with the frequency F(0.257). (b) Sequence homology between domains of S1 proteins from SARS-CoV and 2019-nCoV with essential contribution to the information corresponding to the frequency F(0.257).\n\nFurther, S1 spike proteins from SARS-CoV (Table 1) and 2019-nCoV (Table 2) were compared. The CS of S1 proteins from SARS-CoV (Figure 5a) and 2019-nCoV (Figure 5b) were assessed. Principal information encoded in S1 proteins from SARS-CoV and 2019-nCoV is represented with two different frequencies F(0.222) and F(0.478), respectively. This result indicates some potential difference(s) in the virus-host interaction of these two viruses although they apparently use the same receptor ACE2.\n\n(a) CS of S1 proteins from human SARS-CoV; (b) CS of S1 proteins from 2019-nCoV; (c) CS of mammalian actin proteins. The abscissa and the ordinate are as described in Figure 1.\n\nTo identify the host proteins involved in the attachment and/or internalization of the 2019-nCoV, the UniProt database (https://www.uniprot.org) was screened by ISM for human proteins with the dominant peak on the frequency F(0.478). The list of human proteins that have a dominant peak in IS at the frequency F(0.478) are given in Table 3. According to the IS criterion, these proteins are potential candidate interactors with the 2019-nCoV S1 protein. Further, literature data mining was performed to identify which proteins presented in Table 3 might be involved in the processes of infection with human coronaviruses. This analysis revealed that the actin protein plays an important role in the early entry events during human coronavirus infections12. Actin proteins were selected as the best candidate interactors for the 2019-nCoV among the host proteins that are characterized with frequency F(0.478). Figure 5c shows that CS of actins from different mammalian species (Table 4) contains the dominant peak on F(0.478), suggesting that these proteins probably encode the conserved information important for their biological function.\n\nThe data mining of the PubMed database (www.ncbi.nlm.nih.gov/pubmed/) also showed that actin protein plays an important role in the rapid virus cell-to-cell spread and dissemination of infection13. Additionally, the actin filament reorganization is a key step in lung inflammation induced by systemic inflammatory responses caused by infectious agents14. These findings indicate that interaction between actin proteins and the S1 could be involved in the infection and pathogenesis of 2019-nCoV. In consequence, the possibility to interfere on this interaction might represent a valid hypothesis for development of promising prevention and therapeutic strategies.\n\nInterestingly, further data mining revealed that ibuprofen (FDA approved drug with excellent safety record) attenuates interleukin-1β-induced inflammation as well as actin reorganization15. Actin was also found to be the primary component by which ibuprofen can bind to the tissue in different organs16. This suggests that ibuprofen might impact the 2019-nCoV-induced disease by indirect interaction with actin proteins. Previously, ibuprofen was predicted as a candidate entry inhibitor for Ebola virus using the same in silico approach17, and this prediction was confirmed experimentally at a later time point18,19. These results prompt the possibility to experimentally test the effects of ibuprofen on 2019-nCoV infection under in vitro and in vivo conditions.\n\nIn silico methods are considered very important tools to generate first hypotheses and identify first drug candidates against newly discovered agents, like in the case of 2019-nCoV, especially in the short-term. ISM, a technology based on electronic biology, allowed identifying potential importance of human actin proteins for viral infection/dissemination as well as one FDA approved drug that may have an indirect antiviral activity within weeks of the initial outbreak. However, additional experiments are required to confirm our initial findings.\n\nIn conclusion, results of the presented in silico analysis suggest the following: (i) the newly emerging 2019-nCoV is highly related to SARS-CoV and, to a lesser degree, MERS-CoV, and ACE2 is a likely receptor of it; (ii) civets and poultry are potential candidates for the natural reservoir of the 2019-nCoV, (iii) human actin proteins possibly participate in attachment/internalisation of 2019-nCoV, (iv) drugs which interact with actin proteins (e.g. ibuprofen) should be investigated as possible therapeutics for treatment of 2019-nCoV infection, and (v) domain 266-330 of S1 protein from the 2019-nCoV represents promising therapeutic and/or vaccine target. Further research on these issues are needed, including the development of reverse genetics and animal models to study the biology of 2019-nCoV.\n\n\nData availability\n\nSequence data of the viruses were obtained from the GISAID EpiFlu™ Database. To access the database each individual user should complete the “Registration Form For Individual Users”, which is available alongside detailed instructions. After submission of the Registration form, the user will receive a password. There are not any other restrictions for the access to GISAID. Conditions of access to, and use of, the GISAID EpiFlu™ Database and Data are defined by the Terms of Use.",
"appendix": "References\n\nVeljković V, Cosić I, Dimitrijević B, et al.: Is it possible to analyze DNA and protein sequences by the methods of digital signal processing? IEEE Trans Biomed Eng. 1985; 32(5): 337–41. PubMed Abstract | Publisher Full Text\n\nVeljkovic V, Niman HL, Glisic S, et al.: Identification of hemagglutinin structural domain and polymorphisms which may modulate swine H1N1 interactions with human receptor. BMC Struct Biol. 2009; 9: 62 . PubMed Abstract | Publisher Full Text | Free Full Text\n\nVeljkovic V, Glisic S, Muller CP, et al.: In silico analysis suggests interaction between Ebola virus and the extracellular matrix. Front Microbiol. 2015; 6: 135. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaessler S, Veljkovic V: Prediction of influenza vaccine effectiveness for the influenza season 2017/18 in the US [version 1; peer review: 2 approved]. F1000Res. 2017; 6: 2067. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKoma T, Veljkovic V, Anderson DE, et al.: Zika virus infection elicits auto-antibodies to C1q. Sci Rep. 2018; 8(1): 1882. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVeljković V, Slavić I: Simple general-model pseudopotential. Phys Rev Let. 1972; 29: 105. Publisher Full Text\n\nPerovic VR, Muller CP, Niman HL, et al.: Novel phylogenetic algorithm to monitor human tropism in Egyptian H5N1-HPAIV reveals evolution toward efficient human-to-human transmission. PLoS One. 2013; 8(4): e61572. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchmier S, Mostafa A, Haarmann T, et al.: In Silico Prediction and Experimental Confirmation of HA Residues Conferring Enhanced Human Receptor Specificity of H5N1 Influenza A Viruses. Sci Rep. 2015; 5: 11434. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi W, Moore MJ, Vasilieva N, et al.: Angiotensin-converting enzyme 2 is a functional receptor for the SARS coronavirus. Nature. 2003; 426(6965): 450–4. PubMed Abstract | Publisher Full Text\n\nRaj VS, Mou H, Smits SL, et al.: Dipeptidyl peptidase 4 is a functional receptor for the emerging human coronavirus-EMC. Nature. 2013; 495(7440): 251–4. PubMed Abstract | Publisher Full Text\n\nMüller MA, Raj VS, Muth D, et al.: Human coronavirus EMC does not require the SARS-coronavirus receptor and maintains broad replicative capability in mammalian cell lines. mBio. 2012; 3(6): pii: e00515-12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOwczarek K, Szczepanski A, Milewska A, et al.: Early events during human coronavirus OC43 entry to the cell. Sci Rep. 2018; 8(1): 7124. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDoceul V, Hollinshead M, van der Linden L, et al.: Repulsion of superinfecting virions: a mechanism for rapid virus spread. Science. 2010; 327(5967): 873–876. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDu L, Zhou J, Zhang J, et al.: Actin filament reorganization is a key step in lung inflammation induced by systemic inflammatory response syndrome. Am J Respir Cell Mol Biol. 2012; 47(5): 597–603. PubMed Abstract | Publisher Full Text\n\nLi R, Song X, Li G, et al.: Ibuprofen attenuates interleukin-1β-induced inflammation and actin reorganization via modulation of RhoA signaling in rabbit chondrocytes. Acta Biochim Biophys Sin (Shanghai). 2019; 51(10): 1026–1033. PubMed Abstract | Publisher Full Text\n\nKunsman GW, Rohrig TP: Tissue distribution of ibuprofen in a fatal overdose. Am J Forensic Med Pathol. 1993; 14(1): 48–50. PubMed Abstract | Publisher Full Text\n\nVeljkovic V, Goeijenbier M, Glisic S, et al.: In silico analysis suggests repurposing of ibuprofen for prevention and treatment of EBOLA virus disease [version 1; peer review: 2 approved]. F1000Res. 2015; 4: 104. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhao Y, Ren J, Harlos K, et al.: Toremifene interacts with and destabilizes the Ebola virus glycoprotein. Nature. 2016; 535(7610): 169–172. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaessler S, Huang C, Sencanski M, et al.: Ibuprofen as a template molecule for drug design against Ebola virus. Front Biosci (Landmark Ed). 2018; 23: 947–953. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "60149",
"date": "18 Mar 2020",
"name": "Rita Casadio",
"expertise": [
"Reviewer Expertise Computational Biology",
"Structural bioinformatics",
"Functional annotation",
"Machine and deep learning"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nPossibly the ISM method invented by the authors has high potentiality. Here, apparently, the method is applied to answer urgent questions in relation to sequence analysis of the novel coronavirus 2019-nCoV. It is interesting to verify that different papers at the moment seem to have reached the same conclusions, although in my opinion a comparison with other methods which are already published would add to the paper. On top, R.Yan et al,1 Science 2020 4 March, beautifully detailed the putative region of the virus/ACE2 interaction. I recommend the authors quote this finding as well and if possible add the reference. This would help in validating a quite interesting method of sequence analysis.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Partly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "61263",
"date": "16 Apr 2020",
"name": "Maria Salvato",
"expertise": [
"Reviewer Expertise Immunology",
"viral pathogenesis",
"viral genetics"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript presents a machine-learning analysis of the published sequences of the novel 2019-nCoV. The authors use the Informational Spectrum Method (ISM), a virtual spectroscopy method for protein analysis based on the electronic properties of each amino acid. Their goal is to identify sites on the virus most likely to interact with other molecules like drugs, antibodies or viral receptors.\n\nDue to the rapidity of this field and the time elapsed since the manuscript was submitted (Jan 27), most of their conclusions are no longer new: nCoV is most related to SARS-CoV and less to MERS-CoV, ACE2 is a likely receptor, the natural reservoir might be civets and poultry, human actin proteins participate in internalization, ibuprofen that interacts with actin proteins should be investigated as a therapeutic, and finally that domain 266-330 of the S1 protein should be targeted by drugs or vaccines.\n\nIt is a nice piece of work. The conclusions could be updated, for example they could say that the first of these predictions are supported by recent publications, that bat CoVs now appear to be the most closely related and bats are more likely to be the natural reservoir, and the link between ibuprofen/actin interactions and viral entry remains an exciting path for future therapeutics.\n\nMinor corrections\nIn the first sentence of “Update” the authors refer to ‘these proteins’…do they mean Actin and SARS-CoV proteins? If they only mean actin they should say so, and then the sentence would read: “…actin protein is suggested as a host factor that participates in infection and pathogenesis of 2019-nCoV. Drugs modulating the biological activity of actin (e.g., ibuprofen) were suggested as candidates that should be investigated for the treatment of 2019-nCoV infection.“\n\nIn the UPDATE, last sentences should say \"...which are presented in Figure 5.\"\n\nIntroduction pg 3 “Fears are mounting worldwide over the cross-border spread of the new strain of coronavirus (denoted as 2019-nCoV) originated in Wuhan,….” Instead say “…that originated in Wuhan…”\n\nIntroduction pg 3 “…eight human 2019-nCoV isolates are available without any additional information about biological properties of the virus, beyond the morphological confirmation…” Replace “morphology” with “morphological”, it is an adjective.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 2
|
https://f1000research.com/articles/9-52
|
https://f1000research.com/articles/9-1286/v1
|
30 Oct 20
|
{
"type": "Research Article",
"title": "Age, neutrophil lymphocyte ratio, and radiographic assessment of the quantity of lung edema (RALE) score to predict in-hospital mortality in COVID-19 patients: a retrospective study",
"authors": [
"Anggraini Dwi Sensusiati",
"Muhammad Amin",
"Nasronudin Nasronudin",
"Alfian Nur Rosyid",
"Nanda Aulya Ramadhan",
"Rofida Lathifah",
"Eva Puspitasari",
"Ria Indah Wahyuningtyas",
"Erika Soebakti",
"Muhammad Amin",
"Nasronudin Nasronudin",
"Alfian Nur Rosyid",
"Nanda Aulya Ramadhan",
"Rofida Lathifah",
"Eva Puspitasari",
"Ria Indah Wahyuningtyas",
"Erika Soebakti"
],
"abstract": "Background: Case fatality rate of COVID-19 patients in Surabaya is higher than global cases. Thus, it is important to identify risk factors to reduce the mortality rate. This study aimed to assess the factors associated with hospital mortality of COVID-19 patients, and develop a prediction score based on these findings. Methods: We analyzed 111 patients, who were diagnosed with COVID-19 based on reverse-transcriptase polymerase chain reaction. The following patient characteristics were obtained from records: age, gender, type of symptoms, onset of symptoms, neutrophil lymphocyte ratio (NLR), absolute lymphocyte count, chest x-ray abnormalities, lung involvement, type of lesion, radiographic assessment of the quantity of lung edema (RALE) score, and mortality. Data were analyzed using SPSS 25.0. Results Multivariate analysis showed that age >50 years (p=0.043), NLR score >5.8 (p=0.016) and RALE score >2 (p=0.002) can predict the mortality of COVID-19 patients in the hospital. ROC curve analysis of the score ability to predict mortality showed an area under the curve of 0.794. The cut-off point is 4.5, with a sensitivity of 96.7% and specificity of 49.4% to predict the mortality of COVID-19 patient in the hospital. Conclusions Age, NLR score and RALE score were associated with mortality of COVID-19 patients in the hospital and could be used as a predictor for discharge probability of COVID-19 patients in low health care resource setting. The prediction score may be useful for frontline physicians to effectively manage patients with a higher score to prevent mortality.",
"keywords": [
"COVID-19 mortality",
"NLR",
"RALE Score"
],
"content": "Introduction\n\nCases of pneumonia of unknown etiology were first reported in the city of Wuhan, China, at the end of December 2019. After identification, the etiology of these cases was a new type of coronavirus (Wang et al., 2020). The World Health Organization officially called this coronavirus disease 2019 (COVID-19) (WHO, 2020). COVID-19 is caused by severe acute respiratory syndrome coronavirus-2 (SARS-Co-V2) virus (Gorbalenya et al., 2020).\n\nSince December 2019, more than 20 million people have been diagnosed with COVID-19, and more than 700,000 have died (Worldometers, 2020). In Indonesia, 128,776 people have been diagnosed with COVID-19, and 83,710 have died since March 2020 (Gugus Tugas Percepatan Penanganan COVID-19, 2020). The global case fatality rate (CFR) of COVID-19 in August was 3.7%. In Indonesia, the CFR is slightly higher (4.5%). Meanwhile, the CFR in Surabaya (the second biggest city in Indonesia) is almost twice that of Indonesia for the same period (8.9%). For hospitals in Surabaya, finding characteristics and risk factors to predict mortality is of utmost importance in order to reduce the mortality rate in the future. It is also important to develop a prediction score to assess patients during early stages, when patients receive treatment at tertiary hospitals.\n\nSeveral studies have found risk factors related to COVID-19 mortality (Gupta et al., 2020; Galloway et al., 2020; Jalili et al., 2020; Lippi & Plebani, 2020; Zhao et al., 2020). However, currently, there is no study that analyzes risk factors for COVID-19 mortality in hospitals for the Indonesian population. Existing prediction scores also based on complex laboratory findings, which are less feasible to be used in low resource health centers. Therefore, the purpose of this study was to assess the factors associated with COVID-19 mortality in hospital patients and develop a prediction score based on these findings.\n\n\nMethods\n\nThis was a retrospective study that was conducted in the Emergency Department of Airlangga University Teaching Hospital. We included all the patients who fulfilled the inclusion and exclusion criteria, who were admitted for COVID-19 from March 13, 2020 to May 15, 2020. A total of 111 patients met the criteria and included in this study.\n\nWe collected data regarding clinical symptoms and the date of onset of symptoms from emergency department medical records.\n\nThe inclusion criteria for this retrospective study were: (1) patients with a chief complaint of one COVID-19-related symptom, including: fever, dry cough, tiredness, aches and pains, nasal congestion, headache, conjunctivitis, sore throat, diarrhea, loss of taste or smell, a rash on skin, or discoloration of fingers or toes (WHO, 2020); (2) confirmed SARS-CoV-2 infection by reverse-transcriptase polymerase chain reaction (RT PCR) using nasopharyngeal and oropharyngeal specimens; (3) patients who underwent chest x-ray (CXR) on the day of admission to the hospital.\n\nThe following patients were excluded: (1) asymptomatic patients; (2) those with negative results for SARS-CoV-2 infection by RT-PCR.\n\nAll patients underwent CXR and laboratory examination on the day of admission to the emergency department.\n\nPatients were divided into two groups based on the outcome of the patient: group I (discharged with negative results for SARS-CoV-2 infection by RT-PCR test); and group II (died).\n\nThe following patient characteristics were obtained from the medical records: age, gender, type of symptoms, onset of symptoms, neutrophil lymphocyte ratio (NLR), absolute lymphocyte count (ALC), CRX abnormalities, lung involvement, type of lesion, radiographic assessment of the quantity of lung edema (RALE) score, and mortality.\n\nImage acquisition and evaluation. Radiology data was collected from the radiology department. All the patients underwent an anteroposterior projection chest radiography at full inspiration where possible. The results were reviewed by two radiologists (A.D.S, a radiologist with 25 years of experience and E.S, a radiologist with four years of experience) based on consensus.\n\nCXRs were evaluated for the presence of pulmonary alterations, type of pulmonary alterations, and their distribution. CXR alterations that were found specifically in COVID-19 patients were defined according to the Fleischner Society’s nomenclature, available in the Glossary of Terms for Thoracic Imaging (Hansell et al., 2008), as follows:\n\n- Reticular alteration: a reticular pattern is a collection of innumerable small linear opacities that, by summation, produce an appearance resembling a net\n\n- Consolidation: as a homogeneous increase in pulmonary parenchymal attenuation that obscures the margins of vessels and airway walls\n\n- Ground-glass opacity (GGO): an area of hazy increased lung opacity, usually extensive, within which margins of pulmonary vessels may be indistinct\n\nThe distribution of pulmonary alterations was classified as lung involvement unilateral (right/left) or bilateral. Other features, such as pleural effusion, were also recorded.\n\nRadiograph scoring. A severity score was calculated to quantify the extent of the infection by adapting and simplifying the RALE score proposed by Warren et al. (2018). A score of 0-4 was assigned to each lung depending on the extent of involvement by consolidation or GGO (0 = no involvement; 1 = <25%; 2 = 25-50%; 3 = 50-75%; 4 = >75% involvement). The scores for each lung were summed to produce the final RALE score.\n\nStatistical analysis was performed using IBM SPSS Statistics Version 25.0. Continuous variables were expressed as mean ± standard deviation values. The frequency of symptoms, laboratory findings and CXR findings was shown as the number of incidence and percentage per cluster of groups. The correlation between the patient characteristic, symptoms, laboratory findings and CXR findings with the outcome was analyzed by logistic regression. We also conducted the Hosmer-Lemeshow test to evaluate the goodness of fit of the scoring model and conducted receiver operating characteristic (ROC) analysis to evaluate the sensitivity and specificity of the model.\n\nThis study received ethical approval from the Ethical Committee of Airlangga University Teaching Hospital (approval number 147/KEP/2020). Consent from the participants was waived by the committee.\n\n\nResults\n\nA total of 111 COVID-19 patients were evaluated. In total, 72.9% (n=81) of patients were discharged from the hospital, while 27.1% (30) patients died during their hospitalization. There were 45 men (48.6%) and 47 women (51.4%). Average patient age was 51±14.2 years old. The mean age for the patients who died was higher (55±12.8 years) compared to the patients who were discharged (48±14.8 years). Fever (32.6%) and shortness of breath (26%) were the most frequent symptoms for all patients. However, among the patients who died, shortness of breath was the most frequent symptom (40.7%). The mean NLLR score is higher on the patient who died (10.1 ±10.5) compared to patients who were discharged (5.3 ±5.4). In contrast, the mean ALC score was lower in patients who died (1130±252) compared to patients who were discharged (1349±702). Most of the patients showed GGO on their CXR (67.6%). Mean RALE score was higher in patients who died (5.3±2.5) compared to patients who were discharged (2.7±2.7) (Table 1).\n\nFrom Table 2, no significant relationship was observed between the outcome of the patients based on gender, type of symptoms, onset of symptoms, ALC score, x-ray abnormalities, and type of lesion. This suggested that those variables have no predictive value for mortality outcome for COVID-19 patients. There is a significant relationship between age, NLR score, lung involvement and RALE score towards the outcome of the patients (p-value <0.05). All of these variables were analyzed using logistic regression, which revealed there are three variables with p<0.05: age, NLR score and RALE score (Table 3).\n\nBased on the odds ratio (OR) value of each variable, the mortality risk for patients who were of an older age is 2.787 times higher than patients who were of a younger age. A higher NLR score has a 3.246 times higher mortality risk than patients who have lower NLR score. Mortality risk for patients who have a higher RALE score is 6.826 times higher than patients who have lower RALE score. It should be noted that three patients died while showing no abnormalities on their CXR.\n\nWe form a scoring model based on the OR of each variable, which can be seen in Table 4. If the variable were present, we gave the value of “1”, and if absent we gave the value of “0”. We also performed the Hosmer-Lemeshow test and concluded that the model is fit (p=0.802). Based on our scoring model, each score from 111 patients was calculated. The percentage of mortality was then calculated for each score. Table 5 shows that high mortality (>60%) was seen for a total score of 13, while low mortality (<10%) was seen for a total score ≤3.\n\nThe total score was then analyzed with ROC analysis to predict the sensitivity and specificity for the probability of mortality of COVID-19 patients in the hospital. Figure 1 shows that based on the ROC curve, the area under the curve (AUC) of the score is 0.794. The score has a cut off point of 4.5, with a sensitivity of 96.7% and specificity of 49.4%, to predict the mortality of COVID-19 patients in the hospital.\n\n\nDiscussion\n\nThis is the first study to analyze the radiologic and laboratory findings of COVID-19 patients in Indonesia. The main result of our study was that age, NLR score and RALE score have a strong relationship for mortality risk of COVID-19 patients in the hospital.\n\nThis study found that the average age of COVID-19 patients who died is 55 ±12.8 years. This finding is younger than a study in Iraq, who found that the average age of COVID-19 patients who died in hospital is 67.49 ± 15.28 years (Jalili et al., 2020). This study also showed that the mortality risk is three times higher in older age patient (≥50 years), and this finding is similar to a study in the United States, which found that an older age (≥60 years) is associated with a higher risk of death (Gupta et al., 2020). The younger average age of the patients who died might be due to the co-morbidities that also develop in the relatively younger age of the Indonesian population, especially for type 2 diabetes (Arifin et al., 2019).\n\nBased on our findings, there is no correlation between gender and patient mortality. The result is different from studies in the United States and Iran, which conclude that men has a higher risk of death than women (Gupta et al., 2020; Jalili et al., 2020). This could be due to the characteristic of the co-morbidities of the Indonesian population, which affect both men and women at an almost similar number (Kemenkes, 2019). We also found that there is no correlation between type and onset of symptoms to the outcome of the patients.\n\nThere is a correlation between NLR score and the patient outcome. Previously, it was concluded that higher NLR score related to the severe illness of the patients and could be an early identification for patient assessment in the early stage for intensive care unit (Liu et al., 2020; Yang et al., 2020). Our study showed that the higher NLR score (>5.8) would increase the risk of mortality by three times. In contrast, we found no correlation between ALC score and patient mortality in the hospital. However, various other research shows that low ALC score is related to disease severity for patients admitted to the hospital for COVID-19 (Wagner et al., 2020) and a higher risk of mortality (Lippi & Plebani, 2020).\n\nThis study showed that there is no correlation for CXR abnormalities, lung involvement and type of lesion to COVID-19 mortality in the hospital. Interestingly, we also found that three patients (10%) who died showed normal CXR. This could be due to the lower sensitivity of CXR to detect lung lesions compare to CT scan. However, it also shows that COVID-19 could become severe rapidly even though patients have normal CXR on the day of the admission.\n\nInterestingly, we found that RALE score is correlated with patient mortality, and a higher score (>2) will increase mortality risk by seven times. This finding is consistent with another study in the United Kingdom with a larger sample size, which showed that higher radiological severity score (>3) increased the incidence of critical admission or death (Galloway et al., 2020). CXR could become a key predictor of mortality because of the simplicity of modality and is broadly available across health care provider.\n\nThe prediction score developed in this study demonstrated good accuracy (AUC of 79.4) to predict the discharged probability of COVID-19 patients in the hospital with only three simple parameters. Another score showed good accuracy (AUC of 83) with more complicated parameters, such as health failure, procalcitonin, lactate dehydrogenase, chronic obstructive pulmonary disease, pulse oxygen saturation, heart rate, and age (Zhao et al., 2020). This suggests that our score can be used in a broader setting such as locations with low resource of health care service.\n\nThis study only analyzes patient characteristics, and laboratory and radiology findings related to COVID-19 mortality in the hospital. We do not evaluate the co-morbidities of the patient and the hospital setting, such as the location of the treatment where the patient died (emergency room, intensive care, low care), or the resource (facilities, human resource, intervention).\n\n\nConclusions\n\nAge, NLR score and RALE score were associated with mortality of COVID-19 patients in our hospital setting in an Indonesian population. It may be used as a predictor for mortality of COVID-19 patients in low health care resource settings. The prediction score may be useful for physicians to determine the mortality risk of a patient with COVID-19.\n\n\nData availability\n\nFigshare: datasheet f1000research 2909.xlsx, https://doi.org/10.6084/m9.figshare.13017899.v1 (Sensusiati, 2020).\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "References\n\nArifin B, van Asselt ADI, Setiawan D, et al.: Diabetes distress in Indonesian patients with type 2 diabetes: a comparison between primary and tertiary care. BMC Health Serv Res. 2019; 19(1): 773. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGalloway JB, Norton S, Barker RD, et al.: A clinical risk score to identify patients with COVID-19 at high risk of critical care admission or death: An observational cohort study. J Infect. 2020; 81(2): 282–288. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGorbalenya AE, Baker SC, Baric RS, et al.: Coronaviridae Study Group of the International Committee on Taxonomy of Viruses: Coronaviridae Study Group of the International Committee on Taxonomy of the species Severe acute respiratory syndrome-related coronavirus: classifying 2019-nCoV and naming it SARS-CoV-2. Nat Microbiol. 2020; 5(4): 536–544. Publisher Full Text\n\nGugus Tugas Percepatan Penanganan COVID-19: Peta Sebara. 2020. Reference Source\n\nGupta S, Hayek SS, Wang W, et al.: Factors Associated With Death in Critically Ill Patients With Coronavirus Disease 2019 in the US. JAMA Intern Med. 2020; e203596. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHansell DM, Bankier AA, MacMahon H, et al.: Fleischner Society: Glossary of Terms for Thoracic Imaging. Radiology. 2008; 246(3): 697–722. PubMed Abstract | Publisher Full Text\n\nJalili M, Payandemehr P, Saghaei A: Characteristics and Mortality of Hospitalized Patients With COVID-19 in Iran: A National Retrospective Cohort Study. Ann Intern Med. 2020; M20–2911. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKemenkes RI: Hasil Utama Riskesdas 2018. 2019. Reference Source\n\nLippi G, Plebani M: Laboratory abnormalities in patients with COVID-2019 infection. Clin Chem Lab Med. 2020; 58(7): 1131–1134. PubMed Abstract | Publisher Full Text\n\nLiu J, Liu Y, Xiang P, et al.: Neutrophil-to-lymphocyte ratio predicts critical illness patients with 2019 coronavirus disease in the early stage. J Transl Med. 2020; 18(1): 206. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSensusiati AD: datasheet f1000research 2909.xlsx. figshare. Dataset. 2020. http://www.doi.org/10.6084/m9.figshare.13017899.v1\n\nWagner J, DuPont A, Larson S, et al.: Absolute lymphocyte count is a prognostic marker in Covid-19: A retrospective cohort review. Int J Lab Hematol. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang C, Horby PW, Hayden FG, et al.: A novel coronavirus outbreak of global health concern. Lancet. 2020; 395(10223): 470–473. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWarren MA, Zhao Z, Koyama T, et al.: Severity scoring of lung oedema on the chest radiograph is associated with clinical outcomes in ARDS. Thorax. 2018; 73(9): 840–846. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorldometers: COVID-19 Coronavirus Pandemic. 2020. Reference Source\n\nWorld Health Organization: Statement on the Second Meeting of the International Health Regulations, Emergency Committee regarding the outbreak of novel coronavirus (2019-nCoV). 2020. Reference Source\n\nYang AP, Liu JP, Tao WQ, et al.: The diagnostic and predictive role of NLR d-NLR and PLR in COVID-19 patients. Int Immunopharmacol. 2020; 84: 106504. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhao Z, Chen A, Hou W, et al.: Prediction model and risk scores of ICU admission and mortality in COVID-19. PLoS One. 2020; 15(7): e0236618. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "74045",
"date": "20 Nov 2020",
"name": "Hamzaini Abdul Hamid",
"expertise": [
"Reviewer Expertise General Radiology",
"Paediatric Radiology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nCongratulation and I think the paper has been presented very well. The statistical analysis and tools were done appropriately.\n\nThis paper definitely will contribute to the management of COVID 19 patients i.e. in predicting which patient may deteriorate and need to be monitor closely.\nAlthough in this paper there is no correlation for CXR abnormalities, lung involvement and type of lesion to COVID-19 mortality. However, I think all the above CXR changes would also contribute collectively to the RALE score.\n\nOverall, well done and congratulation to the author. Just a minor error in spelling.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "74043",
"date": "20 Nov 2020",
"name": "Wiwien Heru Wiyono",
"expertise": [
"Reviewer Expertise Pulmonary medicine"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThere are some statements in this paper that are not so accurate.\nIntroduction:\nMy comment: To reduce the mortality rate in some hospitals in Surabaya is not by finding characteristics and risk factors to predict mortality. As I know at that time tracing in Surabaya still low in subjects, so the CFR in this city is still higher than Jakarta and the mean CFR in Indonesia.\n\"the purpose of this study was to assess the factors associated with COVID-19 mortality in hospital patients and develop a prediction score based on these findings.\" My comment: To assess the factors associated with mortality even to develop prediction score should use all factors. But this study actually only uses some parameters not include commorbidities.\nDiscussion: The statement:\n\" …….this finding is similar to a study in the US, which found that an older age (> 60 years) is associated with a higher risk of death is not so accurate, because:\"\nMy comments:\nThe difference between mean ages of >50 and >60 is so significant.\n\nCriteria for elder people in Indonesia is >60 years old.\n\nThe younger average of the patients in this study is not representative of Indonesia Covid patients.\n\nThe statement :\n\"CXR could become a key predictor of mortality because……….\"\n\nMy comment:\nThis statement is over confident and is in contradiction with the statement: \"This study showed that there is no correlation for CXR abnormalities……\"\n\nConclusion:\nThe statement:\n\"It may be used as a predictor for mortality of COVID 19 patients in low health care resource setting.\"\nMy comment :\nIt is not so true. The RALE score should be made by radiologists who have many experiences, so it only applicable in a referral hospital setting.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6219",
"date": "06 Jan 2021",
"name": "Anggraini Dwi Sensusiati",
"role": "Author Response",
"response": "The authors thank the Honorable Reviewer 2 for the excellent review of the paper and in-depth suggestions made to improve the quality of the paper. Followings are the responses to the suggestions of the Honorable Reviewer. The valuable suggestions and corrections are incorporated cautiously, and the paper is made clearer while revising the manuscript. Introduction: Comment: To reduce the mortality rate in some hospitals in Surabaya is not by finding characteristics and risk factors to predict mortality. As I know at that time tracing in Surabaya still low in subjects, so the CFR in this city is still higher than Jakarta and the mean CFR in Indonesia. *We acknowledged the possible cause of high fatality rate and added the sentence \"Several factors may associate with the higher CFR such as low testing, low contact tracing and poor COVID-19 register (Sorci et al., 2020). However, finding characteristics and risk factors to predict mortality is of utmost importance in order to prevent the mortality.\" \"the purpose of this study was to assess the factors associated with COVID-19 mortality in hospital patients and develop a prediction score based on these findings.\" Comment: To assess the factors associated with mortality even to develop prediction score should use all factors. But this study actually only uses some parameters not include commorbidities. *We acknowledged the study limitation which did not evaluate the commorbidities in the limitations section. We focused the research on the demographical characteristic, laboratory and radiological findings associated with COVID-19 mortality in hospital patients and develop a prediction score based on these findings. Discussion: The statement: \" …….this finding is similar to a study in the US, which found that an older age (> 60 years) is associated with a higher risk of death is not so accurate, because:\" Comments: The difference between mean ages of >50 and >60 is so significant. Criteria for elder people in Indonesia is >60 years old. The younger average of the patients in this study is not representative of Indonesia Covid patients. *We found similar finding in another research in Indonesia where the average age of COVID-19 patients who died is 58.2 years old (Rozaliyani et al., 2020). We also found that this finding is similar to a study in the United States, which found that the mortality risk of COVID-19 was 8.1 times higher among those ≥55 years compared with individuals ages <54 years (Yanez et al., 2020). The statement : \"CXR could become a key predictor of mortality because……….\" Comment: This statement is over confident and is in contradiction with the statement: \"This study showed that there is no correlation for CXR abnormalities……\" *We meant that the RALE score from CXR might become a key predictor Conclusion: The statement: \"It may be used as a predictor for mortality of COVID 19 patients in low health care resource setting.\" Comment: It is not so true. The RALE score should be made by radiologists who have many experiences, so it only applicable in a referral hospital setting. *We changed the \"low health care resource setting\" into \"health care centre where radiologists are available.\" We thank Prof. Wiwien for raising these points which we feel have improved the paper considerably. Sensusiati AD et al."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1286
|
https://f1000research.com/articles/9-1107/v1
|
09 Sep 20
|
{
"type": "Systematic Review",
"title": "Predictors of COVID-19 severity: a systematic review and meta-analysis",
"authors": [
"Mudatsir Mudatsir",
"Jonny Karunia Fajar",
"Laksmi Wulandari",
"Gatot Soegiarto",
"Muhammad Ilmawan",
"Yeni Purnamasari",
"Bagus Aulia Mahdi",
"Galih Dwi Jayanto",
"Suhendra Suhendra",
"Yennie Ayu Setianingsih",
"Romi Hamdani",
"Daniel Alexander Suseno",
"Kartika Agustina",
"Hamdan Yuwafi Naim",
"Muchamad Muchlas",
"Hamid Hunaif Dhofi Alluza",
"Nikma Alfi Rosida",
"Mayasari Mayasari",
"Mustofa Mustofa",
"Adam Hartono",
"Richi Aditya",
"Firman Prastiwi",
"Fransiskus Xaverius Meku",
"Monika Sitio",
"Abdullah Azmy",
"Anita Surya Santoso",
"Radhitio Adi Nugroho",
"Camoya Gersom",
"Ali A. Rabaan",
"Sri Masyeni",
"Firzan Nainu",
"Abram L. Wagner",
"Kuldeep Dhama",
"Harapan Harapan",
"Jonny Karunia Fajar",
"Gatot Soegiarto",
"Muhammad Ilmawan",
"Yeni Purnamasari",
"Bagus Aulia Mahdi",
"Galih Dwi Jayanto",
"Suhendra Suhendra",
"Yennie Ayu Setianingsih",
"Romi Hamdani",
"Daniel Alexander Suseno",
"Kartika Agustina",
"Hamdan Yuwafi Naim",
"Muchamad Muchlas",
"Hamid Hunaif Dhofi Alluza",
"Nikma Alfi Rosida",
"Mayasari Mayasari",
"Mustofa Mustofa",
"Adam Hartono",
"Richi Aditya",
"Firman Prastiwi",
"Fransiskus Xaverius Meku",
"Monika Sitio",
"Abdullah Azmy",
"Anita Surya Santoso",
"Radhitio Adi Nugroho",
"Camoya Gersom",
"Ali A. Rabaan",
"Sri Masyeni",
"Firzan Nainu",
"Abram L. Wagner",
"Kuldeep Dhama",
"Harapan Harapan"
],
"abstract": "Background: The unpredictability of the progression of coronavirus disease 2019 (COVID-19) may be attributed to the low precision of the tools used to predict the prognosis of this disease. Objective: To identify the predictors associated with poor clinical outcomes in patients with COVID-19. Methods: Relevant articles from PubMed, Embase, Cochrane, and Web of Science were searched and extracted as of April 5, 2020. Data of interest were collected and evaluated for their compatibility for the meta-analysis. Cumulative calculations to determine the correlation and effect estimates were performed using the Z test. Results: In total, 19 papers recording 1,934 mild and 1,644 severe cases of COVID-19 were included. Based on the initial evaluation, 62 potential risk factors were identified for the meta-analysis. Several comorbidities, including chronic respiratory disease, cardiovascular disease, diabetes mellitus, and hypertension were observed more frequent among patients with severe COVID-19 than with the mild ones. Compared to the mild form, severe COVID-19 was associated with symptoms such as dyspnea, anorexia, fatigue, increased respiratory rate, and high systolic blood pressure. Lower levels of lymphocytes and hemoglobin; elevated levels of leukocytes, aspartate aminotransferase, alanine aminotransferase, blood creatinine, blood urea nitrogen, high-sensitivity troponin, creatine kinase, high-sensitivity C-reactive protein, interleukin 6, D-dimer, ferritin, lactate dehydrogenase, and procalcitonin; and a high erythrocyte sedimentation rate were also associated with severe COVID-19. Conclusion: More than 30 risk factors are associated with a higher risk of severe COVID-19. These may serve as useful baseline parameters in the development of prediction tools for COVID-19 prognosis.",
"keywords": [
"SARS-CoV-2",
"COVID-19",
"prognosis",
"severity",
"clinical outcome"
],
"content": "Introduction\n\nThe coronavirus disease 2019 (COVID-19) pandemic, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), is a global crisis across health, economic, and educational dimensions1,2. The disease has spread rapidly, can cause severe illness, and is characterized by a high mortality rate in certain groups. Mortality is particularly high in the absence of proven effective standard management measures3. One of the problems with the management of this disease is the absence of standardized methods for diagnosis and the inability to estimate prognosis based on clinical features. Certain reports have shown that poor prognostic prediction has correlated with high mortality among patients with COVID-194,5. Among patients with similar clinical characteristics and with similar treatment regiments, there may be a diversity in clinical outcomes6. Therefore, the development and use of an accurate predictor for COVID-19 prognosis will be beneficial for the clinical management of patients with COVID-19, and will help reduce the mortality rate. Successful implementation of such a prediction mechanism could have a large public health impact. Better understanding of clinical progression could also improve public health messaging, particularly as many individuals may consider COVID-19 to not be severe.\n\nPrognostic tools for the prediction of COVID-19 severity in patients have been in development since January 2020. At least nine studies proposed the use of prognostic tools for the prediction of COVID-19 severity7–15. However, a recent systematic review and critical appraisal study evaluated the accuracy of these tools using prediction model risk of bias assessment tool (PROBAST) and reported a high risk of bias16. The establishment of a prediction model for the estimation of disease prognosis may help health workers segregate patients according to prediction status. However, the high risk of bias in these prediction tools might lead to inaccurate prediction of COVID-19 severity. A comprehensive study of the identification of risk factors that might play a significant role in determining the severity of patients with COVID-19 is necessary. We performed a systematic review and meta-analysis to assess the risk factors associated with poor clinical outcomes among patients with COVID-19. To the best of our knowledge, this is the first meta-analysis to assess the comprehensive risk factors that might affect the severity of COVID-19 in patients. The results of our study might serve as preliminary data for the compilation or improvement of the scoring system in the prediction of COVID-19 severity.\n\n\nMethods\n\nWe performed a systematic review and meta-analysis to evaluate potential risk factors that might influence the severity of COVID-19. These risk factors include comorbidities, clinical manifestations, and laboratory findings. Accordingly, we searched the relevant studies from major scientific websites and databases to collect the data of interest, and determined the association and effect estimates by calculating the combined odds ratio (OR) and 95% confidence intervals (95% CI). The protocols for the systematic review and meta-analysis were similar to those used in previous studies17–23, as well as to those recommended by Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA)24.\n\nStudies were included in this review if they met the following inclusion criteria: (1) assessed the clinical manifestations and laboratory findings of patients with mild to severe COVID-19; (2) provided adequate data for the calculation of OR and 95% CI. Review articles, articles with non-standard data presentation, and duplicate publications were excluded.\n\nMajor scientific databases (PubMed, Embase, Cochrane, and Web of Science) were searched for articles as of April 5, 2020. A comprehensive initial search was performed to identify the potential predictors, and a final search was performed to identify the relevant papers that could be included in the meta-analysis. We used the keywords adapted from medical subject headings: [\"COVID-19\" or \"Coronavirus disease-19\" or \"SARS-CoV-2\"] and [\"mild\" or \"severe\" or \"prognosis\" or \"clinical outcome\"] and [\"clinical manifestation\" or \"morbidity\" or \"laboratory findings\"]. Only studies written in English were included. If a duplicate publication was found, the article with the larger sample size was included. We also searched for relevant studies from the reference lists in the articles. During data extraction, the following information of interest was extracted: (1) first author name; (2) publication year; (3) sample size of mild and severe cases, (4) clinical manifestations, (5) morbidities, and (6) laboratory findings. Data extraction was performed by two independent investigators (JKF and MI) using a pilot form.\n\nBefore inclusion in the meta-analysis, the methodological quality of the articles was assessed using the New Castle-Ottawa scale (NOS). NOS scores range from 0 to 9 and consider three items: selection of patients (4 points), comparability of the groups (2 points), and ascertainment of exposure (3 points). Each study was interpreted to be of low quality (for scores ≤ 4), moderate quality (for scores between 5–6), or high quality (for scores ≥ 7)25. Articles with moderate to high quality were included in the analysis. The study assessment was conducted by two independent investigators (MI and YP) using a pilot form. The discrepancies between the findings of the two investigators were solved by consulting with another investigator (JKF).\n\nThe outcome measure of the study was the severity of COVID-19 (mild vs. severe). The risk factors or predictors included three major groups: comorbidities, clinical manifestations, and laboratory parameters. Comorbid factors such as chronic kidney disease, chronic liver disease, chronic respiratory disease, cerebrovascular accident, cardiovascular disease, diabetes mellitus, hypertension, and malignancy were compatible with the analysis. For clinical manifestations, fever, cough, dry cough, expectoration, sore throat, dyspnea, diarrhea, myalgia, nasal congestion, anorexia, abdominal pain, fatigue, dizziness, headache, fever, heart rate, respiratory rate, systolic blood pressure, and diastolic blood pressure were included in this study. Among laboratory characteristics, the presence of leukocytosis, leukocytopenia, anemia, lymphocytopenia; the levels or the counts of white blood cell (WBC), hemoglobin, neutrophil, lymphocyte, monocyte, platelet, activated partial thromboplastin time (aPTT), partial thromboplastin time (PTT), aspartate aminotransferase (AST), alanine aminotransferase (ALT), total bilirubin, albumin, serum creatinine, blood urea nitrogen (BUN), high-sensitivity (Hs)-troponin I, creatine kinase, high-sensitivity C-reactive protein (Hs-CRP), C-reactive protein (CRP) >8 mg/L, interleukin 6 (IL-6), glucose, D-dimer, serum ferritin, sodium, potassium, lactate dehydrogenase, and procalcitonin, CD4 and CD8; erythrocyte sedimentation rate (ESR); elevated IL-16; and elevated ESR were all included.\n\nThe significant risk factors that might govern the severity of COVID-19 were determined by the calculation of a pooled OR and 95% CI. The significance of the pooled ORs was determined using the Z test (p<0.05 was considered statistically significant). Prior to identification of the significant risk factors, data were evaluated for heterogeneity and potential publication bias. The heterogeneity among included studies was evaluated using the Q test. If heterogeneity existed (p<0.10), a random effect model was adopted; if not, a fixed effect model was adopted. Egger’s test and a funnel plot were used to assess the reporting or publication bias (p<0.05 was considered statistically significant). Furthermore, we performed a moderator analysis to identify the independent predictors of poor clinical outcomes among patients with COVID-19. The data were analyzed using Review Manager version 5.3 (Revman Cochrane, London, UK). To prevent analytical errors, statistical analysis was performed by two authors (JKF and MI). The cumulative calculation was presented in a forest plot.\n\n\nResults\n\nOur searches yielded 6,209 potentially relevant studies, of which 6,170 studies were excluded after assessment of the titles and abstracts. Subsequently, further review of the complete texts was performed for 39 potential studies. In the full text review, we excluded 20 studies because they were reviews articles (n = 9), inadequacy of data for the calculation of OR and 95% CI (n = 7), and poor quality (n = 4). Eventually, 19 papers were included in our meta-analysis26–42 The paper selection process adopted in our study is summarized in Figure 1, and the characteristics of studies included in our analysis are outlined in Table 1.\n\nNote: ICU, intensive care unit; CT, computed tomography; NOS, Newcastle Ottawa Scale.\n\nWe found that eight comorbidities, 19 clinical manifestations, and 35 laboratory parameters were available for the meta-analysis (Table 2 and Table 3). Among the comorbid factors, chronic respiratory disease (OR: 2.48; 95% CI: 1.44, 4.27), cardiovascular disease (OR: 1.70; 95% CI: 1.05, 2.78), diabetes mellitus (OR: 2.10; 95% CI: 1.33, 3.34), and hypertension (OR: 2.33; 95% CI: 1.42, 3.81) were associated with a greater risk of severe COVID-19 (Figure 2A-D).\n\nNote, Value, data were presented in number [%] or mean ± SD; NS, number of studies; pE, p Egger; pHet, p heterogeneity; OR, odd ratio; CI, confidence interval; SBP, systolic blood pressure; DBP, diastolic blood pressure.\n\nNote: Value, data were presented in number [%] or mean ± SD; NS, number of studies; pE, p Egger; pHet, p heterogeneity; OR, odd ratio; CI, confidence interval; CBC, complete blood count; WBC, white blood cells; AST, aspartate transaminase; ALT, alanine transaminase; aPTT, activated partial thromboplastin time; PTT, partial thromboplastin time; BUN, blood urea nitrogen; Hs-CRP, high sensitivity C reactive protein; ESR, erythrocyte sedimentation rate; IL, interleukin.\n\nA) Chronic respiratory disease; B) Cardiovascular diease; C) Diabetes mellitus; D) Hypertension.\n\nAmong the clinical manifestations, dyspnea (OR: 3.28; 95% CI: 2.09, 5.15), anorexia (OR: 1.83; 95% CI: 1.00, 3.34), fatigue (OR: 2.00; 95% CI: 1.25, 3.20), and dizziness (OR: 2.67; 95% CI: 1.18, 6.01) were associated with severe COVID-19 (Figure 3A-D). In addition, increased respiratory rate (OR: 2.85; 95% CI: 1.28, 6.33) and increased systolic blood pressure (OR: 1.84; 95% CI: 1.31, 2.60) were also associated with severe COVID-19 (Figure 4A and B). Compared to productive cough, dry cough was associated with a lower risk of severe COVID-19 (OR: 0.66; 95% CI: 0.44, 0.97).\n\nA) Dyspnea; B) Anorexia; C) Fatique; D) Dizziness.\n\nA) Respiratory rate; B) Systolic blood pressure.\n\nAmong laboratory characteristics, severe COVID-19 was associated with elevated WBC count (OR: 4.92; 95% CI: 2.12, 11.31), increased neutrophil count (OR: 5.45; 95% CI: 2.04, 14.54), lymphocytopenia (OR: 3.19; 95% CI: 1.14, 7.07), and decreased hemoglobin levels (OR: 0.76; 95%CI: 0.58, 1.00) (Figure 5A-D). Elevated levels of AST, ALT, and serum creatinine increased the risk for severe manifestations of COVID-19 (ORs 4.91, 3.23, and 2.14, respectively; Figure 6A-C). Elevated levels of BUN (OR: 6.15; 95% CI: 3.05, 12.37), Hs-troponin I (OR: 9.25; 95% CI: 3.51, 24.37), creatine kinase (OR: 2.44; 95% CI: 1.65, 3.62), Hs-CRP (OR: 14.27; 95% CI: 5.13, 39.71), IL-6 (OR: 6.68; 95% CI: 3.20, 13.94), D-dimer (OR: 6.19; 95% CI: 4.22, 9.08), ferritin (OR: 1.96; 95% CI: 1.06, 3.62), lactate dehydrogenase (OR: 8.28; 95% CI: 4.75, 14.46), procalcitonin (OR: 6.62; 95% CI: 3.32, 13.21), ESR (OR: 4.45; 95% CI: 2.56, 7.76), and CRP >8 (OR: 8.34; 95% CI: 1.85, 37.62) were also associated with severe COVID-19 (Figure 7–Figure 9). A low risk of severe COVID-19 was associated with low leukocyte levels (OR: 0.59; 95% CI: 0.41, 0.87) and elevated lymphocyte levels (OR: 0.34; 95% CI: 0.23, 0.50).\n\nA) White blood cells; B) Neutrophil count; C) Lymphocytopenia; D) Hemoglobin.\n\nA forest plot of the association between the risk of severe COVID-19 and the levels of AST (A), ALT (B), and serum creatinine (C).\n\nA forest plot of the association between the risk of severe COVID-19 and the levels of BUN (A), Hs-troponin (B), and creatine kinase (C).\n\nA forest plot of the association between the risk of severe COVID-19 and the levels of CRP (A), Hs-CRP (B), ESR (C), and IL-6 (D).\n\nA forest plot of the association between the risk of severe COVID-19 and the levels of D-dimer (A), serum ferritin (B), lactate dehydrogenase (C), and procalcitonin (D).\n\nHeterogeneity was detected in the data of chronic kidney disease, cerebrovascular disease, cardiovascular disease, diabetes mellitus, hypertension, and malignancy among the comorbid factors analyzed. Therefore, we used the random effect model to analyze the data. The fixed effect model was used to analyze the data on chronic liver disease and chronic respiratory disease, as there was no evidence of heterogeneity. For clinical manifestations, the data on fever, cough, sore throat, dyspnea, diarrhea, anorexia, fatigue, temperature >38°C, respiratory rate, and diastolic blood pressure were analyzed using the random effect model while the rest of clinical manifestation data were analyzed using the fixed effect model.\n\nAmong laboratory parameters, evidence of heterogeneity was found in count of WBC, neutrophil, monocyte, lymphocyte, platelet, CD4, and CD8; the presence of lymphocytopenia and anemia; the levels of AST, ALT, total bilirubin, albumin, aPTT, PTT, serum creatinine, BUN, Hs-Troponin I, creatine kinase, IL-6, Hs-CRP, glucose, D-dimer, sodium, potassium, lactate dehydrogenase, and procalcitonin; elevated CRP; and ESR. Accordingly, the data were analyzed using the random effect model. The data for the remaining parameters were analyzed using the fixed effect model.\n\nWe used Egger's test to assess the potential publication bias. Our cumulative calculation revealed that reporting or publication bias (p<0.05) existed with respect to chronic liver disease, expectoration, myalgia, abdominal pain, heart rate, leukocytosis, elevated ESR, and elevated IL-6 levels.\n\n\nDiscussion\n\nOur data suggest that comorbidities, such as chronic respiratory disease, cardiovascular disease, diabetes, and hypertension, were associated with a higher risk of severe COVID-19, among which, hypertension was the strongest risk factor. These results are consistent with those of previous meta-analyses43,44 that indicated that chronic respiratory disease, cardiovascular disease, diabetes, and hypertension are significantly associated with higher COVID-19 mortality. Hypertension and diabetes are also associated with higher mortality among patients with dengue fever, West Nile virus infection, Zika virus infection, and yellow fever45. To date, no study has reported details of the primary mechanism underlying the association between severe COVID-19 and comorbid factors. However, immune responses might be the most crucial factor underlying this association. Patients with comorbidities such as cardiovascular disease, chronic respiratory disease, hypertension, and diabetes were observed to have a lower immunity status than healthy individuals46–48. Since COVID-19 primarily affects the respiratory tract49, patients with chronic respiratory diseases might be at a higher risk of contracting severe COVID-19. In addition, endothelial dysfunction might also play a pivotal role50.\n\nCOVID-19 is a novel disease, and the immune response of this disease is not completely understood. Our data suggest that elevated leukocyte and neutrophil levels and reduced lymphocyte levels are associated with severe COVID-19. In other viral infections, such as influenza, elevated leukocyte and neutrophil levels serve as important predictors of disease severity51. The role of leukocytes in the pathogenesis of COVID-19 is conflicting. In most cases, viral infections have been observed to cause leukopenia52. Furthermore, a study also reported that leukopenia was observed at a significantly higher frequency among COVID-19 patients than among non-COVID-19 patients53. However, in our present study, we did not compare COVID-19 and non-COVID-19 patients. The major factor that seemed to affect our findings was the occurrence of cytokine storm in patients. In COVID-19, there is an immune system overreaction, which results in a cytokine storm. In this condition, leukocytes might be over-activated, which might lead to the release of high levels of cytokines54. Consistent with our data, a study has confirmed that cytokine storm is significantly associated with severe COVID-1955. The theory underlying the role of neutrophils in COVID-19, as reported in our study, remains unclear. The speculations might be attributed to the involvement of neutrophil extracellular traps (NETs). While no study has assessed the precise role of NETs in COVID-19 pathogenesis, certain researchers speculate that SARS-CoV-2 might stimulate neutrophils to produce NETs, similar to several other viral pathogens56. Furthermore, this might lead to neutrophil infiltration in pulmonary capillaries, organ damage, and the development of acute respiratory distress syndrome57.\n\nLow lymphocyte levels were observed in patients with severe COVID-19 compared with those with mild COVID-19. In the context of the immunological mechanism, our results might be contradictory. Lymphocyte subsets are known to play an important role in the action against bacterial, viral, fungal, and parasitic infections58; therefore, the levels of circulating lymphocytes should increase. The immunological response in COVID-19 is unique and remains unclear. However, certain propositions might help describe our findings. First, coronaviruses infect human cells through ACE2 receptors59. Since ACE2 receptors are also expressed by lymphocytes60, the coronaviruses may enter lymphocytes and induce apoptosis. Second, the feedback mechanism between pro-inflammatory cytokines (such as IL-6) and lymphocytes might also explain our results. A study revealed that elevation in the levels of pro-inflammatory cytokines correlated with reduction in the levels of lymphocytes61. Moreover, our findings also confirmed the significant elevation in the levels of IL-6. Third, ACE2 receptors are expressed by cells from various organs, including the thymus and spleen62. As coronaviruses infect human cells through the ACE2 receptors, the spleen and thymus might also be damaged in patients with COVID-19, which would lead to lower levels of lymphocyte production. Fourth, lymphocyte proliferation requires a balanced metabolism, and metabolic disorders such as hyperlactic acidemia have been reported to disturb lymphocyte proliferation63. Hyperlactic acidemia has been observed in patients with severe COVID-1964.\n\nThe studies included in this systematic review also suggest that the levels of D-dimer were significantly higher in patients with severe COVID-19. Coagulation in patients with COVID-19 has been a major concern, and the lack of reliable data and meta-analyses prevents a holistic comparison. Certain infectious diseases that cause abnormal coagulation have been associated with poor clinical outcomes65. The theory behind this mechanism is not understood clearly. It is widely known that ACE2 receptors are important for the infection of host cells by SARS-CoV-2, and ACE2 receptors are expressed in various cells in the human body, including endothelial cells66. Consequently, a massive inflammatory reaction may occur in endothelial cells owing to SARS-CoV-2 infection67, which may lead to increased coagulation, disseminated intravascular coagulation68, and increased fibrin degradation69. High fibrin degradation leads to elevated levels of fibrinogen and D-dimer70, which might also explain the occurrence of venous thromboembolism in critical patients of COVID-1971. In addition, a study with a short follow-up period also reported the existence of a dynamic correlation between the D-dimer levels and the severity of COVID-1972. Furthermore, pulmonary embolism and deep vein thrombosis were also observed in patients with severe COVID-1973,74, which suggests that D-dimer might play a prominent role in governing the severity of COVID-19 patients.\n\nWe also observed that inflammatory markers, including elevated levels of CRP, ESR, and IL-6, were found both in patients with severe and mild COVID-19, with a significant increase detected in patients with severe COVID-19. Other variables associated with adverse outcomes, such as ferritin, lactate dehydrogenase, and procalcitonin levels, were found to be elevated predominantly in patients with severe COVID-19. Our findings were consistent with those of a previous meta-analysis75, and indicated that high levels of CRP, lactate dehydrogenase, and ESR were associated with adverse outcomes in COVID-19. Another meta-analysis had also confirmed that elevated levels of IL-6 were observed in patients with COVID-19 who exhibited poor clinical outcomes76. Therefore, the levels of CRP, ESR, IL-6, ferritin, procalcitonin, and lactate dehydrogenase can serve as potential markers for the evaluation of COVID-19 prognosis.\n\nThe high mortality rate and treatment failure in patients with COVID-19 can be attributed to the fact that COVID-19 affects multiple organs, including the lung, heart, kidney, and liver77. Our data suggest that elevated levels of urea and creatinine, and not chronic kidney disease, were associated with severe COVID-19, which indicates that acute inflammation might be caused by SARS-CoV-2 infection. Previous meta-analyses have also reported findings consistent with our results78,79. Moreover, anatomical studies have reported significant renal inflammation in patients with severe COVID-1975,80,81. There might be two mechanisms by which SARS-CoV-2 induces renal inflammation. First, SARS-CoV-2 might directly infect renal tubular epithelial cells and podocytes through ACE2 receptors, which facilitates the targeted infection of certain cells by the virus. Consequently, acute tubular necrosis, podocytopathy, microangiopathy, and collapsing glomerulopathy might occur owing to the massive inflammation in renal tubular epithelial cells and podocytes82,83. Second, the binding between SARS-CoV-2 and ACE2 receptors might activate angiotensin II and induce cytokine production, which may lead to hypercoagulopathy and microangiopathy, and eventually cause renal hypoxia84,85.\n\nConversely, with respect to liver function, we observed that the levels of liver enzymes were higher in patients with severe COVID-19. Previous studies in this context have elucidated that ACE2 receptors are highly expressed in bile duct cells; therefore, infection of these cells by coronaviruses might lead to abnormalities in the levels of liver enzymes86. However, a recent anatomical study on liver biopsy specimens from patients with severe COVID-19 revealed that moderate microvascular steatosis and mild lobular and portal activities were observed87. These data suggest that it cannot be determined clearly whether the elevated levels of liver enzymes in patients with severe COVID-19 are caused by direct infection or by drug-induced liver injury. Therefore, further studies are required to elucidate the precise mechanism underlying the elevation of liver enzymes levels in patients with severe COVID-19.\n\nMeta-analyses on this topic have been performed previously43,44,75,76,88–91. However, compared to previous studies, our study has the following strengths. The previous studies only reported limited factors, such as clinical manifestations43,88,90,91, laboratory findings76,89, or a combination of only clinical manifestations and laboratory findings75. In our study, we included all comorbidities, clinical manifestations, and laboratory characteristics. Additionally, compared to previous studies, this study has a larger sample size; the data on 1,934 patients with mild and 1,644 patients with severe COVID-19 treated across 19 hospitals were retrieved. However, this study also has certain limitations. Certain crucial factors that might play an important role in the pathogenesis of COVID-19, including secondary infection, treatment, and immunological status were not controlled for. Our current findings should be interpreted with caution because the majority of studies included were cross-sectional, and the samples corresponding to the data analyzed originated only in China. Longitudinal studies may reveal more long-term impacts of SARS-CoV-2 infection92.\n\n\nConclusion\n\nCOVID-19 is an emergent infectious disease, and the major problem associated with it is the unknown pattern of disease development. We identified 34 factors that are associated with severe COVID-19. This might improve our understanding of COVID-19 progression and provide baseline data to compile or improve the prediction models for the estimation of COVID-19 prognosis.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\nFigshare: PRISMA checklist for ‘Predictors of COVID-19 severity: a systematic review and meta-analysis’, https://doi.org/10.6084/m9.figshare.12813683.v193\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "References\n\nAcikgoz O, Gunay A: The early impact of the Covid-19 pandemic on the global and Turkish economy. Turk J Med Sci. 2020; 50(SI-1): 520–526. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNicola M, Alsafi Z, Sohrabi C, et al.: The socio-economic implications of the coronavirus pandemic (COVID-19): A review. Int J Surg. 2020; 78: 185–193. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAl-Tawfiq JA, Leonardi R, Fasoli G, et al.: Prevalence and fatality rates of COVID-19: What are the reasons for the wide variations worldwide? Travel Med Infect Dis. 2020; 35: 101711. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShojaee S, Pourhoseingholi MA, Ashtari S, et al.: Predicting the mortality due to Covid-19 by the next month for Italy, Iran and South Korea; a simulation study. Gastroenterol Hepatol Bed Bench. 2020; 13(2): 177–179. PubMed Abstract | Free Full Text\n\nWang L, Li J, Guo S, et al.: Real-time estimation and prediction of mortality caused by COVID-19 with patient information based algorithm. Sci Total Environ. 2020; 727: 138394. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSiordia JA: Epidemiology and clinical features of COVID-19: A review of current literature. J Clin Virol. 2020; 127: 104357. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBai X, Fang C, Zhou Y, et al.: Predicting COVID-19 malignant progression with AI techniques. 2020. Publisher Full Text\n\nXie J, Hungerford D, Chen H, et al.: Development and external validation of a prognostic multivariable model on admission for hospitalized patients with COVID-19. Korean J Radiol. 2020; 21(8): 1007–1017. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCaramelo F, Ferreira N, Oliveiros B: Estimation of risk factors for COVID-19 mortality-preliminary results. MedRxiv. 2020. Publisher Full Text\n\nQi X, Jiang Z, Yu Q, et al.: Machine learning-based CT radiomics model for predicting hospital stay in patients with pneumonia associated with SARS-CoV-2 infection: A multicenter study. Ann Transl Med. 2020; 8(14): 859. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYan L, Zhang HT, Xiao Y, et al.: Prediction of criticality in patients with severe Covid-19 infection using three clinical features:a machine learning-based prognostic model with clinical data in Wuhan. MedRxiv. 2020. Publisher Full Text\n\nYuan M, Yin W, Tao Z, et al.: Association of radiologic findings with mortality of patients infected with 2019 novel coronavirus in Wuhan, China. PLoS One. 2020; 15(3): e0230548. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLu J, Hu S, Fan R, et al.: ACP risk grade: a simple mortality index for patients with confirmed or suspected severe acute respiratory syndrome coronavirus 2 disease (COVID-19) during the early stage of outbreak in Wuhan, China. MedRxiv. 2020. Publisher Full Text\n\nGong J, Ou J, Qiu X, et al.: A tool to early predict severe 2019-novel coronavirus pneumonia (COVID-19):a multicenter study using the risk nomogram in Wuhan and Guangdong, China. Clin Infect Dis. 2020; 71(15): 833–840. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShi Y, Yu X, Zhao H, et al.: Host susceptibility to severe COVID-19 and establishment of a host risk score: findings of 487 cases outside Wuhan. Crit Care. 2020; 24(1): 108. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWynants L, Van Calster B, Collins GS, et al.: Prediction models for diagnosis and prognosis of covid-19 infection: systematic review and critical appraisal. BMJ. 2020; 369: m1328. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFajar J, Mahdi B, Heriansyah T, et al.: Length of stay and major adverse cardiac events: Comparison between percutaneous coronary intervention and thrombolytic therapy in patients with ST-elevation myocardial infarction Ιmplications for cost effectiveness. Archives of Hellenic Medicine/Arheia Ellenikes Iatrikes. 2019; 36(4). Reference Source\n\nFajar JK, Andalas M, Harapan H: Comparison of Apgar scores in breech presentations between vaginal and cesarean delivery. Tzu-Chi Medical Journal. 2017; 29(1): 24–29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFajar JK, Harapan H: Socioeconomic and attitudinal variables associated with acceptance and willingness to pay towards dengue vaccine :a systematic review. Arch Clin Infect Dis. 2017; 12(3): e13914. Publisher Full Text\n\nFajar JK, Mahendra AI, Tamara F, et al.: The association between complete blood count and the risk of coronary heart disease. Turkiye Klinikleri J Med Sci. 2019; 39(1): 56–64. Publisher Full Text\n\nFajar JK, Taufan T, Syarif M, et al.: Hip geometry and femoral neck fractures: A meta-analysis. J Orthop Translat. 2018; 13: 1–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPrihatiningsih S, Fajar JK, Tamara F, et al.: Risk factors of tuberculosis infection among health care workers: a meta-analysis. Indian J Tuberc. 2020; 67(1): 121–129. PubMed Abstract | Publisher Full Text\n\nRohman MS, Purnamasari Y, Ilmawan M, et al.: Comparison of major bleeding in patients with acute coronary syndrome that underwent coronary artery bypass grafting treated with clopidogrel or ticagrelor: a systematic review and meta-analysis [version 1; peer review: 1 approved, 1 approved with reservations]. F1000Res. 2020; 9(99): 99. Publisher Full Text\n\nMcInnes MDF, Moher D, Thombs BD, et al.: Preferred Reporting Items for a Systematic Review and Meta-analysis of Diagnostic Test Accuracy Studies:The PRISMA-DTA Statement. JAMA. 2018; 319(4): 388–396. PubMed Abstract | Publisher Full Text\n\nStang A: Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol. 2010; 25(9): 603–5. PubMed Abstract | Publisher Full Text\n\nBai T, Tu S, Wei Y, et al.: Clinical and laboratory factors predicting the prognosis of patients with COVID-19: an analysis of 127 patients in Wuhan, China. China (2/26/2020). 2020. Publisher Full Text\n\nCai Q, Huang D, Ou P, et al.: COVID-19 in a designated infectious diseases hospital outside Hubei Province, China. Allergy. 2020; 75(7): 1742–1752. PubMed Abstract | Publisher Full Text\n\nChen G, Wu D, Guo W, et al.: Clinical and immunological features of severe and moderate coronavirus disease 2019. J Clin Invest. 2020; 130(5): 2620–2629. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen X, Zheng F, Qing Y, et al.: Epidemiological and clinical features of 291 cases with coronavirus disease 2019 in areas adjacent to Hubei, China: a double-center observational study. MedRxiv. 2020. Publisher Full Text\n\nChen X, Ling J, Mo P, et al.: Restoration of leukomonocyte counts is associated with viral clearance in COVID-19 hospitalized patients. MedRxiv. 2020. Publisher Full Text\n\nDuan Q, Guo G, Ren Y, et al.: Treatment Outcomes, Influence Factors of 116 Hospitalized COVID-19 Patients with Longer/Prolonged Treatment Course in Wuhan, China. Influence Factors. 2020; 116.\n\nGao Y, Li T, Han M, et al.: Diagnostic utility of clinical laboratory data determinations for patients with the severe COVID-19. J Med Virol. 2020; 92(7): 791–796. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGuan WJ, Ni ZY, Hu Y, et al.: Clinical Characteristics of Coronavirus Disease 2019 in China. N Engl J Med. 2020; 382(18): 1708–1720. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuang C, Wang Y, Li X, et al.: Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. Lancet. 2020; 395(10223): 497–506. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJian-ya G: Clinical characteristics of 51 patients discharged from hospital with COVID-19 in Chongqing, China. medRxiv. 2020. Publisher Full Text\n\nLiu T, Zhang J, Yang Y, et al.: The potential role of IL-6 in monitoring coronavirus disease 2019. 2020. Publisher Full Text\n\nShi Q, Zhao K, Yu J, et al.: Clinical characteristics of 101 non-surviving hospitalized patients with COVID-19: A single center, retrospective study. medRxiv. 2020. Reference Source\n\nWang D, Hu B, Hu C, et al.: Clinical Characteristics of 138 Hospitalized Patients With 2019 Novel Coronavirus-Infected Pneumonia in Wuhan, China. JAMA. 2020; 323(11): 1061–1069. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang H, Luo S, Shen Y, et al.: Multiple enzyme release, inflammation storm and hypercoagulability are prominent indicators for disease progression in COVID-19: a multi-centered, correlation study with CT imaging score. 2020. Publisher Full Text\n\nXu Y, Li Y-r, Zeng Q, et al.: Clinical characteristics of SARS-CoV-2 pneumonia compared to controls in Chinese Han population. medRxiv. 2020. Publisher Full Text\n\nZhang JJ, Dong X, Cao YY, et al.: Clinical characteristics of 140 patients infected with SARS-CoV-2 in Wuhan, China. Allergy. 2020; 75(7): 1730–1741. PubMed Abstract | Publisher Full Text\n\nZhou F, Yu T, Du R, et al.: Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study. Lancet. 2020; 395(10229): 1054–1062. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi J, He X, Zhang W, et al.: Meta-analysis investigating the relationship between clinical features, outcomes, and severity of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pneumonia. Am J Infect Control. 2020; S0196-6553(20)30369-2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang J, Zheng Y, Gou X, et al.: Prevalence of comorbidities and its effects in patients infected with SARS-CoV-2: a systematic review and meta-analysis. Int J Infect Dis. 2020; 94: 91–95. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBadawi A, Velummailum R, Ryoo SG, et al.: Prevalence of chronic comorbidities in dengue fever and West Nile virus: A systematic review and meta-analysis. PLoS One. 2018; 13(7): e0200200. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVasdev S, Stuckless J, Richardson V: Role of the immune system in hypertension: modulation by dietary antioxidants. Int J Angiol. 2011; 20(4): 189–212. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFerlita S, Yegiazaryan A, Noori N, et al.: Type 2 Diabetes Mellitus and Altered Immune System Leading to Susceptibility to Pathogens, Especially Mycobacterium tuberculosis. J Clin Med. 2019; 8(12): 2219. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLazzerini PE, Hamilton RM, Boutjdir M: Editorial: Cardioimmunology: Inflammation and Immunity in Cardiovascular Disease. Front Cardiovasc Med. 2019; 6: 181. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPerlman S, Dandekar AA: Immunopathogenesis of coronavirus infections: implications for SARS. Nat Rev Immunol. 2005; 5(12): 917–27. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGreen SJ: Covid-19 accelerates endothelial dysfunction and nitric oxide deficiency. Microbes Infect. 2020; 22(4–5): 149–150. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRussell CD, Parajuli A, Gale HJ, et al.: The utility of peripheral blood leucocyte ratios as biomarkers in infectious diseases: A systematic review and meta-analysis. J Infect. 2019; 78(5): 339–348. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPascutti MF, Erkelens MN, Nolte MA: Impact of Viral Infections on Hematopoiesis: From Beneficial to Detrimental Effects on Bone Marrow Output. Front Immunol. 2016; 7: 364. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi Y, Wang H, Wang F, et al.: Comparison of Hospitalized Patients with pneumonia caused by COVID-19 and influenza A in children under 5 years. Int J Infect Dis. 2020; 98: 80–83. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCoperchini F, Chiovato L, Croce L, et al.: The cytokine storm in COVID-19: An overview of the involvement of the chemokine/chemokine-receptor system. Cytokine Growth Factor Rev. 2020; 53: 25–32. PubMed Abstract | Publisher Full Text | Free Full Text\n\nQuartuccio L, Sonaglia A, McGonagle D, et al.: Profiling COVID-19 pneumonia progressing into the cytokine storm syndrome: results from a single Italian Centre study on tocilizumab versus standard of care. J Clin Virol. 2020; 129: 104444. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPapayannopoulos V: Neutrophil extracellular traps in immunity and disease. Nat Rev Immunol. 2018; 18(2): 134–147. PubMed Abstract | Publisher Full Text\n\nMikacenic C, Moore R, Dmyterko V, et al.: Neutrophil extracellular traps (NETs) are increased in the alveolar spaces of patients with ventilator-associated pneumonia. Crit Care. 2018; 22(1): 358. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKoyasu S, Moro K: Role of innate lymphocytes in infection and inflammation. Front Immunol. 2012; 3: 101. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu M, Wang T, Zhou Y, et al.: Potential Role of ACE2 in Coronavirus Disease 2019 (COVID-19) Prevention and Management. J Transl Int Med. 2020; 8(1): 9–19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTo KF, Lo AW: Exploring the pathogenesis of severe acute respiratory syndrome (SARS): the tissue distribution of the coronavirus (SARS-CoV) and its putative receptor, angiotensin-converting enzyme 2 (ACE2). J Pathol. 2004; 203(3): 740–3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShachar I, Karin N: The dual roles of inflammatory cytokines and chemokines in the regulation of autoimmune diseases and their clinical implications. J Leukoc Biol. 2013; 93(1): 51–61. PubMed Abstract | Publisher Full Text\n\nHamming I, Timens W, Bulthuis M, et al.: Tissue distribution of ACE2 protein, the functional receptor for SARS coronavirus. A first step in understanding SARS pathogenesis. J Pathol. 2004; 203(2): 631–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFischer K, Hoffmann P, Voelkl S, et al.: Inhibitory effect of tumor cell-derived lactic acid on human T cells. Blood. 2007; 109(9): 3812–9. PubMed Abstract | Publisher Full Text\n\nChhetri S, Khamis F, Pandak N, et al.: A fatal case of COVID-19 due to metabolic acidosis following dysregulate inflammatory response (cytokine storm). IDCases. 2020; 21: e00829. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLevi M, Keller TT, van Gorp E, et al.: Infection and inflammation and the coagulation system. Cardiovasc Res. 2003; 60(1): 26–39. PubMed Abstract | Publisher Full Text\n\nLovren F, Pan Y, Quan A, et al.: Angiotensin converting enzyme-2 confers endothelial protection and attenuates atherosclerosis. Am J Physiol Heart Circ Physiol. 2008; 295(4): H1377–84. PubMed Abstract | Publisher Full Text\n\nVarga Z, Flammer AJ, Steiger P, et al.: Endothelial cell infection and endotheliitis in COVID-19. Lancet. 2020; 395(10234): 1417–1418. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarietta M, Ageno W, Artoni A, et al.: COVID-19 and haemostasis: a position paper from Italian Society on Thrombosis and Haemostasis (SISET). Blood Transfus. 2020; 18(3): 167–169. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBecker RC: COVID-19 update: Covid-19-associated coagulopathy. J Thromb Thrombolysis. 2020; 50(1): 54–67. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMatsuo T, Kobayashi H, Kario K, et al.: Fibrin D-dimer in thrombogenic disorders. Semin Thromb Hemost. 2000; 26(1): 101–7. PubMed Abstract | Publisher Full Text\n\nKhan IH, Savarimuthu S, Leung MST, et al.: The need to manage the risk of thromboembolism in COVID-19 patients. J Vasc Surg. 2020; 72(3): 799–804. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGarcia-Olive I, Sintes H, Radua J, et al.: D-dimer in patients infected with COVID-19 and suspected pulmonary embolism. Respir Med. 2020; 169: 106023. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUllah W, Saeed R, Sarwar U, et al.: COVID-19 complicated by Acute Pulmonary Embolism and Right-Sided Heart Failure. JACC Case Rep. 2020; 2(9): 1379–1382. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNahum J, Morichau-Beauchant T, Daviaud F, et al.: Venous Thrombosis Among Critically Ill Patients With Coronavirus Disease 2019 (COVID-19). JAMA Netw Open. 2020; 3(5): e2010478. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRodriguez-Morales AJ, Cardona-Ospina JA, Gutierrez-Ocampo E, et al.: Clinical, laboratory and imaging features of COVID-19: A systematic review and meta-analysis. Travel Med Infect Dis. 2020; 34: 101623. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAziz M, Fatima R, Assaly R: Elevated interleukin-6 and severe COVID-19: A meta-analysis. J Med Virol. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZaim S, Chong JH, Sankaranarayanan V, et al.: COVID-19 and Multiorgan Response. Curr Probl Cardiol. 2020; 45(8): 100618. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang X, Jin Y, Li R, et al.: Prevalence and impact of acute renal impairment on COVID-19: a systematic review and meta-analysis. Crit Care. 2020; 24(1): 356. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen YT, Shao SC, Hsu CK, et al.: Incidence of acute kidney injury in COVID-19 infection: a systematic review and meta-analysis. Crit Care. 2020; 24(1): 346. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRossi GM, Delsante M, Pilato FP, et al.: Kidney biopsy findings in a critically ill COVID-19 patient with dialysis-dependent acute kidney injury: a case against \"SARS-CoV-2 nephropathy\". Kidney Int Rep. 2020; 5(7): 1100–1105. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHarapan H, Itoh N, Yufika A, et al.: Coronavirus disease 2019 (COVID-19): A literature review. J Infect Public Health. 2020; 13(5): 667–673. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBatlle D, Soler MJ, Sparks MA, et al.: Acute kidney injury in COVID-19: emerging evidence of a distinct pathophysiology. J Am Soc Nephrol. 2020; 31(7): 1380–1383. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNasr SH, Kopp JB: COVID-19-Associated Collapsing Glomerulopathy: An Emerging Entity. Kidney Int Rep. 2020; 5(6): 759–761. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHenry BM, Vikse J, Benoit S, et al.: Hyperinflammation and derangement of renin-angiotensin-aldosterone system in COVID-19: A novel hypothesis for clinically suspected hypercoagulopathy and microvascular immunothrombosis. Clin Chim Acta. 2020; 507: 167–173. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKai H, Kai M: Interactions of coronaviruses with ACE2, angiotensin II, and RAS inhibitors-lessons from available evidence and insights into COVID-19. Hypertens Res. 2020; 43(7): 648–654. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChai X, Hu L, Zhang Y, et al.: Specific ACE2 expression in cholangiocytes may cause liver damage after 2019-nCoV infection. bioRxiv. 2020. Publisher Full Text\n\nXu Z, Shi L, Wang Y, et al.: Pathological findings of COVID-19 associated with acute respiratory distress syndrome. Lancet Respir Med. 2020; 8(4): 420–422. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPark JH, Jang W, Kim SW, et al.: The Clinical Manifestations and Chest Computed Tomography Findings of Coronavirus Disease 2019 (COVID-19) Patients in China: A Proportion Meta-Analysis. Clin Exp Otorhinolaryngol. 2020; 13(2): 95–105. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang ZL, Hou YL, Li DT, et al.: Laboratory findings of COVID-19: a systematic review and meta-analysis. Scand J Clin Lab Invest. 2020; 1–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJain V, Yuan JM: Predictive symptoms and comorbidities for severe COVID-19 and intensive care unit admission: a systematic review and meta-analysis. Int J Public Health. 2020; 1–14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi LQ, Huang T, Wang YQ, et al.: COVID-19 patients' clinical characteristics, discharge rate, and fatality rate of meta-analysis. J Med Virol. 2020; 92(6): 577–583. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTenforde MW, Kim SS, Lindsell CJ, et al.: Symptom Duration and Risk Factors for Delayed Return to Usual Health Among Outpatients with COVID-19 in a Multistate Health Care Systems Network - United States, March-June 2020. MMWR Morb Mortal Wkly Rep. 2020; 69(30): 993–998. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMudatsir M, Fajar J: PRISMA CHECKLIST FOR \"Predictors of COVID-19 severity: a systematic review and meta-analysis\". figshare. Media. 2020. http://www.doi.org/10.6084/m9.figshare.12813683.v1"
}
|
[
{
"id": "71054",
"date": "21 Sep 2020",
"name": "Morteza Arab-Zozani",
"expertise": [
"Reviewer Expertise Systematic review and meta-analysis in health and medical intervention"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this meta-analysis, you investigated the predictors of COVID-19 severity through the literature. You considered a topic of interest and provided a well-written manuscript. However, there are some things that will improve your reporting.\nAbstract, method section, please insert detail about critical/quality appraisal of the included studies.\n\nAbstract, method section, line 1, please remove \" and extracted\" from the text. It maybe causes a misunderstanding between this step and the data extraction step.\n\nMethod section, please remove line five. \"the protocols for the ...\". Mentioning the PRISMA is enough.\n\nMethod section, eligibility criteria, (2) please mention the type of data for adequate data. what is adequate data?\n\nMethod section, search strategy, why is Scopus not searched? You may have missed some articles that are only indexed in Scopus.\n\nMethod section, search strategy, this sentence not related to this section. If you limit the search to EN publication then you need to change the verb. If not this sentence related to inclusion criteria.\n\nMethod section, search strategy, based on PRISMA, add at least one search strategy for one database as a supplement.\n\nMethod section, data extraction, please added the country of origin for each study. The predictors may be different from one setting to another setting.\n\nMethod section, data extraction, please add details about how resolved disagreement between reviewers.\n\nMethod section, how did you handle the publication bias?\n\nResult section, there is some problem in figure 1. Please fill it considering other related studies. The number for \"record screened\" is incorrect.\n\nResult section, table 1, all studies are from China. If all studies are from China it is better to change the title. these are a predictor of severity in China. In my opinion, this is a limitation of your study.\n\nCheers\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Partly\n\nIs the statistical analysis and its interpretation appropriate? Yes\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes",
"responses": [
{
"c_id": "6089",
"date": "06 Jan 2021",
"name": "Mudatsir Mudatsir",
"role": "Author Response",
"response": "Response to comments from reviewers: Reviewer 1# 1. In this meta-analysis, you investigated the predictors of COVID-19 severity through the literature. You considered a topic of interest and provided a well-written manuscript. However, there are some things that will improve your reporting. Abstract, method section, please insert detail about critical/quality appraisal of the included studies. Response: The description of the quality assessment of included papers has been added in the method of abstract. 2. Abstract, method section, line 1, please remove \" and extracted\" from the text. It maybe causes a misunderstanding between this step and the data extraction step. Response: We have removed “and extracted”. 3. Method section, please remove line five. \"the protocols for the ...\". Mentioning the PRISMA is enough. Response: PRISMA checklist may be interpreted as the general guideline in meta-analysis. The specific protocols may differ among meta-analysis studies; for example, the protocols of meta-analysis in gene polymorphism may differ from the protocols of meta-analysis in risk factors identification. In our manuscript, we referred to previous meta-analysis studies in the context of risk factors identification. 4. Method section, eligibility criteria, (2) please mention the type of data for adequate data. what is adequate data? Response: The additional information related to adequate data has been provided. 5. Method section, search strategy, why is Scopus not searched? You may have missed some articles that are only indexed in Scopus. Response: We also performed the searching strategy in Scopus as of 5 April 2020, however, we did not find additional articles. 6. Method section, search strategy, this sentence not related to this section. If you limit the search to EN publication then you need to change the verb. If not this sentence related to inclusion criteria. Response: English publication language has been added to eligibility criteria. 7. Method section, search strategy, based on PRISMA, add at least one search strategy for one database as a supplement. Response: The additional database has been added as the additional database. 8. Method section, data extraction, please added the country of origin for each study. The predictors may be different from one setting to another setting. Response: Country of origin has been added in data extraction. 9. Method section, data extraction, please add details about how resolved disagreement between reviewers. Response: The additional sentence has been added to describe how to resolve the disagreement. 10. Method section, how did you handle the publication bias? Response: The assessment of publication bias has been described in statistical analysis using Egger test. In the results, we presented in Tables 2 & 3. 11. Result section, there is some problem in figure 1. Please fill it considering other related studies. The number for \"record screened\" is incorrect. Response: In Figure 1, we used the template from PRISMA for the article selection pathway. 12. Result section, table 1, all studies are from China. If all studies are from China it is better to change the title. these are a predictor of severity in China. In my opinion, this is a limitation of your study. Response: We tried to search articles in all regions, however, at the time frame of our searching, we only found articles in China."
}
]
},
{
"id": "72568",
"date": "02 Nov 2020",
"name": "Annelies Wilder-Smith",
"expertise": [
"Reviewer Expertise COVID-19",
"Zika and dengue"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe strength of this paper is the meta-analysis in terms of effect estimates. The weakness is the focus of data from China, while we should learn more from global data including the comparison between HIC and LMIC.\nIn China, severity was also found to correlate with the force of infection, eg those in high transmission areas had more severe disease outcomes than those from lower transmission areas in China, see: Exposure to SARS-CoV-2 in a high transmission setting increases the risk of severe COVID-19 compared with exposure to a low transmission setting?\nChen D, Hu C, Su F, Song Q, Wang Z. J Travel Med. 2020 Aug 20;27(5):taaa094. doi: 10.1093/jtm/taaa094.1\nThe authors highlight the need for a scoring system for the prediction of severity. There is another reason why it is important to identify risk factors for severe disease: to guide prioritization of high risk target populations for vaccination\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Yes\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes",
"responses": [
{
"c_id": "6090",
"date": "06 Jan 2021",
"name": "Mudatsir Mudatsir",
"role": "Author Response",
"response": "Response to comments from reviewers: Reviewer 2# 1. The strength of this paper is the meta-analysis in terms of effect estimates. The weakness is the focus of data from China, while we should learn more from global data including the comparison between HIC and LMIC. In China, severity was also found to correlate with the force of infection, eg those in high transmission areas had more severe disease outcomes than those from lower transmission areas in China, see: Exposure to SARS-CoV-2 in a high transmission setting increases the risk of severe COVID-19 compared with exposure to a low transmission setting? Chen D, Hu C, Su F, Song Q, Wang Z. J Travel Med. 2020 Aug 20;27(5):taaa094. doi: 10.1093/jtm/taaa094.1 Response: The additional limitation has been added, as suggested 2. The authors highlight the need for a scoring system for the prediction of severity. There is another reason why it is important to identify risk factors for severe disease: to guide prioritization of high risk target populations for vaccination. Response: The additional clinical implication has been added, as suggested."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1107
|
https://f1000research.com/articles/10-3/v1
|
05 Jan 21
|
{
"type": "Research Article",
"title": "Energy-dependent protein folding: modeling how a protein folding machine may work",
"authors": [
"Harutyun Sahakyan",
"Karen Nazaryan",
"Arcady Mushegian",
"Irina Sorokina",
"Harutyun Sahakyan",
"Karen Nazaryan",
"Arcady Mushegian"
],
"abstract": "Background: Proteins fold robustly and reproducibly in vivo, but many cannot fold in vitro in isolation from cellular components. Despite the remarkable progress that has been achieved by the artificial intelligence approaches in predicting the protein native conformations, the pathways that lead to such conformations, either in vitro or in vivo, remain largely unknown. The slow progress in recapitulating protein folding pathways in silico may be an indication of the fundamental deficiencies in our understanding of folding as it occurs in nature. Here we consider the possibility that protein folding in living cells may not be driven solely by the decrease in Gibbs free energy and propose that protein folding in vivo should be modeled as an active energy-dependent process. The mechanism of action of such a protein folding machine might include direct manipulation of the peptide backbone.\nMethods: To show the feasibility of a protein folding machine, we conducted molecular dynamics simulations that were augmented by the application of mechanical force to rotate the C-terminal amino acid while simultaneously limiting the N-terminal amino acid movements.\nResults: Remarkably, the addition of this simple manipulation of peptide backbones to the standard molecular dynamics simulation indeed facilitated the formation of native structures in five diverse alpha-helical peptides. Steric clashes that arise in the peptides due to the forced directional rotation resulted in the behavior of the peptide backbone no longer resembling a freely jointed chain.\nConclusions: These simulations show the feasibility of a protein folding machine operating under the conditions when the movements of the polypeptide backbone are restricted by applying external forces and constraints. Further investigation is needed to see whether such an effect may play a role during co-translational protein folding in vivo and how it can be utilized to facilitate folding of proteins in artificial environments.",
"keywords": [
"Protein folding",
"ribosome function",
"chaperone",
"computer modeling",
"molecular dynamics",
"energy-dependent protein folding",
"co-translational protein folding",
"nascent peptide rotation",
"peptide backbone manipulation",
"protein folding machine"
],
"content": "Introduction\n\nOnce they are synthesized in a living cell, the majority of proteins rapidly attain their distinctive biologically active three-dimensional structures, called native conformations. These conformations are robustly achieved in vivo via a folding process that involves interactions of the folding chain with molecular chaperones and other maturation factors. The folding process often cannot be reproduced in vitro, in the absence of chaperones and other cellular components1–5. However, some small proteins fold spontaneously in vitro in the absence of any other macromolecules6.\n\nWhat exactly happens during the folding of a linear polypeptide chain into a native conformation either in vivo or in vitro remains largely unknown. Despite decades of intense laboratory research, theory development and computer simulations, we still cannot recapitulate complete folding trajectories in silico, except for those of a few relatively short polypeptides7. Knowledge of the intermediates in the folding pathways and the mechanisms that enable them is essential for determining the points of intervention at which folding and misfolding processes can be altered.\n\nThe painfully slow progress in our ability to fold in silico all but the shortest polypeptides could be due to the sheer complexity of the system: the number of possible conformations of a polypeptide chain, and the number of interactions between the atoms of all amino acid residues within the polypeptide itself and with the surrounding solvent, are so astronomically high that the existing computational power is not yet sufficient, and might never become sufficient, to capture the folding trajectories for longer proteins8. It is also possible, however, that there are fundamental deficiencies in our understanding of folding as it occurs in nature, and progress in recapitulating protein folding pathways requires a more realistic physical model of folding than the one we have been relying upon.\n\nThe current dominant model of protein folding was prompted by early observations that some small proteins are able to fold in vitro into their native conformations spontaneously, in isolation from other proteins or cellular components (reviewed in 6). These observations gave rise to the thermodynamic hypothesis of protein folding6,9, which in turn led to the development of the physical model that describes protein folding as a thermodynamically favorable, unassisted process. In a more recent, refined form, this model includes the description of a rugged funnel-shaped energy landscape, in which the various unfolded, unstructured conformations occupy the high-free-energy brim of the funnel10–13. As the polypeptide chains fold, they sample conformations with progressively decreasing Gibbs free energy until they reach the native conformation, which is presumed to occupy the global thermodynamic minimum at the bottom of the funnel. The sampling of conformations during the folding process is assumed to occur via random thermal motions14. The driving force of protein folding is assumed to be the decrease in free energy to the global minimum.\n\nIn summary, the current general physical model of protein folding describes a process that occurs in a closed system in the absence of external sources of energy. It assumes that folding starts from a random, unstructured conformation and proceeds unassisted, with no apparent requirement for the folding chain to interact with other proteins or macromolecular cellular components. This model describes an extremely artificial process that is only likely to occur in vitro and has little resemblance to what takes place during the folding of all proteins in the living cell.\n\nIn nature, folding of the majority of proteins occurs in the environment of a living cell, which is an open system with a constant flow of energy and shifting chemical composition. In a cell, a polypeptide starts folding while it is still being synthesized on a ribosome, where it occupies a tight space that allows it to adopt only a limited set of conformations. The nascent peptide emerges into a crowded, viscous environment outside of the ribosomal tunnel and interacts with multiple proteins, including chaperones, and with other cellular components, at all stages of folding. In the course of peptide synthesis and co-translational folding, a large amount of energy is released by GTP hydrolysis. This energy is not required for the formation of peptide bonds15, but may be spent, at least partially, on various motions and adjustments of the ribosomal components, directly affecting the folding environment of the nascent peptide16–18. It is difficult to escape the conclusion that protein folding in vivo must be described by a physical model that takes into account the interactions of a folding polypeptide chain with its complex dynamic cellular environment.\n\nWe have recently proposed that a more realistic physical model of protein folding might be built on the assumption that protein folding in vivo is an active, energy-dependent process. In this alternative model, proteins that are not able to fold spontaneously must rely on additional external forces to achieve native conformations19. We hypothesized that the mechanism of action of such a protein folding machine might include direct mechanical manipulation of the peptide backbone by the concerted actions of the ribosome and chaperone complexes20,21. During translation in the peptidyl transferase center of the ribosome, the 3’ terminus of the tRNA in the A-site swings by nearly 180 degrees in every elongation cycle22,23. We hypothesized that this motion might lead to the rotation of the C-terminus of the nascent peptide. Simultaneously, the movements of the N-terminal regions of the nascent peptides may be restricted, first, by occlusions in the ribosome exit tunnel and then by steric capture mediated by the ribosome-associated “nascent chain welcoming committee”, such as the trigger factor in bacteria and the nascent polypeptide-associated complex in archaea and eukaryotes21. As a result, the folding polypeptide may experience transient strained conformations with elevated free energy19.\n\nAs the first step in exploring the feasibility of a protein folding machine capable of facilitating the attainment of native structure by mechanical manipulation of the peptide backbone, we performed molecular dynamics simulations augmented by application of torsion to the peptide backbones. During the simulations, the C-termini of various polypeptides were mechanically rotated either clockwise or counterclockwise, while the motions of their N-termini were restricted. We compared the trajectories of both types of simulations with the folding of the same peptides without the application of torque. In our experiments, directional rotation of the C-terminal amino acids with simultaneous limitation of the movements of the N-termini indeed facilitated the formation of native structures in five diverse alpha-helical peptides.\n\n\nMethods\n\nThe initial stretched structures of peptides (Table 1) with four additional alanine residues, two at each end, were generated using ICM software24. These alanines were attached as handles to which the rotation or restraint could be applied directly without affecting the sequence whose folding was investigated, and were not considered in the RMSD calculations. We aligned a peptide along the X-axis and solvated it in a dodecahedron box in the case of the simulations of unassisted folding and triclinic box in all other cases, with minimum distance of 1.5 nm between a peptide and the simulations box. Potassium and sodium ions were added to neutralize the charges in the system. The system was then minimized with the steepest descent algorithm, equilibrated for 100 ps in the NVT ensemble using V-rescale thermostat25 for temperature coupling, and continued in the NPT ensembles for 1 ns using V-rescale thermostat and Berendsen barostat26. After the equilibration, we kept temperature and pressure constant at 300 K and 1 bar respectively, using Nose−Hoover thermostat27,28 and isotropic Parinello−Rahman barostat29.\n\nFor all simulations, we used the ff14SB force field30 with the TIP3P water model31 and ion parameters modified by Joung and Cheatham32. Electrostatic interactions were calculated using particle-mesh Ewald (PME) summation33 with a Fourier grid spacing of 0.135 nm. For non-bonded Coulomb and Lennard-Jones interactions, 1 nm cutoff was used. We constrained the hydrogen bonds with the LINCS algorithm34 and used a 2-fs integration time step.\n\nTo exert an external mechanical torque to the C-termini of the peptides, we adopted the enforced rotation method, originally designed to study rearrangements during the rotation of a folded protein within the F1-ATPase assembly35, implemented in the GROMACS molecular dynamics package. To this end, we restrained the positions of the O and N atoms of the C-terminal alanine to keep it aligned with the X-axis, about which the rotation was applied. The restraints with a force constant of 10000 kJ/mol*nm2 were applied only for the YZ-plane, so the C-terminal amino acid could move along the X-axis. In addition, we restrained the O atom of the C-terminal amino acid in the X direction with a force constant of 5 kJ/mol*nm2 and N and Cα atoms of the N-terminal alanine with a force constant 10000 kJ/mol*nm2 in all directions. The C-terminal amino acid was rotated using a flexible axis approach (Vflex2) with a reference rotation rate of 60 degrees/ps and a force constant of 1500 kJ/mol*nm2.\n\nThe GROMACS package version 2020.236 was used for all simulations and trajectory analyses. The simulations were carried out on CUDA-enabled GPUs with Turing architecture, running Ubuntu 18.04. For visualization of protein structures and trajectories, the programs ICM-Pro 3.924 and VMD 1.9.337 were used.\n\n\nResults\n\nWe performed atomistic molecular dynamics simulations to study peptide folding under conditions when, throughout the simulation, an external mechanical torque was applied to the C-terminal amino acid of a peptide and the motions of the N-terminal amino acid were restrained (Figure 1). We compared the folding trajectories of the peptide to which a mechanical force was applied to rotate the C-terminal amino acid in one of the two possible directions – either clockwise as in Figure 1, or counterclockwise – with the trajectories for the same peptide which was allowed to fold without any motion restriction or application of any mechanical force (referred to as “unassisted folding” below). As an additional control, we ran a fourth type of simulation, where motion restraints were applied to both ends of each peptide but the torque was omitted. The details of the simulations are described in the Methods section. Each of the four types of simulations were repeated three times, giving 12 simulations for each peptide.\n\nThe force vectors applied to the C- and N-termini of a peptide in the simulation box are shown by black arrows. All force values are in kJ/mol*nm2. The purple curled arrow indicates the direction of the clockwise rotation of the peptides that resulted in the accelerated productive folding of all peptides to their helical conformations. The restrained groups are shown by green outline.\n\nThe experiments were run on five peptides that are known to adopt alpha-helical conformations in their folded form (Table 1). Two of these, P1 and P2, have been designed de novo, and the other three, P3-P5, are parts of naturally occurring proteins. The folding of the peptides was monitored by calculating the root mean square deviation (RMSD) distance of the peptide backbone from the native structure of the same fragment determined by X-ray crystallography (peptides P2-P5), or computed ab initio (peptide P1). The results of the simulations for each peptide when folded unassisted in the standard force field, and when an external torque force was added to the field, are presented in Table 2 and Figure 2. All folding trajectories and the additional information on the properties of all simulation boxes are available at Zenodo38.\n\nThe first number indicates the time (ns) spent before reaching the RMSD of 0.2 nm from the native conformation, and the second number indicates the duration of the experiment. 500/500 and 1500/1500 values indicate that folding was not observed in this simulation.\n\nEach horizontal pane represents molecular dynamics simulations for one peptide, numbered P1 through P5 (Table 1). On the left side, top three curves (dark blue, orange, and yellow) indicate three independent runs for one peptide in the standard force field without externally applied backbone rotation, and the bottom three curves (purple, green, and light blue) indicate three runs in the presence of the clockwise rotational force. On the right side, the bottom three curves are the same as in the corresponding left pane (three runs in the presence of the clockwise rotational force), and the top three curves (dark blue, orange, and yellow) indicate three runs for the same peptide in the presence of the counterclockwise rotational force.\n\nWithin our simulation lengths, we observed the completion of unassisted folding into the native-like alpha-helical structure only in some runs for one peptide, P4, which represents the third helix and preceding loop in the villin headpiece domain HP35. Other peptides remained essentially unfolded throughout the 500–1500-nanosecond runs. The peptides also failed to fold when their ends were restricted in mobility but torque was not applied (Table 2). In contrast, when the external torsion force was applied to the C-termini of the peptides in the clockwise direction, as described in Methods and illustrated in Figure 1, peptides P1-P4 all folded into alpha-helical structures and were brought within 0.2 nm RMSD from their native structures in every run, typically within the first 100–200 ns of simulation. These peptides stayed in the native or nearly-native conformations for the remainder of the experiments. Peptide P5 was a special case; similarly to P1-P4, it adopted a compact conformation early in the experiments, but remained only partially folded for the duration of all runs (Figure 2).\n\nFor all five peptides, folding was observed when the rotation force was applied to the C-terminal amino acid in the clockwise direction (Figure 1). In contrast, the torque applied to the C-terminus counterclockwise with the same force constant did not facilitate folding of P1-P3 and P5, and may have inhibited folding of P4 (Figure 2).\n\n\nDiscussion and conclusions\n\nTo test the idea that inclusion of external forces can improve modeling of protein folding pathways in silico, we performed molecular dynamics simulations in which a standard force field was augmented by the application of external mechanical forces to the polypeptide backbone. We compared these simulations to control runs without any additional external forces. The directional rotation of the C-terminal amino acid with simultaneous restriction of the movements of the N-terminal amino acid facilitated the formation of native structures in five diverse alpha-helical peptides, confirming that such constraints can have significant consequences for folding dynamics. Strikingly, application of mechanical force accelerated the folding of P4, a fragment of an on-pathway folding intermediate of the well-studied villin headpiece domain HP35, which is one of the fastest-folding protein domains known7,39,40. The several-fold increase in the rate of P4 folding that was achieved in our experiments seems to suggest that the postulated “physical limit of folding” of HP35 as a whole39,41 could be overcome by a protein folding machine. The other four peptides in our experiments likewise attained their alpha-helical structure in the presence of the rotating force, but did not reach their native conformations when allowed to fold unassisted, even though we ran the control unassisted simulations for ~10 times longer than the simulations that included the application of the external force (Table 2). Some of those peptides might take a very long time to reach their native conformations without application of an external force, whereas others might never fold unassisted, if their unfolded states are more stable than the folded conformations.\n\nThese results are in line with our protein folding machine hypothesis19. They also support a hypothetical mechanism through which the machine would directly alter the conformations of proteins by applying mechanical force to the peptide backbone20,21. The feasibility of such a mechanism, however, is dependent on whether the torsion applied at one point of a peptide would propagate through the rest of the peptide chain and affect the movements of the distal parts of the peptide. The peptide backbone is often viewed as a freely jointed chain, due to the 360-degrees rotation ability around the phi- and psi-bonds within each amino acid42. If the peptides in our simulations were to behave as freely jointed chains, the rotation of a single amino acid at the end of the peptide would not have any appreciable effect on the motions of the rest of the peptide. However, if a mechanical torque were applied to a peptide while it was being folded in a viscous crowded environment (e.g., co-translationally in a living cell), we predicted that the free rotation of the phi- and psi-bonds in the peptide backbone would be hindered enough that escape from the forbidden sections of the Ramachandran plots would become difficult for many residues, and as a result, the entire peptide backbone may experience transient strained conformations. Although our simulation could not account for all the details of the protein folding environment in vivo, we were able to devise a set of conditions under which the peptide indeed did not behave as a freely jointed chain. When a force was applied to a single amino acid residue, and the motion of just one other residue at least 15 amino acids apart was restricted, the folding trajectory of the entire peptide was affected dramatically, leading to the rapid attainment of the native helical conformation. Some of the steric hindrances that make this rapid folding possible involve amino acid side chains, and therefore the effect might be sequence-specific. For example, glycine residues are more likely to experience the full 360-degree rotation around the phi- and psi-bonds, relieving the strain in the main chain; this might explain why P5, a peptide with an internal glycine, first acquired and then partially lost its folded conformation in our experiments (Figure 2).\n\nIt remains unclear whether our simulation captures the main features of the folding process as it occurs in nature. For example, one of the parameters that differs between our simulations and real co-translational protein folding process is their characteristic times. The rotation of the backbone in our system occurs at the submicrosecond time scale, whereas the addition of amino acids to the nascent peptide is much slower, on the order of subseconds43–45. Molecular dynamics simulations have been known to model, at a fast scale, the essential parts of the molecular processes that are much slower when observed with bulk kinetics or single-molecule methods17,40, but the effect of the rotation rate on the peptide folding trajectory remains to be investigated.\n\nThe key feature of the hypothetical mechanism of co-translational protein folding that we simulated is the directional rotation of the peptide backbone. As discussed above, the 3’ terminus of the tRNA in the A-site of the ribosome peptidyl transferase center turns by nearly 180 degrees in every translation elongation cycle. Only a 45-degree swing is necessary to achieve the proper stereochemistry of the peptide bond formation46; the function of the remaining portion of the turn is unknown, and we have hypothesized that it may be needed to facilitate co-translational folding20,21. It is notable, however, that the tRNA within the translating ribosome appears to turn in the counterclockwise direction when looking from the C-terminus of the nascent peptide22,23. In contrast, folding of all peptides into the right-handed alpha-helices in our experiments took place only with clockwise rotation of the C-termini (Figure 1 and Figure 2). It remains to be determined what, exactly, happens to the nascent peptide in the peptidyl transferase center and in the ribosome exit tunnel. The nascent peptide might be rotated counterclockwise (in the direction of the tRNA swing), or clockwise (as a result of a gear-like interaction with the tunnel walls), or might not be rotated at all but rearranged in a more complex way, being subject to pushing and pulling forces as well as interactions with the exit tunnel walls and other components of the ribosomal complex.\n\nRegardless of whether the peptide torsion mechanism operates during co-translational folding on the ribosome in vivo, we demonstrate that it is possible to facilitate protein folding under conditions when an external mechanical force is applied to the peptide backbone. Importantly, we show that the peptide does not always behave as a freely jointed chain, opening the possibility that in vivo the peptide backbone can be manipulated into conformations that cannot be reached without assistance because they are either thermodynamically unstable or kinetically inaccessible. The results of our simulations thus demonstrate the feasibility of a protein folding machine. Some recently published results, including studies of the role of the exit tunnel in nascent chain folding47–51 and of direct coupling between ATP hydrolysis and protein refolding by the chaperones of the HSP70 family52–54, may be also interpreted as evidence of protein folding in vivo being an active process.\n\nThe notion of an active, energy-dependent protein folding mechanisms in vivo is better compatible with the current understanding of evolution than the generally accepted, standard thermodynamic hypothesis of protein folding. Although it is accepted that the ability of proteins to attain their native conformations must have evolved by natural selection of sequences that fold quickly and correctly (“evolution solved the protein folding problem”55), models of unassisted folding sidestep the fact that ribosomes and translation factors are among the oldest molecular machines shared by all extant cellular life56, and were present during much of the evolution of proteins and of their folding pathways. The evolutionary optimization of the tempo and mode of protein folding, for at least 3.5 billion years of biological evolution, has taken place not in dilute solutions of isolated proteins, but in a dynamic environment of living cells with their constant flow of matter and energy. Thus, the ability of any present-day protein to fold in isolation and without assistance is likely to be either an incidental or derived property, not shared by most other proteins. Realistic computational modeling of protein folding must therefore take into account the presence of a multitude of external forces. Further studies should attempt to more closely recreate the conditions of protein folding in vivo.\n\n\nData availability\n\nZenodo: Energy-dependent protein folding: modeling how a protein folding machine may work. http://doi.org/10.5281/zenodo.439295938\n\nThis project contains the following underlying data:\n\n- trajectories.zip (xtc files of the folding trajectories obtained in the molecular dynamics simulations)\n\n- m2020nepfSF1.pdf (pdf file of the properties of all simulation boxes)\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).\n\n\nAuthors’ contributions\n\nConceptualization: INS; Formal analysis: HKS, INS; Funding acquisition: KBN, INS; Investigation: HKS, KBN, ARM, INS; Methodology: HKS, INS; Visualization: HKS; Writing – original draft preparation: ARM, INS; Writing – review and editing: HKS, KBN, ARM, INS.",
"appendix": "Acknowledgments\n\nWe are grateful to many colleagues, most of all Dr. Yuri Wolf, for useful discussions. We thank Dr. Alexandra Mushegian for critical reading of this manuscript. INS is grateful to an anonymous angel investor for their support of Strenic, LLC. An earlier version of this article can be found on BiorXiv (doi:10.1101/2020.09.01.277582).\n\n\nReferences\n\nMartin J, Langer T, Boteva R, et al.: Chaperonin-mediated protein folding at the surface of groEL through a 'molten globule'-like intermediate. Nature. 1991; 352(6330): 36–42. PubMed Abstract | Publisher Full Text\n\nChow D, Srikakulam R, Chen Y, et al.: Folding of the striated muscle myosin motor domain. J Biol Chem. 2002; 277(39): 36799–36807. PubMed Abstract | Publisher Full Text\n\nBalchin D, Miličić G, Strauss M, et al.: Pathway of actin folding directed by the eukaryotic chaperonin TRiC. Cell. 2018; 174(6): 1507–1521.e16. PubMed Abstract | Publisher Full Text\n\nBauer D, Meinhold S, Jakob RP, et al.: A folding nucleus and minimal ATP binding domain of Hsp70 identified by single-molecule force spectroscopy. Proc Natl Acad Sci U S A. 2018; 115(18): 4666–4671. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTo P, Whitehead B, Tarbox HE, et al.: Non-Refoldability is Pervasive Across the E. coli Proteome. bioRxiv. 2020; 273110. Publisher Full Text\n\nAnfinsen CB: Principles that govern the folding of protein chains. Science. 1973; 181(4096): 223–30. PubMed Abstract | Publisher Full Text\n\nLindorff-Larsen K, Piana S, Dror RO, et al.: How fast-folding proteins fold. Science. 2011; 334(6055): 517–20. PubMed Abstract | Publisher Full Text\n\nLevinthal C: How to fold graciously. Mössbauer Spectroscopy in Biological Systems Proceedings (1969). University of Illinois Bulletin. 1969; 67(41): 22–24. Reference Source\n\nHaber E, Anfinsen CB: Side-chain interactions governing the pairing of half-cystine residues in ribonuclease. J Biol Chem. 1962; 237(6): 1839–1844. PubMed Abstract\n\nBryngelson JD, Onuchic JN, Socci ND, et al.: Funnels, pathways, and the energy landscape of protein folding: a synthesis. Proteins. 1995; 21(3): 167–95. PubMed Abstract | Publisher Full Text\n\nOnuchic JN, Luthey-Schulten Z, Wolynes PG: Theory of protein folding: the energy landscape perspective. Annu Rev Phys Chem. 1997; 48: 545–600. PubMed Abstract | Publisher Full Text\n\nDill KA, Ozkan SB, Shell MS, et al.: The Protein Folding Problem. Annual Rev Biophys. 2008; 37: 289–316. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRollins GC, Dill KA: General Mechanism of Two-State Protein Folding Kinetics. J Am Chem Soc. 2014; 136(32): 11420–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDill KA, MacCallum JL: The protein-folding problem, 50 years on. Science. 2012; 338(6110): 1042–6. PubMed Abstract | Publisher Full Text\n\nSpirin AS: Ribosomes. Kluwer Academic/Plenum Publishers, 1999. Reference Source\n\nLim VI, Curran JF, Garber MB: Ribosomal Elongation Cycle: Energetic, Kinetic and Stereochemical Aspects. J Mol Biol. 2005; 351(3): 470–480. PubMed Abstract | Publisher Full Text\n\nBock LV, Blau C, Schroder GF, et al.: Energy barriers and driving forces in tRNA translocation through the ribosome. Nat Struct Mol Biol. 2013; 20(12): 1390–6. PubMed Abstract | Publisher Full Text\n\nMaracci C, Rodnina MV: Translational GTPases. Biopolymers. 2016; 105(8): 463–75. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSorokina I, Mushegian A: Modeling protein folding in vivo. Biol Direct. 2018; 13(1): 13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSorokina I, Mushegian A: The role of the backbone torsion in protein folding. Biol Direct. 2016; 11(1): 64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSorokina I, Mushegian A: Rotational restriction of nascent peptides as an essential element of co-translational protein folding: possible molecular players and structural consequences. Biol Direct. 2017; 12(1): 14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBashan A, Agmon I, Zarivach R, et al.: Structural basis of the ribosomal machinery for peptide bond formation, translocation, and nascent chain progression. Mol Cell. 2003; 11(1): 91–102. PubMed Abstract | Publisher Full Text\n\nAgmon I, Bashan A, Zarivach R, et al.: Symmetry at the active site of the ribosome: structural and functional implications. Biol Chem. 2005; 386(9): 833–44. PubMed Abstract | Publisher Full Text\n\nAbagyan R, Totrov M, Kuznetsov D: ICM - A new method for protein modeling and design: Applications to docking and structure prediction from the distorted native conformation. J Comput Chem. 1994; 15(5): 488–506. Publisher Full Text\n\nBussi G, Donadio D, Parrinello M: Canonical sampling through velocity rescaling. J Chem Phys. 2007; 126(1): 014101. PubMed Abstract | Publisher Full Text\n\nBerendsen HJC, Postma JPM, van Gunsteren WF, et al.: Molecular dynamics with coupling to an external bath. J Chem Phys. 1984; 81(8): 3684–3690. Publisher Full Text\n\nNose S: A molecular dynamics method for simulations in the canonical ensemble. Molecular Physics. 1984; 52(2): 255–268. Publisher Full Text\n\nHoover WG: Canonical dynamics: Equilibrium phase-space distributions. Phys Rev A Gen Phys. 1985; 31: 1695–1697. PubMed Abstract | Publisher Full Text\n\nParrinello M, Rahman A: Polymorphic transitions in single crystals: A new molecular dynamics method. J Appl Phys. 1981; 52: 7182–7190. Publisher Full Text\n\nMaier JA, Martinez C, Kasavajhala K, et al.: ff14SB: Improving the Accuracy of Protein Side Chain and Backbone Parameters from ff99SB. J Chem Theory Comput. 2015; 11(8): 3696–3713. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJorgensen WL, Chandrasekhar J, Madura JD, et al.: Comparison of simple potential functions for simulating liquid water. J Chem Phys. 1983; 79(2): 926–935. Publisher Full Text\n\nJoung IS, Cheatham TE: Determination of Alkali and Halide Monovalent Ion Parameters for Use in Explicitly Solvated Biomolecular Simulations. J Phys Chem B. 2008; 112(30): 9020–9041. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDarden T, York D, Pedersen L: Particle mesh Ewald: An Nlog(N) method for Ewald sums in large systems. J Chem Phys. 1993; 98(12): 10089–10092. Publisher Full Text\n\nHess B, Bekker H, Berendsen HJC, et al.: LINCS: A linear constraint solver for molecular simulations. J Comput Chem. 1997; 18(12): 1463–1472. Publisher Full Text\n\nKutzner C, Czub J, Grubmüller H: Keep it flexible: driving macromolecular rotary motions in atomistic simulations with GROMACS. J Chem Theory Comput. 2011; 7(5): 1381–1393. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAbraham MJ, Murtola T, Schultz R, et al.: GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers. SoftwareX. 2015; 1–2: 19–25. Publisher Full Text\n\nHumphrey W, Dalke A, Schulten K: VMD: Visual molecular dynamics. J Mol Graph. 1996; 14(1): 33–38. PubMed Abstract | Publisher Full Text\n\nSahakyan HK, Nazaryan KB, Mushegian AR, et al.: Energy-dependent protein folding: modeling how a protein folding machine may work. [Data set]. Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4392959\n\nEnsign DL, Kasson PM, Pande VS: Heterogeneity even at the speed limit of folding: large-scale molecular dynamics study of a fast-folding variant of the villin headpiece. J Mol Biol. 2007; 374(3): 806–816. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang E, Tao P, Wang J, et al.: A novel folding pathway of the villin headpiece subdomain HP35. Phys Chem Chem Phys. 2019; 21(33): 18219–18226. PubMed Abstract | Publisher Full Text\n\nKubelka J, Hofrichter J, Eaton WA: The protein folding “speed limit.” Curr Opin Struct Biol. 2004; 14(1): 76–88. PubMed Abstract | Publisher Full Text\n\nFleming PJ, Rose GD: Conformational properties of unfolded proteins. Protein Science Encyclopedia. A. R. Fersht, Ed. Wiley-VCH Verlag GmbH, 2008; 2: 710–736. Publisher Full Text\n\nTalkad V, Schneider E, Kennell D: Evidence for variable rates of ribosome movement in Escherichia coli. J Mol Biol. 1976; 104(1): 299–303. PubMed Abstract | Publisher Full Text\n\nOlofsson SO, Boström K, Carlsson P, et al.: Structure and biosynthesis of apolipoprotein B. Am Heart J. 1987; 113(2): 446–452. PubMed Abstract | Publisher Full Text\n\nLi GW, Burkhardt D, Gross C, et al.: Quantifying Absolute Protein Synthesis Rates Reveals Principles Underlying Allocation of Cellular Resources. Cell. 2014; 157(3): 624–635. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGindulyte A, Bashan A, Agmon I, et al.: The transition state for formation of the peptide bond in the ribosome. Proc Natl Acad Sci U S A. 2006; 103(36): 13327–13332. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNilsson OB, Hedman R, Marino J, et al.: Cotranslational protein folding inside the ribosome exit tunnel. Cell Rep. 2015; 12(10): 1533–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFarías-Rico JA, Goetz SK, Marino J, et al.: Mutational analysis of protein folding inside the ribosome exit tunnel. FEBS Lett. 2017; 591(1): 155–163. PubMed Abstract | Publisher Full Text\n\nKemp G, Kudva R, de la Rosa A, et al.: Force-Profile Analysis of the Cotranslational Folding of HemK and Filamin Domains: Comparison of Biochemical and Biophysical Folding Assays. J Mol Biol. 2019; 431(6): 1308–1314. PubMed Abstract | Publisher Full Text\n\nBitran A, Jacobs WM, Zhai X, et al.: Cotranslational folding allows misfolding-prone proteins to circumvent deep kinetic traps. Proc Natl Acad Sci U S A. 2020; 117(3): 1485–1495. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTao P, Wang E, Xiao Y: Pathway regulation mechanism revealed by cotranslational folding of villin headpiece subdomain HP35. Phys Rev E. 2020; 101(5–1): 052403. PubMed Abstract | Publisher Full Text\n\nBarducci A, De Los Rios P: Non-equilibrium conformational dynamics in the function of molecular chaperones. Curr Opin Struct Biol. 2015; 30: 161–169. PubMed Abstract | Publisher Full Text\n\nGoloubinoff P, Sassi A, Fauvet B, et al.: Chaperones convert the energy from ATP into the nonequilibrium stabilization of native proteins. Nat Chem Biol. 2018; 14(4): 388–395. PubMed Abstract | Publisher Full Text\n\nAssenza S, Sassi AS, Kellner R, et al.: Efficient conversion of chemical energy into mechanical work by Hsp70 chaperones. eLife. 2019; 8: e48491. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWolynes PG: Evolution, energy landscapes and the paradoxes of protein folding. Biochimie. 2015; 119: 218–30. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBowman JC, Petrov AS, Frenkel-Pinter M, et al.: Root of the tree: the significance, evolution, and origins of the ribosome. Chem Rev. 2020; 120(11): 4848–4878. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "76648",
"date": "25 Jan 2021",
"name": "Lisa J. Lapidus",
"expertise": [
"Reviewer Expertise I study protein folding using ultra-fast experimental methods. I have pioneered the method of measuring intramolecular diffusion of unfolded proteins leading to the estimates of protein folding speed limits."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper presents an intriguing hypothesis, that torque applied by the ribosome at the A-site, actively assists folding. To demonstrate this, the folding of 5 small proteins were simulated from an extended state. When clockwise torque is applied to the C-terminus while constraining the N-terminus (so that the torque is maintained within the backbone), the proteins typically folded within 1.5 microseconds and did not fold within that window without torque, without restraint, and if the torque was applied counter-clockwise.\n\nThe results are quite convincing but a few details are missing. Below are some questions whose answers would add to the understanding of the reader:\n1. What happens to the backbone after the torque is applied? It seems like the N-terminus is constrained for the entire length of the simulation but this should be made more clear.\n\n2. How does the collapse of the backbone proceed with the constraints compared to without?\n3. Do the helical segments form simultaneously or do helices propagate down the length of the chain? Another figure showing native structures of the sequences color coded by when native structure is formed might by useful (or something equivalent).\nThis paper has an interesting premise with a lot of caveats given the simplified results supporting the conclusion. However, the authors do a good job of addressing some of them. I am not completely convinced that crowded cell conditions could provide the N-terminal constraint required to acquire structure rapidly, but this is a testable result for current co-translational folding simulations. The authors discuss that the torque applied by the ribosome tunnel is a open question and may not be mimicked by the simple torque in the simulations. Another question that could be answered by more realistic simulations is what is the effect of torque at the A-site on the nascent chain outside the ribosome tunnel. While helices can form in the tunnel, there is not a lot of evidence that helices emerge from the tunnel for most sequences. So is the net torque on the last part of the chain before leaving the tunnel what really matters?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "6511",
"date": "29 Mar 2021",
"name": "Irina Sorokina",
"role": "Author Response",
"response": "We are grateful to Dr. Lapidus for their support of our work, for the criticism and for interesting discussion. To answer the three numbered questions: Yes, in this study the position of the N-terminus was constrained in all directions for the entire length of the simulation. Without constraints, only one target (Helix 3 of villin HP35 domain; our P4) collapsed to the helical conformation, albeit several times slower than with our constrains. Please see Figure 2 and Table 2 for values. In our simulations the helical segments did not form simultaneously; their behavior was very different and sequence dependent – some propagated from the C to the N terminus, in others the helical structure started forming in the middle of the peptide and propagated towards the ends, and in one case two small helical segments started closer to the termini and then merged in the middle. The trajectories are available on Zenodo and can be visualized. We are currently working on a good way of presenting the mode and timing of the different parts of the targets attaining their native conformations. Regarding the ability of the crowded cell conditions to provide the N-terminal constraints during co-translational folding: as part of the protein folding machine hypothesis we consider the possibility that the most important function of the various chaperones that meet the nascent peptide at the end of the ribosomal exit tunnel, such as trigger factor in bacteria and its eukaryotic analogs, may be the immobilization of the N-terminus of the nascent peptide to achieve the effect similar to what is observed in our current simulations. In our previous publications (references 19, 20 and in particular ref. 21 in the main text) we discussed in some detail the possible involvement of a number of other specific cellular components. “Crowdedness factor”, acting through the excluded volume effect or other mechanisms, is just one of these possibilities. As for more realistic simulations of the co-translational folding process, those are certainly needed. The postulated torque on the nascent peptide chain may be applied as part of polypeptide synthesis, whereas in this study we simplified by working with the entire chains and did not model stepwise addition of the amino acids. The force that is applied also may need to be adjusted, as well as its timing and multiple other factors. The question of the residual torque on the nascent peptide outside the ribosome is interesting. In our simulations we observed that the twisted peptides did not behave as freely jointed chains. It means that if some movement restrictions (e.g., by interactions with chaperones) are present outside of the ribosome, the residual twist of the backbone will still be present and have an impact on the folding trajectory well beyond the ribosomal tunnel."
}
]
},
{
"id": "77095",
"date": "02 Feb 2021",
"name": "Stephen D. Fried",
"expertise": [
"Reviewer Expertise Protein folding biophysics",
"proteomics and mass spectrometry."
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this short article, Sahakyan et al. present molecular dynamics (MD) simulations that seek to question the overarching paradigm that protein folding is a predominantly thermodynamically-driven process in which native states are achieved exclusively by identifying states with minimal free energy in the absence of any energy dissipation or physical forces acting on them.\nIn contrast, by presenting several µs-length trajectories of several short (20mer) alpha-helix forming peptides, the authors show that the ability for the peptide to rapidly adopt a helical conformation is dependent on the application of an externally applied torque, rotating the C-terminus clockwise. The authors draw the connection that the C-terminus of a nascent protein might also experience rotary forces during co-translational folding induced by the movement of tRNA molecules transiting through the A-site and P-site\nThis article presents an interesting and novel idea that the protein folding community should be aware of, as recent years have seen a growing number of contributions highlighting the potential difference between co-translational folding and ‘classical’ folding experiments. I would say that rather than ‘answering’ a problem, it is more an invitation for future work and thought, given the relatively few results that have been reported. That being said, the results presented are intriguing and internally consistent, and deserve the consideration of any worker in the protein folding field interested in understanding how biological folding could differ from the classic scenario of an ergodic search on a constant free energy landscape.\nIn evaluating this work, I have a number of conceptual comments, suggestions for further study, and recommendations on reproducibility.\nConceptual comments:\nAt several points, the authors mention that the 3’ terminus of the A-site tRNA undergoes a large-scale (ca. 180˚) rotation, and that this forms the basis of their hypothesis that rotary motions could be transduced to a nascent chain during translation. The motion that the authors are referring to – if I’m not mistaken – is the accommodation of the tRNA from the A/T-hybrid state to the A/A-canonical state upon hydrolysis of GTP by EF-Tu (though they do not mention this process by name). My understanding is that the nature of this large rotation *is* understood; it results from the fact that the presence of EF-Tu precludes the ‘top half’ of the tRNA from entering the A-site on the 50S subunit whilst the anticodon engages with the decoding center; release of EF-Tu then allows the 3’-end of the A-site tRNA to swing into the PTC. That being said, this large movement would not be experienced by the nascent chain per se, because it is occurring on the A-site tRNA (not the P-site tRNA carrying the nascent chain). Hence, the authors may need to clarify what rotational conformational changes are experienced by the P-site tRNA, and provide structural data that directly support this.\n\nIs the rate of the rotary movement relevant to the timescale of translation? If not, this should be commented on directly, or at least probed by varying its rotational frequency.\nFurther study:\nSeeing as the larger rotational movement is associated with the A-site tRNA (before its aminoacylated acid is added to the peptidyl chain), the authors may want to use simulation to directly test their hypothesis that “this motion might lead to the rotation of the C-terminus of the nascent peptide.” I appreciate the challenge involved in such, as it would surely require performing relatively long simulations on the ribosome. Though this would represent a truly important contribution and should be considered. If simulations of this size/timescale are too infeasible, potentially a careful analysis of the many ribosome structures with P-site tRNA-nascent chains bound could also serve as a reasonable way to interrogate this question.\n\nThe authors seem to assume that all the peptides they studied *could* become alpha helical under the present forcefield, and that this is accelerated by the presence of the external torque. I suppose the first part of this should probably be shown explicitly, perhaps by a free-energy perturbation or umbrella sampling approach, just to show that the minimum free-energy states are in fact what we would think they are.\n\nWhat is the mechanism whereby the rotary force induces folding? Careful analysis of the trajectories could probably support or refute the hypothesis that it helps ‘seed’ the formation of the alpha helix, which then ‘zippers’ up. (Actually the zippering model for alpha-helix formation is probably very relevant to the authors’ discussion, and they might consider seeing if their simulations reproduce the parameters from classical statistical mechanical treatments of the helix-coil problem).\nReproducibility:\nSeeing as performing MD simulations in the presence of external torques is fairly non-standard, the authors may want to provide the actual GROMACS input files for any computational biochemist interested in performing similar simulations.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "6512",
"date": "29 Mar 2021",
"name": "Irina Sorokina",
"role": "Author Response",
"response": "We thank Dr. Fried for their positive assessment of our study, and agree that it is just a first step and an invitation for future work and thought, as any new observation surely is. Regarding Dr. Fried's long and interesting Conceptual Comment: Several distinct motions of the components of translation machinery occur in each elongation cycle. One such motion is what Dr. Fried is referring to, namely the accommodation of the aa-tRNA in the 50S subunit, which follows the codon recognition in the decoding center of the 30S subunit. The end point of this accommodation is the placement of the charged 3'-end of aa-tRNA into the A-site of the peptidyl transferase center (PTC). The aa-tRNA moves inside the “accommodation corridor” made of ~20 nucleotides, most of which are not part of the PTC. The pendulum-like “swing” in this motion appears to involve mostly the acceptor stem of aa-tRNA, which changes its angle with regards to the rest of tRNA. The entire accommodation process precedes the peptide bond formation (details in Rodnina and Wintermeyer, 2001, doi:10.1146/annurev.biochem.70.1.415; Sanbonmatsu et al., 2005, doi: 10.1073/pnas.0503456102; Rodnina et al., 2017, doi: 1098/rstb.2016.0182). Another, apparently distinct, motion is the one that we discuss in the manuscript, i.e., the rotational swing of the 3'-terminus of the aa-tRNA as it transitions from the A-site to P-site. This motion occurs entirely within the PTC, the “swing” involves the 3'-terminus of tRNA rotating about the two-fold symmetry axis of the PTC, and the peptidyl transfer occurs concomitantly with the motion (references 22, 23 and 46 in the main text of the article). The accommodation process appears to be slower than the rotation in the PTC (Rodnina and Wintermeyer, 2001). Thus, the two motions appear to differ in several ways, though it is quite possible that they are coupled. In any case, even if the former motion does not affect the nascent chain per se, the latter motion is likely to. Yet another rotational motion even later in the elongation cycle is ratcheting of the entire two ribosome subunits; it is not discussed here in detail, though it also may be of interest for understanding the mechanisms of energy-dependent protein folding. Regarding the plausibility of the alpha-helical conformation in our targets when folding unassisted in the same force field: we have selected our targets because their alpha-helical conformation either have been known from the native structure of the domains they are found within (targets P2-P5) or computed using molecular dynamics (target P1). Since the peptides that we used are fragments of larger proteins, their conformations in isolation may or may not correspond to the thermodynamic minimum, and they may not necessarily be expected to remain stable in the absence of the entire set of native contacts supporting the folded conformation in the full length proteins. That said, we did observe unassisted folding of Helix 3 of villin domain HP35 (our target P4) to nearly-native conformation, but the process was several-fold slower than with application of the torque force. We also agree that longer simulations of larger molecular complexes would be quite desirable (and possibly tractable with specialized hardware and with improved physical models of folding), and that given the complexity of the task, variation of rotational frequency may be a practical workaround in the meantime. The all-atom simulation of the entire ribosome and reconstructions from the available structures is the ultimate goal. The amount of work involved in such studies may be sufficient for the entire community interested in the mechanisms of protein folding on the ribosome. The GROMACS input files for all targets with enforced rotation applied to the C-terminal amino acid of a peptide can be found at https://zenodo.org/record/4625519"
}
]
},
{
"id": "78915",
"date": "19 Feb 2021",
"name": "Antonio Trovato",
"expertise": [
"Reviewer Expertise Computational biophysics",
"protein folding and aggregation"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this short contribution the authors set out to validate, by means of all-atom molecular dynamics simulations in explicit solvent, the idea that the folding of proteins into their biologically relevant native state depends crucially on the presence of the \"in vivo\" cellular machinery. They have been arguing through several papers, in recent years, that only a minor subset of short proteins conforms to the widely accepted thermodynamic hypothesis (i.e. the native state is the global free energy minimum for a protein in the test tube). In their view, most proteins are not able to fold \"in vitro\" and instead evolved to be able to fold while they are synthesized at the ribosome. The energy flow provided by the cellular machinery through mechanical manipulations would than make protein folding an activated process \"in vivo\".\nHere, in particular, they consider short peptides (around 20 residues) with helical native states. Their molecular dynamics simulations, 1.5 microseconds long, show that folding to the native state rarely occurs in the absence of any restraint. On the other hand, when restraints on the position of the N-terminal residue and a constant clockwise torque on the C-terminal residue are applied, folding to the native helical state readily occurs for all peptides.\nThe paper is well written and makes for an interesting read. The authors are very careful in making it clear that any connection between the mechanical restraints used in their simulations and the ones present during co-translational folding is yet to be substantiated. Their findings are in principle interesting since they show how energy pumping through a torque may in fact improve the ability of the folding process to reach a given configuration.\nHowever, the authors should consider carefully the following remarks, which could undermine, to some extent, their conclusions:\nI think it is important to check whether the helical native states are stable for the considered all-atom force field in the absence of any restraint (e.g. running a microsecond long simulation with the native state as initial condition); if not, failure to fold into the native state could be more simply ascribed to incorrect force field parametrization. I am assuming that the stability of the native state in solution, at least on the microsecond time scale, should be guaranteed also within the unorthodox view advocated by the authors; a comment by them on this point would be useful.\n\nConsidering only helical native states may bias the conclusion of this study, due to their definite right-handed chirality. Applying a clockwise torque may be effective in biasing any polymer chain towards adopting a right-handed helical configuration. The authors should then consider, at least for future work, to run simulations for peptides forming beta-hairpins or beta-sheets in their native states.\n\nAlong the same lines, it had been in fact observed that helical shapes are kinetically favored for growing polymers attached to a moving end1, even in the absence of an explicit torque. My point is again that what found by the authors could be a generic feature shared by any polymer chain, not a property specific to peptide sequences adopting a helical native state.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "6513",
"date": "29 Mar 2021",
"name": "Irina Sorokina",
"role": "Author Response",
"response": "We are grateful to Dr. Trovato for their close reading of our published work, for a great concise summary of the paper and for important points of discussion. To answer specific questions: Regarding the stability of the native helical structures of our target peptides in our experimental conditions: we did not study their destabilization from folded forms in our experiments and we should indeed perform this study. We must note, however, that the peptides that we used are fragments of larger proteins, and they may be expected to be unstable or transiently stable in the absence of the entire set of native contacts supporting the folded conformation in the full length protein. That said, we did observe unassisted folding of Helix 3 of villin domain HP35 (our target P4) to nearly-native conformation in 300-500 ns. We are now experimenting with larger protein domains. Their stability in the simulations, as well as the dependence of their stability on the force field parametrization, are important goals for the future studies. Since the publication of this paper, we have started studying the effect of torque on folding of protein domains from different structural classes. The main new result is that the application of torque facilitates folding of some protein domains containing not only alpha helices, but also beta strands (manuscript in preparation). Thank you for pointing out this interesting study! Each and every mechanism that might play a role in the co-translational folding needs to be considered and studied – there must be multiple elements contributing to the working of the folding machine."
}
]
},
{
"id": "78384",
"date": "22 Feb 2021",
"name": "José Arcadio Farías Rico",
"expertise": [
"Reviewer Expertise Cotranslational protein folding",
"biophysics",
"synthetic biology",
"protein evolution."
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis short MD publication by Sahakyan et al., proposes an interesting alternative to the classic thermodynamic theory of protein folding in which the polypeptide chain acquires its native conformation spontaneously without any external aid. The authors argue that protein folding must be modelled as an active energy dependent process assisted by a hypothetical folding machine (the ribosome) that applies torsion to the polypeptide chain.\nThey present folding trajectories (up to 1500 ns) of five short peptides (15-21 aa) under four different conditions: a) unassisted folding without restrictions and rotation, b) motions restraints to both ends of the peptide, c) a mechanical force is applied to rotate the C-terminus counterclockwise while the N-terminus is restrained and d) a mechanical force is applied to rotate the C-terminus clockwise while the N-terminus is restrained. Most of the peptides adopt an α helical conformation only in the last condition. The authors suggest that the ribosome, and specially the tRNA molecules moving from the A-site to the P-site, apply rotary forces to the nascent chain during cotranslational folding.\n\nThe idea is interesting and the protein folding community would benefit from the indexing of this study. Currently, it is well established that some proteins fold differently while being synthesized by the ribosome than refolded in diluted buffer conditions. The work acceptably adds to these new ideas, the MD experiments are consistent and its properly presented. This study certainly provides food for thought and it does not intend to present a complete picture of the forces acting on the polypeptide chain during cotranslational folding.\n\nThe exit tunnel is a complex environment where transient interactions can be stablished between every residue of the growing chain and the macromolecules lining the tunnel. A myriad of factors other than torsion can aid the folding of macromolecules in the tunnel (conformational entropy reduction, transient salt-bridges and Van Der Waals interactions, for instance). Therefore, torsion might play a role but more experiments are needed to fully support this assumption. Also, since the force is applied in the same direction as the alpha helix formation it would definitely accelerate the folding process.\nThe weakest point of the publication is the timescale. The rotation (60 degrees/ps) is several orders of magnitud faster than tRNA rotation (1aa/50ms) given a bacterial translation rate of 20aa/s. It would be interesting to see another setup where they rotate the C-terminal residue by 180 degrees only once, and then follow the equilibration of the peptide conformation for some microseconds (helices take microseconds to milliseconds to fold). If the hypothesis is correct one should see the same increased tendency for the peptide to become helical with clockwise rotation compared to counterclockwise as they do under continuous rotation.\nSeveral further studies are possible, such as testing a bigger number of peptides with different helix propensities to draw statistics on how likely is that torque in the chain produce helix formation versus the effect of the simple reduction on the conformational entropy by confining the peptide in a closed environment (as it is the ribosomal tunnel).\nAlso one could try to perform MD with torsion on small protein domains that have already proven to fold very deep of the ribosomal tunnel. There is a very small domain (29 aa) (ref 47) for which its folding has been tested. How difficult is to test the same conditions used for the peptides with this domain? This domain is almost the same size than the peptides and it is a much better representative of a folded domain.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "6514",
"date": "29 Mar 2021",
"name": "Irina Sorokina",
"role": "Author Response",
"response": "We are grateful to Dr. Farias Rico for their criticism and for helpful suggestions of possible future experiments. We agree that many different physical effects may take place in the environment of the exit tunnel. We also agree that it would be important to study the torque-facilitated folding at the microsecond scale. Interrupted rotation is a good idea that we have considered already; in our current work, we have been applying the torque within the first 100-200 nanoseconds of simulation followed by the release of both ends of the polypeptide, and under such regime the acceleration of folding of peptides and some small protein domains is also observed (manuscript in preparation). It is certainly will be interesting to simulate repeated very short duration applications of the torque. We also have some new results showing that the application of torque facilitates folding of some protein domains containing not only alpha helices, but also beta strands (manuscript in preparation)."
}
]
}
] | 1
|
https://f1000research.com/articles/10-3
|
https://f1000research.com/articles/9-1161/v1
|
21 Sep 20
|
{
"type": "Research Article",
"title": "Development of species-specific SCAR markers for identification and authentication of three rare Peninsular Malaysian endemic Coelogyne (Orchidaceae) orchids",
"authors": [
"Yoh Kok Hon",
"Christina Seok-Yien Yong",
"Janna Ong Abdullah",
"Rusea Go",
"Yoh Kok Hon",
"Christina Seok-Yien Yong",
"Janna Ong Abdullah"
],
"abstract": "Background: Coelogyne kaliana, Coelogyne stenochila and Coelogyne tiomanensis are three valuable rare orchid species endemic to Peninsular Malaysia, currently rampantly traded illegally via the internet and through local nurseries, which label them as hybrids to avoid enforcement detection. Drastic measures to ensure the continued existence of their populations in the wild should be introduced as they are rapidly diminishing into extinction, including the development of rapid and accurate species-specific identification tools. These three orchid species are highly similar morphologically and currently it is impossible to distinguish among them without their reproductive structures. Methods: RAPD-based species-specific SCAR markers were developed to distinguish and authenticate the identity of these three endemic Peninsular Malaysian Coelogyne species. Results: Three SCAR markers were successfully developed in this study. SCAR marker primer pair, CKL_f / CKL_r was specific to C. kaliana as it produced a unique single band of 271 bp but not in C. stenochila and C. tiomanensis. SCAR marker primer pair CST_f / CST_r amplified a single band of 854 bp in C. stenochila and two bands of different sizes (372 bp and 858 bp) in C. tiomanensis, but no amplification in C. kaliana. The third SCAR marker primer pair, CTI_f / CTI_r produced a single band (about 500 bp) for both C. stenochila and C. tiomanensis, but showed no amplification in C. kaliana. Conclusions: Although not all these SCAR markers were species amplification specific, they could be used to discriminate among the three Coelogyne species effectively. Accurate species identification is one of the most important steps to allow a proper management plan to be established in the effort to conserve these three endangered orchid species of Peninsular Malaysia. Besides, it could effectively put a stop to the illegal trading of these rare endangered orchid species worldwide.",
"keywords": [
"Coelogyne kaliana",
"Coelogyne stenochila",
"Coelogyne tiomanensis",
"endemic species",
"RAPD",
"SCAR",
"species identification",
"taxon delimitation"
],
"content": "Introduction\n\nThe genus Coelogyne L. belongs to the Orchidaceae family, which is comprised of about 200 sympodial species distributed throughout India, southwest China, southeast Asia and the Fiji Islands. Their main centres of diversity are Borneo, Papua New Guinea, Sumatra and the Himalayan range (Butzin, 1992; Gravendeel et al., 2001). There are 26 species of Coelogyne in Peninsular Malaysia based on the latest Checklist of Orchids of Peninsular Malaysia (Ong et al., 2017). Amongst them, Coelogyne kaliana, Coelogyne stenochila and Coelogyne tiomanensis are endemic to Peninsular Malaysia (Figure 1). They are mostly found in the highland regions at elevations of 1200 m above sea level. These three Peninsular Malaysian endemic species are very rare and are present in small populations. Endemic orchid species are national treasures, which have high commercial values among orchid collectors and enthusiasts. This leads to their illegal and indiscriminate collections from the wild. Moreover, the survival of many endemic species may be jeopardized due to rapid climate change, which causes their populations to decline gradually. This eventually would reduce the likelihood of them to being found in their natural habitats (Ong, 2013). Scientific and efficient discriminations of morphologically similar species have been indispensable in achieving the ultimate goal of preservation and conservation of endemic orchid species. Apart from their distinctive floral appearances, C. kaliana, C. stenochila and C. tiomanensis are morphologically similar with indistinguishable common vegetative structures. Thus, it is crucial that molecular markers be developed, which would enable us to easily distinguish among them taxonomically.\n\n(A) Coelogyne tiomanensis; (B) Coelogyne kaliana and (C) Coelogyne stenochila.\n\nRandom Amplified Polymorphic DNA (RAPD) is a polymerase chain reaction (PCR)-based method (Williams et al., 1990) that uses short and arbitrary oligonucleotides with GC contents of at least 50% to produce amplification products at random. The short oligonucleotide primer of approximately 10 bp in length is called a decamer and serves as both the forward and reverse primers. The amplification process in RAPD analysis is performed on total genomic DNA. Thus, it can be used to study genetic polymorphisms within the whole genome instead of just polymorphisms within a single genetic region. It also allows the amplification of random fragments of the genome without prior sequence knowledge. The main constraint of RAPD is its low fidelity. However, the conversion of RAPD markers into new, longer and specific Sequence Characterized Amplified Region (SCAR) markers can significantly improve reliability and reproducibility. Some other PCR-based methods, such as single locus microsatellites, Inter-Simple Sequence Repeats (ISSR) and Amplified Fragment Length Polymorphisms (AFLP), show better reproducibility of amplifications than the RAPD method. But, SCAR marker developments from these methods are often costlier, time consuming and laborious (Kumar & Gurusubramanian, 2011) when compared to its development from the RAPD method.\n\nThe SCAR marker (Paran & Michelmore, 1993) is a robust and reliable method used to detect and differentiate different samples by using specific primers derived from RAPD, ISSR, AFLP and other DNA markers. SCAR markers can discriminate among closely related samples or species by amplifying products of different sizes or no amplification in non-targeted samples and positive amplifications in targeted samples. Over the past few years, this method had been widely adopted in the identification of morphologically similar but genetically different organisms. A large number of robust and reliable SCAR markers had been successfully developed through the RAPD-based method for Pisum sativum (Srivastava et al., 2012), Oryza sativa (Semsang et al., 2013), Sorghum halepense (Zhang et al., 2013) and even for orchids, such as Phalaenopsis (Goh et al., 2005; Niknejad et al., 2009) and Paphiopedilum (Sun et al., 2011). However, currently there is still no genetic marker for the species identification of Peninsular Malaysia’s endemic Coelogyne species.\n\nHence, in this present investigation, RAPD-based SCAR markers were developed to distinguish among the three endemic Coelogyne species in Peninsular Malaysia. RAPD primers were used to screen the DNA samples of these three species to amplify reproducible species-specific bands. A single or a combination of two SCAR markers, which we developed, allowed us to successfully distinguish among the three morphologically similar and closely related Peninsular Malaysian endemic Coelogyne species. To our best knowledge, this is the first study on developing species-specific SCAR markers to identify the three endemic Coelogyne orchids of Peninsular Malaysia.\n\n\nMethods\n\nDue to the rarity of these species, only two samples of C. kaliana (YKH 020, Genting Highlands and YKH 004, Cameron Highlands, Malaysia), one sample of C. stenochila (YKH 031, Gunung Tahan, Malaysia) and one sample of C. tiomanensis (FRI 75329, Tioman Island, Malaysia) were collected from their native locations in Peninsular Malaysia. A small piece (3 cm × 3 cm) of fresh leaf from each sample was used for the DNA extraction. Genomic DNA was extracted based on the conventional CTAB method (Doyle & Doyle, 1987). The DNA pellets were suspended in 50 μl sterile DNase free water and kept at −20 °C until ready for use. The purity and concentration of the DNA samples were determined using a spectrophotometer (NanoDrop, USA).\n\nA total of 16 decamer universal RAPD primers (synthesized by First Base Laboratory, Serdang, Malaysia), as shown in Table 1, were used to amplify the DNA samples of the three Malaysian endemic Coelogyne species. These sequences were obtained from published articles (Table 1) where they have been used successfully in other orchid species. The RAPD amplification reactions were performed in a 10 μl volume containing 50 ng of plant DNA, 1.0 μM of RAPD primer (First Base Laboratory), 1 × PCR buffer (Promega, USA), 2.5 mM of MgCl2 (Promega, USA), 0.1mM dNTPs mix (Promega, USA), 0.5 units of Taq DNA polymerase (Promega, USA), using a thermal cycler machine (Eppendorf Master Cycler Gradient, Hamburg, Germany), The run conditions were: 1 cycle of initial denaturation at 95oC for 10 min; 45 cycles of denaturation at 95oC for 1 min, annealing at 35oC for 2 min and extension at 72oC for 2 min; and 1 cycle of final extension at 72oC for 10 min. The amplicons were electrophoresed on 1.2% TBE (Tris-borate-EDTA) agarose gel for 90 minutes at 70 V and stained with 0.5μg/mL ethidium bromide. The gel was visualised under a UV transilluminator (UVIdoc, USA) and the gel image was taken using a camera (UVItec, UK). Species-specific bands were identified from the bands produced during the RAPD amplifications.\n\nPrimers that consistently generated species specific PCR profiles in all the DNA samples were selected for primer design. The amplified PCR products were cloned and sequenced using the service provided by First Base Laboratory. All species-specific RAPD-generated PCR amplicons were cloned into pJET1.2/blunt cloning vector by chemical transformation into the E. coli competent cells. The white colonies, which contained the recombinant DNA (plasmid), were picked from LB/ampicillin/X-gal plates and the recombinant DNA was isolated from the bacterial culture. The targeted inserts were sequenced using SP6 and T7 universal sequencing primers.\n\nThe DNA sequences obtained were first subjected to nucleotide similarity searches using the BLASTN function of the NCBI database to check for significant similarities of the sequences with any sequences in GenBank Species-specific SCAR marker primer pairs were then designed based on the DNA sequences obtained. Species-specific SCAR marker primer pairs of 20-24 bases in length were designed with high stringency using the program Primer3 (Rozen & Skaletsky, 2000). Then, the designed SCAR marker primer pairs were synthesised by First Base Laboratory. Primer specificities were validated by amplifying the SCAR markers using the DNA samples of the three endemic Coelogyne species and by the same reaction mixture as was described in the section RAPD-PCR Amplification. The PCR profiles used for the amplifications of the designed SCAR primer pairs were: initial denaturation at 95oC for 10 min, 35 cycles of denaturation at 95oC for 1 min, annealing at 51 to 53oC (depending on the SCAR primer pair) for 2 min, extension at 72o C for 2 min, and a final extension cycle at 72oC for 10 min. The amplified PCR products were sequenced by First Base Laboratory, Serdang, Malaysia to confirm sequence specificities.\n\n\nResults\n\nIn the preliminary screenings of 16 random decanucleotide RAPD primers, all primers were able to amplify genomic DNA samples of Coelogyne kaliana, Coelogyne stenochila and Coelogyne tiomanensis, to produce a variable number of bands of different sizes. However, out of the 16 universal RAPD primers screened, only two RAPD primers, namely OPU 08 and OPU 12 showed a high level of consistency, producing distinct and reproducible fingerprint patterns in all the DNA samples tested. Amplification by primer OPU 08 consistently produced a clear band between 1 kb and 1.5 kb (Figure 2), which was unique to C. stenochila but absent in C. kaliana and C. tiomanensis. Primer OPU 08 also consistently amplified an intense band of roughly 750 bp that was specific to C. tiomanensis only and not found in C. stenochila and C. kaliana. Primer OPU 12 consistently amplified a distinct fragment of about 500 bp (Figure 2), which was present in all DNA samples of C. kaliana, but which was not observed in the DNA samples of C. stenochila and C. tiomanensis.\n\nLanes 1 and 2 are biological samples of C. kaliana (YKH004 and YKH020) while lanes 3 and 4 are technical replicates of YKH004 and YKH020 respectively; lane 5 is a biological sample of C. stenochila (YKH031) while lanes 6 to 8 are technical replicates of YKH031; lane 9 is a biological sample of C. tiomanensis (FRI75329) while lanes 10 to 12 are technical replicates of FRI75329. Lane M represents the 1.0 kb DNA ladder (Promega, USA). Fragments indicated by arrows were the specific bands, which were selected and subsequently sequenced for SCAR marker primer pair design.\n\nBased on the RAPD fingerprinting results of OPU 08 and OPU 12, very few species-specific bands were observed. However, in selecting species-specific fragments for cloning and sequencing, other than focusing on bright and clear monomorphic band, fragment size was also taken into consideration. The ideal fragment size usually should range from 500 bp to 1.5 kbp, as too large a fragment faced difficulty in the cloning process; while too small a fragment provided less sequence for designing the SCAR marker primer pair. Therefore, a fragment of between 1 kbp to 1.5 kbp amplified by the OPU 08 primer was selected for C. stenochila, while another of approximately 750 bp was selected for C. tiomanensis. A fragment of about 500 bp amplified by primer OPU 12 was chosen for C. kaliana (Figure 3).\n\nLanes 1 and 2 are biological samples while lanes 3 and 4 are technical replicates of C. kaliana; lane 5 is a biological sample while lanes 6 to 8 are technical replicates of C. stenochila; lane 9 is a biological sample while lanes 10 to 12 are technical replicates of C. tiomanensis. Lane M represents the 1.0 kb DNA ladder (Promega, USA). Fragments indicated by arrows were the specific bands selected and sequenced for SCAR marker primer pair design.\n\nSequencing of the selected fragments yielded a sequence of 517 bp for C. kaliana, 1207 bp for C. stenochila and 742 bp for C. tiomanensis respectively. Homology searches using BLASTN showed that the fragments did not have similarity with any known nucleotide sequences in the NCBI database. The SCAR primer pairs of 20 to 24 bases each were designed with high stringency to ensure the specificity for each endemic Coelogyne species using these primer pair sequences. Details of the three designed primer pairs are shown in Table 2.\n\nThe efficacy and specificity of the designed SCAR primers were screened for using DNA samples of the three endemic Coelogyne species. SCAR primer pair CKL_f and CKL_r amplified a single, clear, distinct and easily identifiable band of 271 bp only in DNA samples of C. kaliana, while no amplification was observed for C. stenochila and C. tiomanensis (Figure 4). Hence, the SCAR primer pair CKL_f and CKL_r is a species-specific marker for C. kaliana.\n\nLane M represents the 1.0 kb DNA ladder (Promega, USA).\n\nThe SCAR primer pair CST_f and CST_r, designed based on the specific fragment selected from the RAPD profile of C. stenochila, was not species amplification specific as anticipated. However, this SCAR marker still allowed the differentiation of the three endemic Coelogyne species by amplifying a single and bright band of 854 bp in C. stenochila but two bands of different sizes (858bp and 372 bp) in C. tiomanensis, and no DNA amplification in C. kaliana (Figure 5).\n\nLane M represents the 1.0 kb DNA ladder (Promega, USA).\n\nThe SCAR primer pair CTI_f and CTI_r, designed from the species-specific RAPD fragment for C. tiomanensis, was also not species amplification specific to C. tiomanensis. Nonetheless, it could still be used to differentiate C. kaliana from the other two endemic species, C. stenochila and C. tiomanensis. This SCAR marker amplified a single band of about 500 bp in both C. stenochila and C. tiomanensis but no DNA amplification in C. kaliana (Figure 6).\n\nLane M represents the 1.0 kb DNA ladder (Promega, USA).\n\n\nDiscussion\n\nBased on the results of the present study, selected bands amplified by universal RAPD primers were successfully converted into specific SCAR markers. These SCAR markers were able to be used for the rapid identification and authentication of three Peninsular Malaysian endemic Coelogyne species, Coelogyne kaliana, Coelogyne stenochila and Coelogyne tiomanensis. Endemic species refers to species that are found exclusively or confined to a restricted geographical area. Both Coelogyne kaliana and Coelogyne tiomanensis were named after the regions from which they were first found. Coelogyne kaliana was first discovered in Gunung Ulu Kali of Genting Highlands in Peninsular Malaysia and was described by Cribb in 1982. Coelogyne tiomanensis, first described by Henderson in 1930, is found exclusively in Gunung Kajang of Tioman Island. Coelogyne stenochila, first described by Hook in 1890, can only be found in Pahang and Selangor states (Seidenfaden & Wood, 1992). The endemism and rarity of these Coelogyne species make their conservation of utmost importance. The accurate identification of these three morphologically similar species is one of the first steps towards achieving this goal by effectively preventing their illegal trading.\n\nTo the best of our knowledge, this is the first reported RAPD-SCAR study on the three endemic Peninsular Malaysian Coelogyne species. Although not all the SCAR markers developed in this study were species amplification specific, they could still elucidate the identity of the three endemic Coelogyne species efficiently and correctly. This was impossible through the traditional vegetative morphological characteristics method usually used by botanists. The main constraint in this study was that the number of samples available for use was very few due to the rarity and protected status of these endemic species. Nonetheless, the SCAR markers developed in this study are new to science and are able to differentiate the three species accurately. This study further reinforced the robustness and suitability of SCAR makers as one of the leading methods for species identification of members of cryptic species complexes. This approach could also be used for other flagship species of economic and aesthetic importance. We hope that with the assistance of these markers, an exhaustive conservation plan could be developed in the immediate future to ensure the continued survival of these three rare endemic orchid species. Currently, the biggest threat to their survival is the high demand for them by orchid enthusiasts worldwide. This has resulted in them being obtained and traded illegally often by the sellers claiming that they are selling hybrid plants. It is expected that with the availability of the SCAR markers, which we have developed, this legal loophole can be successfully plugged.\n\n\nData availability\n\nSequences are available from GenBank:\n\nCoelogyne kaliana voucher YKH 004 maturase K (matK) gene, partial cds; chloroplast, Accession number MK398222.1: https://www.ncbi.nlm.nih.gov/nuccore/MK398222.1\n\nCoelogyne kaliana voucher YKH 020 maturase K (matK) gene, partial cds; chloroplast, Accession number MK398221.1: https://www.ncbi.nlm.nih.gov/nuccore/MK398221.1\n\nCoelogyne kaliana voucher YKH 004 ribulose-1,5-bisphosphate carboxylase/oxygenase large subunit (rbcL) gene, partial cds; chloroplast, Accession number MK398165.1: https://www.ncbi.nlm.nih.gov/nuccore/MK398165.1\n\nCoelogyne kaliana voucher YKH 020 ribulose-1,5-bisphosphate carboxylase/oxygenase large subunit (rbcL) gene, partial cds; chloroplast, Accession number MK398164.1: https://www.ncbi.nlm.nih.gov/nuccore/MK398164.1\n\nCoelogyne kaliana voucher YKH 004 tRNA-Leu (trnL) gene and trnL-trnF intergenic spacer, partial sequence; chloroplast, Accession number MK356248.1: https://www.ncbi.nlm.nih.gov/nuccore/MK356248.1\n\nCoelogyne kaliana voucher YKH 004 small subunit ribosomal RNA gene, partial sequence; internal transcribed spacer 1, 5.8S ribosomal RNA gene, and internal transcribed spacer 2, complete sequence; and large subunit ribosomal RNA gene, partial sequence, Accession number MK356194.1: https://www.ncbi.nlm.nih.gov/nuccore/MK356194.1\n\nCoelogyne kaliana voucher YKH 020 small subunit ribosomal RNA gene, partial sequence; internal transcribed spacer 1, 5.8S ribosomal RNA gene, and internal transcribed spacer 2, complete sequence; and large subunit ribosomal RNA gene, partial sequence, Accession number MK356193.1: https://www.ncbi.nlm.nih.gov/nuccore/MK356193.1\n\nCoelogyne kaliana voucher YKH 020 tRNA-Leu (trnL) gene and trnL-trnF intergenic spacer, partial sequence; chloroplast, Accession number MK356247.1: https://www.ncbi.nlm.nih.gov/nuccore/MK356247.1\n\nCoelogyne tiomanensis voucher FRI 75329 maturase K (matK) gene, partial cds; chloroplast, Accession number MK398239.1: https://www.ncbi.nlm.nih.gov/nuccore/MK398239.1\n\nCoelogyne tiomanensis voucher FRI 75329 small subunit ribosomal RNA gene, partial sequence; internal transcribed spacer 1, 5.8S ribosomal RNA gene, and internal transcribed spacer 2, complete sequence; and large subunit ribosomal RNA gene, partial sequence, Accession number MK356154.1: https://www.ncbi.nlm.nih.gov/nuccore/MK356154.1\n\nCoelogyne tiomanensis tRNA-Leu (trnL) gene and trnL-trnF intergenic spacer, partial sequence; chloroplast, Accession number MK356265.1: https://www.ncbi.nlm.nih.gov/nuccore/MK356265.1\n\nCoelogyne tiomanensis voucher FRI 75329 ribulose-1,5-bisphosphate carboxylase/oxygenase large subunit (rbcL) gene, partial cds; chloroplast, Accession number MK398182.1: https://www.ncbi.nlm.nih.gov/nuccore/MK398182.1\n\nCoelogyne stenochila voucher YKH 031 ribulose-1,5-bisphosphate carboxylase/oxygenase large subunit (rbcL) gene, partial cds; chloroplast, Accession number MK398181.1: https://www.ncbi.nlm.nih.gov/nuccore/MK398181.1\n\nCoelogyne stenochila voucher YKH 031 small subunit ribosomal RNA gene, partial sequence; internal transcribed spacer 1, 5.8S ribosomal RNA gene, and internal transcribed spacer 2, complete sequence; and large subunit ribosomal RNA gene, partial sequence, Accession number MK356153.1: https://www.ncbi.nlm.nih.gov/nuccore/MK356153.1\n\nCoelogyne stenochila voucher YKH 031 maturase K (matK) gene, partial cds; chloroplast, Accession number MK398238.1: https://www.ncbi.nlm.nih.gov/nuccore/MK398238.1\n\nCoelogyne stenochila tRNA-Leu (trnL) gene and trnL-trnF intergenic spacer, partial sequence; chloroplast, Accession number MK356264.1: https://www.ncbi.nlm.nih.gov/nuccore/MK356264.1\n\nOpen Science Framework: Development of species-specific SCAR markers for identification and authentication of three rare Peninsular Malaysian endemic Coelogyne (Orchidaceae) orchids, http://doi.org/10.17605/OSF.IO/35XBG (Go, 2020) (project registered on September 1st, 2020; https://osf.io/dxtb2).\n\nThis project contains the following underlying data:\n\nSequences resulted from cloning.\n\nUncropped/unedited images of RAPD fingerprint blots for the 16 decamer universal RAPD primers.\n\nUncropped/unedited images of SCAR primer pair blots.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Acknowledgements\n\nThe authors are also grateful to the Forest Research Institute Malaysia (FRIM) for providing the Coelogyne tiomanensis plant sample, and Forest Department of Peninsular Malaysia for giving us permission to access the study areas where the other samples were collected. Permit numbers are as follows: (a). National Park, Pulau Tioman - JPHL&TN(IP):80-4/2 Jld 17(11) dated 22 October 2013; (b) National Park, Gunung Tahan - JPHL&TN(IP):80-4/2 Jld 16 dated 3 June 2013.\n\n\nReferences\n\nBenner MS, Braunstein MD, Weisberg MU: Detection of DNA polymorphisms within the genusCattleya (Orchidaceae). Plant Mol Biol Rep. 1995; 13(2): 147–155. Publisher Full Text\n\nButzin F: Coelogyne Lindl. In F. G. Brieger, R. Maatsch, & K. Senghas (Eds.): Die Orchideen Berlin: Verlag Paul Parey. 1992; 919–940.\n\nDoyle JJ, Doyle JL: A rapid DNA isolation procedure from small quantities of fresh leaf tissue. Phytochemical Bulletin. 1987; 19: 11–15. Reference Source\n\nGo R: Development of species-specific SCAR markers for identification and authentication of three rare Peninsular Malaysian endemic Coelogyne (Orchidaceae) orchids. 2020. http://www.doi.org/10.17605/OSF.IO/35XBG\n\nGoh MW, Kumar PP, Lim SH, et al.: Random amplified polymorphic DNA analysis of the moth orchids, Phalaenopsis (Epidendroideae: Orchidaceae). Euphytica. 2005; 141(1–2): 11–22. Publisher Full Text\n\nGravendeel B, Chase MW, de Vogel EF, et al.: Molecular phylogeny of Coelogyne (Epidendroideae; Orchidaceae) based on plastid RFLPS, matK, and nuclear ribosomal ITS sequences: evidence for polyphyly. Am J Bot. 2001; 88(10): 1915–1927. Publisher Full Text\n\nGrünanger P, Caporali E, Marziani G, et al.: Molecular (RAPD) analysis on Italian taxa of theOphrys bertolonii aggregate (Orchidaceae). Plant Systematics and Evolution. 1998; 212(3–4): 177–184. Publisher Full Text\n\nKumar NS, Gurusubramanian G: Random amplified polymorphic DNA (RAPD) markers and its applications. Sci Vis. 2011; 11(3): 116–124. Reference Source\n\nLim S, Chye-peng Teng P, Lee Y, et al.: RAPD Analysis of Some Species in the Genus Vanda (Orchidaceae). Ann Bot. 1999; 83(2): 193–196. Publisher Full Text\n\nNiknejad A, Kadir M, Kadzimin S, et al.: Molecular characterization and phylogenetic relationships among and within species of Phalaenopsis (Epidendroideae: Orchidaceae) based on RAPD analysis. Afr J Biotechnol. 2009; 8(20): 5225–5240. Reference Source\n\nObara-Okeyo P, Kako S: Genetic diversity and identification of cymbidium cultivars as measured by random amplified polymorphic DNA (RAPD) markers. Euphytica. 1998; 99(2): 95–101. Publisher Full Text\n\nOng PT: Orchid Gems on Tioman. Conservation Malaysia Bulletin No. 2013; 13.\n\nOng PT, O’ Bryne P, Saw LG, et al.: Checklist of Orchids of Peninsular Malaysia. Research Pamphlet No. 2017; 136.\n\nParab GV, Krishnan S: Assessment of genetic variation among populations of Rhynchostylis retusa, an epiphytic orchid from Goa, India using ISSR and RAPD markers. Indian Journal of Biotechnology. 2008; 7: 313–319. Reference Source\n\nParan I, Michelmore RW: Development of reliable PCR-based markers linked to downy mildew resistance genes in lettuce. Theor Appl Genet. 1993; 85(8): 985–993. PubMed Abstract | Publisher Full Text\n\nRozen S, Skaletsky HJ: Primer3 on the WWW for general users and for biologist programmers. In S. Krawetz & S. Misener (Eds.): Bioinformatics Methods and Protocols: Methods in Molecular Biology Totowa, N. J.: Humana Press. 2000; 365–386. Reference Source\n\nSeidenfaden G, Wood JJ: The Orchids of Peninsular Malaysia and Singapore. Fredensborg: Olsen and Olsen. 1992; 13(3): 256–256. Reference Source\n\nSemsang N, Chundet R, Phanchisri B: Development of a SCAR Marker for Discrimination of a Thai Jasmine Rice (Oryza sativa L. cv. KDML105) Mutant, BKOS6, and Associated with Purple Color Trait in Thai Jasmine Rice-Related Varieties. Am J Plant Sci. 2013; 4(9): 1774–1783. Publisher Full Text\n\nSrivastava RK, Mishra SK, Singh AK, et al.: Development of a coupling-phase SCAR marker linked to the powdery mildew resistance gene ‘er1’ in pea (Pisum sativum L.). Euphytica. 2012; 186(3): 855–866. Publisher Full Text\n\nSun YW, Liao YJ, Hung YS, et al.: Development of ITS sequence based SCAR markers for discrimination of Paphiopedilum armeniacum, Paphiopedilum micranthum, Paphiopedilum delenatii and their hybrids. Sci Hortic. 2011; 127(3): 405–410. Publisher Full Text\n\nWilliams JG, Kubelik AR, Livak KJ, et al.: DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Res. 1990; 18(22): 6531–6535. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang W, Wei S, Yin L, et al.: RAPD Marker Conversion into a SCAR Marker for Rapid Identification of Johnsongrass [Sorghum halepense (L.) Pers.]. Not Bot Horti Agrobot Cluj Napoca. 2013; 41(1): 306. Publisher Full Text"
}
|
[
{
"id": "71987",
"date": "13 Oct 2020",
"name": "Mohd Norfaizal Ghazali",
"expertise": [
"Reviewer Expertise Plant Systematic",
"Plant Anatomy and Micromorphology",
"Palynology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript discussed the information and identification of Coelogyne kaliana, Coelogyne stenochila and Coelogyne tiomanensis (Orchidaceae) which are considered as three unique rare orchid species. Although the SCAR markers used in this study were not considered as species amplification specific, the information is quite useful in discriminating three Coelogyne species effectively. This paper is recommended for your intended journal, and the authors are required to amend the following as suggested below:\nMy recommendations:\nAbstract section\n\nInsert Orchidaceae after C. tiomanensis (para 2) Keywords: Coelogyne stenochila, C. tiomanensis, C. kaliniana – italic\n\nMethodology\nInsert the plant materials collection and details in a Table\n\nThe table should consist of these information:\n\n1. Date of collection 2. Scientific name 3. Localities 4. Collector’s name 5.Voucher number\nDiscussion section\n\nItalic all the scientific name in the section:\n\n‘Data Availability’ Underlying data\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "71757",
"date": "30 Nov 2020",
"name": "Jean Wan Hong Yong",
"expertise": [
"Reviewer Expertise Orchid biology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting, and timely paper pertaining to the molecular identification of three endemic Coelogyne (Orchidaceae) orchids in Malaysia. When these orchids are not flowering (vegetative stage), it is nearly impossible to identify them taxonomically.\nSequence Characterized Amplified Region (SCAR) markers, an improvement from RAPD, was used to discriminate successfully the three closely related Coelogyne species in Peninsular Malaysia.\nThe literature review carried out by the authors was adequate, appropriate, and organized in a logical fashion. Only minor adjustment is needed (see below).\nThe paper was well written and the experiments were conducted carefully. The sample size, research approaches (standard molecular techniques, following manufacturer’s protocols) and profile analyses, were appropriate.\nScientific suggestion My only technical critique is the lack of “outlier” in the analyses. The type species is Coelogyne cristata (may not be available to the authors). Perhaps, the authors could consider using a commonly available Malaysian Coelogyne (e.g. C. asperata) as an ‘“outlier” in future experiments as a form of control. Hence, the authors could discuss/explain the lack of a “outlier” in their experiments.\nFigure improvement A new figure 2 or an expanded Fig 1 with vegetative features (featuring leaf and pseudobulb) of the three Coelogyne species should be included, in addition to the (beautiful) flowers. See Plate 12 (four Coelogyne species featured, especially C. viscosa), page 732, The Orchids of Peninsular Malaysia and Singapore as a photographic guide1.\nText corrections \"Seidenfaden and Wood, (1992)\"1, the authoritative reference for orchids of Malaysia should feature prominently in the Introduction.\nThere are 26 species of Coelogyne in Peninsular Malaysia based on the latest Checklist of Orchids of Peninsular Malaysia2 (see also Seidenfaden and Wood, 1992)1.\nRecommendation This paper should be indexed with some minor revision.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1161
|
https://f1000research.com/articles/9-1255/v1
|
19 Oct 20
|
{
"type": "Software Tool Article",
"title": "Recursive Cluster Elimination based Rank Function (SVM-RCE-R) implemented in KNIME",
"authors": [
"Malik Yousef",
"Burcu Bakir-Gungor",
"Amhar Jabeer",
"Gokhan Goy",
"Rehman Qureshi",
"Louise C. Showe",
"Burcu Bakir-Gungor",
"Amhar Jabeer",
"Gokhan Goy",
"Rehman Qureshi",
"Louise C. Showe"
],
"abstract": "In our earlier study, we proposed a novel feature selection approach, Recursive Cluster Elimination with Support Vector Machines (SVM-RCE) and implemented this approach in Matlab. Interest in this approach has grown over time and several researchers have incorporated SVM-RCE into their studies, resulting in a substantial number of scientific publications. This increased interest encouraged us to reconsider how feature selection, particularly in biological datasets, can benefit from considering the relationships of those genes in the selection process, this led to our development of SVM-RCE-R. The usefulness of SVM-RCE-R is further supported by development of maTE tool, which uses a similar approach to identify microRNA (miRNA) targets. We have now implemented the SVM-RCE-R algorithm in Knime in order to make it easier to apply and to make it more accessible to the biomedical community. The use of SVM-RCE-R in Knime is simple and intuitive, allowing researchers to immediately begin their data analysis without having to consult an information technology specialist. The input for the Knime tool is an EXCEL file (or text or CSV) with a simple structure and the output is also an EXCEL file. The Knime version also incorporates new features not available in the previous version. One of these features is a user-specific ranking function that enables the user to provide the weights of the accuracy, sensitivity, specificity, f-measure, area under curve and precision in the ranking function, allowing the user to select for greater sensitivity or greater specificity as needed. The results show that the ranking function has an impact on the performance of SVM-RCE-R. Some of the clusters that achieve high scores for a specified ranking can also have high scores in other metrics. This finding motivates future studies to suggest the optimal ranking function.",
"keywords": [
"clustering",
"machine learning",
"recursive",
"gene expression",
"ranking",
"grouping",
"KNIME"
],
"content": "Introduction\n\nThe application of a variety of new technologies for measuring gene expression has generated publicly available datasets with very high feature dimensionalities (tens of thousands of genes)1,2. Because expression of certain groups of genes can be functionally related, they can be grouped according to a specific metric, which can be defined by the biological processes and interactions the group represents. Since most of the existing feature selection approaches have been borrowed from the field of computer science and statistics, they fail to consider the associations between gene expression features. We now propose to address that issue. In our initial study we suggested an algorithm called SVM-RCE3, where genes were grouped using a k-means based clustering algorithm. Our following study, SVM-RNE4 incorporated the possibility of grouping subsets of genes according to gene sub-networks. Our recent tool maTE5 suggested an alternative grouping based on microRNA targets and replaced k-means clustering with ensemble clustering6.\n\nSahu and Mishra7 have stressed the weakness of Signal-to-Noise Ratio (SNR) and t-statistics, which are widely used for gene rankings in the analysis of gene expression data, as using SNR and t-statistics as filtering techniques will likely select redundant features. They instead suggest that the genes are first grouped into clusters based on the similarities of their expression values, followed by the application of different filtering techniques to rank the genes in each cluster. The assigned ranks were then used to select the most informative genes from each cluster resulting in improved classification. The problem of dealing with clusters of features or groups of correlated features, in remote sensing data sets, was also recently addressed by Harris and Niekerk8. They stress the importance of first clustering the features by affinity propagation, and then applying a ranking function to overcome the weakness of the traditional feature selection approaches, which are likely to result in the selection of sub-optimal features.\n\n\nMethods\n\nThe SVM-RCE algorithm can be described by three main steps:\n\n1. The Clustering step combines the genes, based on expression, into groups using a clustering algorithm such as K-means. The merit of this step is to put genes with similar expression patterns into one cluster in order to deal with them together. In general, we refer to this step as a grouping function.\n\n2. The Rank step ranks each cluster using a function we have used in the SVM-RCE3 using Rank(X(S), f, r) as the average accuracy of the linear SVM over the data X represented by the S genes computed as f-folds cross validation repeated r times. We set f to 3 and r to 5 as default values (See Pseudocode 1).\n\n3. The RCE step removes the lower ranked clusters of genes and can be implemented to remove one cluster or a percentage of clusters as specified by the researcher, e.g. removing the lower 10% of the clusters.\n\nWe have applied the step of recursive cluster elimination based on the hypothesis that the clustering algorithm will generate new sets of clusters and that some of the genes will move between clusters and we have shown this to be the case.\n\nRanking Algorithm - R(Xs,M,f,r)\n\nXs: any subset of the input gene expression data X, the features are gene expression values\n\nM {m1,m2,...,mp} is a list of groups produced by k-means.\n\nf is a scalar (0≤f≤1): split into train and test data\n\nr: repeated times (iteration)\n\nres={} for aggregation the scores for each mi\n\nGenerate Rank for each mi, Rank(mi):\n\nFor each mi in M\n\nsmi=0;\n\nPerform r times (here r=5) steps 1̵5:\n\n1. Perform stratified random sampling to split Xs into train Xt and test Xv data sets according to f (here 80:20)\n\n2. Remove all genes (features) from Xt and Xv which are not in the group mi\n\n3. Train classifier on Xt using SVM\n\n4. t = Test classifier on Xv –calculate performance\n\n5. smi = smi + t;\n\nScore(mi)= smi /r ; Aggregate performance\n\nres= ∪i=1pscore(mi)\n\nOutput\n\nReturn res ( res = {Rank(m1),Rank(m2),…,Rank(mp)} )\n\nThe algorithm of Recursive Cluster Elimination3 considers clusters of similar features/genes and applies a rank function to each group as described in Pseudocode 1. Since we are using the clustering algorithm k-means we refer to these groups as clusters, but it could be any other biological or more general function that groups the particular features, such as KEGG pathways or microRNA targets, as we have suggested in several other studies4,5. As illustrated in Pseudocode 1, in the original code of SVM-RCE we used the accuracy as the determinant for ranking the clusters. The data for establishing that ranking was divided between training and testing. The data represented by each gene/feature is then assigned to a specific cluster and the rank function is then applied as the mean of r repeat times of the training-testing performance while recording different measurements of accuracy (sensitivity, specificity, etc.).\n\nIn this new version implemented in Knime9 we have incorporated more user specific ranking functions. The user provides the weights of the following ranking functions that correspond to the mean of each measurement achieved by the r times of the internal:\n\nWhere the acc is the accuracy, sen is the sensitivity, spe is the specificity, fm is the f-measurement, auc is the area under curve and pres is precision.\n\nThe coefficient weights represent the importance of each measurement for searching those clusters of genes that contribute to the final performance requirements. For example, if the user is interested in achieving greater specificity than sensitivity, the user would choose weights of 0.7 for the parameter spe and 0.3 for sen, stating that he is searching for clusters of genes that contribute to high specificity. However, one can also choose all the weights to be zero, with the weight of accuracy is set as 1, the rank function will then only rely on the accuracy.\n\nWe have used the free and open-source platform Knime10 for re-coding SVM-RCE (Figure 1–Figure 3) due to its simplicity and useful graphical presentations. Knime is a highly integrative tool that allows the user to include other programming languages such as R, Python and Java. In addition, one can also add external packages as such WEKA, H2O and so on. Figure 1 presents the workflow that includes SVM-RCE-R as a meta-node. The workflow can be executed on multiple input files. The node “List Files” will be indicated on the folder that has the input files. The workflow loops through those files and runs the SVM-RCE-R meta-node. The “Loop End” is also collecting specific results that can be subjected to further analysis.\n\n(a) The main Knime workflow for RCE based SVM that can be excuted on multiple input files. (b) The internal meta-node SVM-RCE-R that consisits of two compenents.\n\nThe SVM-RCE-R meta node consists of two components (two meta-nodes). The meta-node “Genes Filter t-test” (Figure 1b) is used to reduce the dimension of the features by applying the t-test to the training part of the data. Following that is the RCE component.\n\nThe interface of the SVM-RCE-R is presented in Figure 2. This part of the tool is used to set different parameters. The user can specify the number of iterations for Monte Carlo cross-validation (MCCV). MCCV is the process of randomly selecting (without replacement) some fraction of the data to form the training set, and then assigning the rest to the test set, by configuring the node “Counting Loop Start”. The node “Partitioning” is used to specify the ratio of the training/testing splitting.\n\nThe most important component “Rank Function Weights” is related to the rank function R(), where the user specifies the values of the weights w1,w2,..,w6. We show in the results section that these values have an impact on the performance of the SVM-RCE-R.\n\nFigure 3, meanwhile, shows nodes present in the meta-node SVM-RCE. It is designed so that it follows the pseudocode, thereby making it user-friendly.\n\nThe workflow was developed in KNIME which is compatible with Mac, Linux and Windows OS. We would recommend using a quad core CPU with at least 8 GB of RAM to run the workflow. Moreover, users will need to install Python 3 and R environments, Anaconda is recommended for the installation of Python 3 meanwhile R > 1.5 should be installed with Reserve package which can be found at https://cran.r-project.org/web/packages/Rserve/index.html.\n\n12 human gene expression datasets were downloaded from the Gene Expression Omnibus at NCBI11. For all datasets, disease (positive) and control (negative) data were available (Table 1). Those 12 datasets served to test the SVM-RCE-R tool and to compare its performance with two other approaches; the filter and embedded approaches12,13. The first approach performs feature selection using information gain (SVM IG) on the training part while the second approach is compared with SVM with recursive feature elimination (SVM-RFE)14. We have also implemented a workflow for SVM-RFE that is based on the Scikit-learn package15 in Knime.\n\nThe data sets are obtained from GEO. Each entry has the GEO code the name of the data, the number of samples and the classes of the data.\n\n\nResults\n\nFor the comparison of the three approaches, five datasets are considered for, as listed in Table 2. We have applied SVM-RCE-R, obtaining the performance over 100 iterations. At each iteration we have split the data into 90% for training and 10% for testing. The average of all different performance measurements is then aggregated. For additional comparison we refer to the first study published about SVM-RCE-R3.\n\nThe results indicate that SVM-RCE-R outperforms the other approaches in all the datasets except for GDS3646 with a case to control ratio of 5 to 116. SVM-RFE has a slightly better accuracy although significantly lower specificity than SVM-RCE-R.\n\nWe have considered different values of the rank function R(w_1,w_2,w_3,w_4,w_5,w_6) by specifying different values of the measurements weights, w1,..,w6 and have generated six rank functions as listed in Table 2. For each rank function we have applied the SVM-RCE-R obtaining the performance over 100 iterations. At each iteration we have split the data into 90% for training and 10% for testing. The average of all different performance measurements is then aggregated. A comparison between the performance of six different functions is listed in Table 3 and the results are shown in Figure 4.\n\nACC, accuracy; Spe, specificity; Sen, sensitivity; Auc, area under curve; F1, f-measurement.\n\nFigure 4 shows that there is deviation of the performance measurements for each R. However, we observed that the deviation is clear if we consider each data set individually, as presented in Figure 5.\n\nThe average of 100 iterations if computed for different performance measurements for each R1,…,R6 over the 12 datasets. The results of the level of cluster 2 is presented. #Genes is the average number of genes in level 2. The average accuracy, sensitivity, specificity and area under the curve (AUC) is presented for R1,..R6.\n\nThe results show the increase/decrease of those rank functions on the accuracy, sensitivity and specificity.\n\nIn order to examine the effect of the Rank function, we plotted the results obtained on the cluster level 2 as appears in Figure 5 (See Underlying data for all the results for the 12 datasets16) for each data set. For example, the accuracy obtained with R5 is significantly greater than R4 by about 12%, while reaching 4%–6% more than the other ranks. Interestingly we are getting a 4% improvement over the standard rank we have been using with the old version of SVM-RCE, which was R2.\n\nGDS2547 data reached an accuracy of ~79% applying R6 and 63% with R3, a difference of 16%, which is about 9% over the standard rank using the previous version SVM-RCE. However, for GDS5037 the max performance obtained with the standard rank R2 reached a difference of 16% over the minimum values reached by R5.\n\nWe have calculated the overall difference between the max value of each rank and the R2 that was used in the old version to get 5%.\n\nThis indicates that one can dramatically improve the performance of SVM-RCE-R by searching for the optimal values of the weights of the rank function.\n\nWe also conducted an additional experiment in order to examine the effect of gradually changing the values of sensitivity and specificity weights in the rank function. We ran two experiments on GDS3646 and GDS1962 data considering the values of (1,0) (0,1) (first argument is sensitivity weight while second one is specificity weight) increasing by 0.1 to reach (0,1) for the weights of sensitivity and specificity, respectively. The results are represented in Figure 6 for cluster level 2.\n\nThe axes labels are the values, for example sen01spe09 is associated for weight of 0.1 of sensitivity and 0.9 for specificity. The accuracy (ACC), sensitivity (Sen) and specificity (Spe) are plotted.\n\nFigure 6 shows that the two graphs are behaving differently over the set of weights, showing that the results depend on the specific data. Interestingly we see that for GDS1962 data, the optimal performance for all measurements is with 0.6 and weight 0.4 for sensitivity and specificity, respectively. Although the maximum accuracy is achieved over (0.1,0.9) weights pair, for GDS3646 data, the specificity at this point is very low and not usable for prediction, while (0.5,0.5) seems to provide reasonable performance for both sensitivity and specificity. Additionally, we have computed the number of common genes by considering the top 50 significant genes for each pair (sen01sep09 vs sen02spe08, …) having on average 11 genes. That is another indication that the rank function also has a significant impact on the list of the significant genes.\n\n\nDiscussion\n\nAs gene expression data sets become more complex, new computational tools that deal with features in a non-traditional way are needed to address this complexity. Our approach does not simply tackle the problem of inherent redundant or correlated features, it also suggests that defining the grouping metrics is equally important when searching that specific feature space that each researcher would like to focus on. Different biological systems/problems can require an output with a greater emphasis on either specificity, sensitivity or overall accuracy. Although specifying a certain metric, for instance, specificity, has higher priority during clustering, there can be cases where the clusters have high values for other metrics, which can be inferred from our results. Therefore, finding the optimal ranking will be one of the topics that we will further focus on. We now provide the capability to decide whether the specific problem being addressed will benefit more from reducing false positives or false negatives.\n\nThis new version of RCE now provides the user with the ability to control the analyses and to also design the ranking function that will allow exploration of the data in a way that addresses the specific goals of the analysis. Additionally, since it is easy to change the learning algorithm from SVM or to combine SVM with other machine learning algorithms, it further expands the utility of RCE-R. These additional components will be added to the next version of RCE as well as additional features for optimization procedures. Currently, our program estimates each cluster separately; a future version will combine different numbers of clusters using a search algorithm in order to identify the optimal combination that will return the highest accuracy.\n\n\nData availability\n\nHuman gene expression datasets from Gene Expression Omnibus, Accession numbers: GDS1962, GDS2519, GDS3268, GDS2547, GDS5499, GDS3646, GDS3874, GDS3837, GDS5037, GDS4516_GDS4718, GDS3900, GDS3929\n\nZenodo: Ajabeer/SVM-RCE-R-results-Omnibus-dataset: Supplementary Data for SVM-RCE-R. https://doi.org/10.5281/zenodo.403154616.\n\nThis project contains the following underlying data:\n\n- all_res1_clusters.xlsx files (contains the summary of all res_1 files for all 12 datasets for R1-R6)\n\n- logResults.csv files (contains the scoring values and class labels for each run of the SVM-RCE loop for each of the 12 datasets, R1-R6)\n\n- rankedGenes.xsx files (contains the names of the genes that ranked according to the rank function with their levels, rank function values and scores for each of the 12 datasets, R1-R6)\n\n- res1.xlsx files (contains the mean values of genes and the scoring metrics values calculated: Accuracy, Sensitivity, Specificity, F-measure, AUC, Cohens Kappa, for each cluster level for each of the 12 datasets, R1-R6)\n\n- res2.xlsx files (contains the number of genes for each level, scoring metrics values calculated: Accuracy, Sensitivity, Specificity, F-measure, AUC, Cohens Kappa, for each cluster for each iteration for each of the 12 datasets, R1-R6)\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).\n\n\nSoftware availability\n\nThe SVM-RCE-R Knime workflow, step-by-step tutorial and a detailed documentation are available on the following web site: https://malikyousef.com/svm-rce-in-knime/\n\nSource code available from: https://github.com/malikyousef/SVM-RCE-R-KNIME\n\nArchived source code at time of publication: https://zenodo.org/record/4066639#.X3sQVlLis2w9\n\nLicense: GNU General Public License v3.0\n\nDetailed terms and conditions of KNIME can be found at https://www.knime.com/downloads/full-license.",
"appendix": "References\n\nClough E, Barrett T: The Gene Expression Omnibus Database. Methods Mol Biol. 2016; 1418: 93–110. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrazma A, Sarkans U, Shojatalab M, et al.: ArrayExpress - A public repository for microarray gene expression data at the EBI. Nucleic Acids Res. 2003; 33(Database issue): D553–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYousef M, Jung S, Showe LC, et al.: Recursive Cluster Elimination (RCE) for classification and feature selection from gene expression data. BMC Bioinformatics. 2007; 8: 144. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYousef M, Ketany M, Manevitz L, et al.: Classification and biomarker identification using gene network modules and support vector machines. BMC Bioinformatics. 2009; 10: 337. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYousef M, Abdallah L, Allmer J: maTE: discovering expressed interactions between microRNAs and their targets. Bioinformatics. 2019; 35(20): 4020–4028. PubMed Abstract | Publisher Full Text\n\nAbdAllah L, Khalifa W, Showe LC, et al.: Selection of Significant Clusters of Genes based on Ensemble Clustering and Recursive Cluster Elimination (RCE). J Proteomics Bioinform. 2017; 10(8): 186–192. Publisher Full Text\n\nSahu B, Mishra D: A novel approach for selecting informative genes from gene expression data using Signal-to-Noise Ratio and t-statistics. In, 2011 2nd International Conference on Computer and Communication Technology, ICCCT-2011. 2011. Publisher Full Text\n\nHarris D, VanNiekerk A: Feature clustering and ranking for selecting stable features from high dimensional remotely sensed data. Int J Remote Sens. 2018; 39(23): 8934–8949. Publisher Full Text\n\nmalikyousef: malikyousef/SVM-RCE-R-KNIME: SVM-RCE-R (Version v1.0). Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4066639\n\nBerthold MR, Cebron N, Dill F, et al.: KNIME: The Konstanz Information Miner. In SIGKDD, Explorations. 2008; 319–326. Publisher Full Text\n\nBarrett T, Wilhite SE, Ledoux P, et al.: NCBI GEO: Archive for functional genomics data sets - Update. Nucleic Acids Res. 2013; 41(Database issue): D991–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPan W: A comparative review of statistical methods for discovering differentially expressed genes in replicated microarray experiments. Bioinformatics. 2002; 18(4): 546–554. PubMed Abstract | Publisher Full Text\n\nLazar C, Taminau J, Meganck S, et al.: A survey on filter techniques for feature selection in gene expression microarray analysis. IEEE/ACM Trans Comput Biol Bioinform. 2012; 9(4): 1106–19. PubMed Abstract | Publisher Full Text\n\nGuyon I, Weston J, Barnhill S, et al.: Gene Selection for Cancer Classification using Support Vector Machines. Machine Learning. 2012; 46: 389–422. Publisher Full Text\n\nPedregosa F, Varoquaux G, Gramfort A, et al.: Scikit-learn: Machine learning in Python. J Mach Learn Res. 2011; 12(85): 2825−2830. Reference Source\n\nAjabeer: Ajabeer/SVM-RCE-R-results-Omnibus-dataset: Supplementary Data for SVM-RCE-R (Version v1.0.0). Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4031546"
}
|
[
{
"id": "73564",
"date": "02 Dec 2020",
"name": "Marco Chierici",
"expertise": [
"Reviewer Expertise Bioinformatics",
"artificial intelligence",
"computational biology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors describe the Knime implementation of SVM-RCE-R, an algorithm for feature ranking and selection that is described as an improved version of the previously developed SVM-RCE. The motivation behind this new implementation is clearly stated and mainly resides in Knime’s user-friendliness and ease of use, besides Knime being an Open Source project. Because of this, the software has the potential to be a valuable tool for researchers in the computational biology community.\nMy first concern is about potential confusion: it is not clear whether the manuscript focuses on the Knime implementation of a previously devised algorithm (SVM-RCE-R) or whether it introduces both the algorithm and its implementation. I think this aspect should be better stated in both the Abstract and the Introduction.\nMy other concerns follow.\nConveyance of novelty:\nThe novelty of SVM-RCE-R vs. SVM-RCE, which, to my understanding, is the introduction of the ranking functions, should be pointed out more effectively. In particular, the Methods section opens with the description of the SVM-RCE algorithm: this comes unexpectedly to the reader since the manuscript should focus on describing SVM-RCE-R. The novelty element sneaks into the “Weighted rank function” subsection, while it should be better highlighted. This could be achieved by merely restructuring the Methods section or by adjusting the subsection titles.\n\nData and analysis details:\nIt has to be clearly stated which datasets are used for each experiment discussed in the Results section: as far as I have understood, the experiments concerning the ranking functions (e.g., Figure 4) consider all 12 datasets, while those comparing the three approaches (e.g., Table 3) consider five datasets. Still, Figure 5 presents results for three datasets while the main text states “for each data set”. The beginning sentence of Results can be misleading, as it may seem that all results are obtained on five datasets.\n\nWere the GEO datasets used “as is” or was some preprocessing applied? More details should be included regarding the input format: for example, is it a gene expression table? What should be on the rows and the columns? Should there be row and column names?\n\nPresentation of results:\nA comparison with SVM-RCE is only briefly mentioned, while the discussion would benefit from a more extended comparison (as done with SVM-RFE and SVM-IG), also to show SVM-RCE-R improvements over the previous version.\n\nThe authors state that “SVM-RFE has a slightly better accuracy although significantly lower specificity than SVM-RCE-R”, but according to Table 3 it appears that average SVM-RFE accuracy is always lower than, or equal to, average SVM-RCE-R accuracy, and the same for SVM-RFE specificity (except for one dataset): please elaborate. Moreover, how was the significance assessed?\n\nFor the ranking function, the authors chose accuracy, sensitivity, specificity, F1 score, AUC, and precision. This is quite a comprehensive set of metrics, which the authors could further enrich with the Matthews Correlation Coefficient (MCC), a balanced measure of accuracy and precision that can still be effectively used when sample classes are highly imbalanced.\n\nThe notation used in Table 3 should be improved: it is misleading to indicate the weights using the metric names to which they refer. For example, Acc=0.2 should be replaced by w1=0.2, and so on. Please refer to the rank function notation introduced in “Weighted rank function” and rename the weights accordingly.\n\nFigure 4: since the goal is to compare different ranking functions, and the ranking function includes F1 among its terms, I suggest adding the average F1 to the figure. The same holds for Figure 5.\n\nMisc and minors: In general, please consider having the paper proof checked by a native English speaker to improve overall readability and address typos.\nPlease consider improving the quality of Figures, especially 1-3.\n\nPlease indicate the measure of dispersion represented by the error bars in the Figures.\n\nI am afraid the following sentence is not correct: “As illustrated in Pseudocode 1, in the original code of SVM-RCE we used the accuracy as the determinant for ranking the clusters.” There is no mention of accuracy in Pseudocode 1, only “performance”. Moreover, in Pseudocode 1 please consider improving “t = Test classifier on Xv –calculate performance” as it is not clear, and please change “aggregation the scores” to “aggregating the scores”.\n\nIn Methods (p. 3): “test set performance” would be more accurate than “training-testing performance”.\n\nIn Methods (p. 4): probably “prec” should be used instead of “pres”.\n\nIn “Implementation in Knime”: the sentence “by configuring the node Counting Loop Start” more likely belongs to the end of the preceding paragraph.\n\nFigure 1 caption: please change “RCE based SVM” to “SVM-RCE” for consistency; please address the typo “compenents”.\n\nIn “Operation” (p. 5): “Reserve” should be changed to “Rserve”.\n\nIn Results (p. 5): please check “five datasets are considered for”.\n\nIn Results (p. 6): “... six rank functions as listed in Table 2”; it should be Table 3.\n\nIn Results (p. 9): please check “with 0.6 and weight 0.4”.\n\nPlease use either “meta node” or “meta-node” throughout the manuscript; the same holds for “SVM IG” vs SVM-IG.\n\nPlease use “area under the curve” instead of “area under curve” (also in Figures).\n\nPlease change w_1, w_2, … and w1,..,w6 to w1, w2, … (Results); please mind proper spacing and ellipsis in “w1,..w6”.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "6193",
"date": "05 Jan 2021",
"name": "Malik Yousef",
"role": "Author Response",
"response": "Dear Marco Chierici, Thank you for your constructive feedback. We found your insights helpful for improving the clarity of the manuscript. Below, we have addressed each of the items raised: My first concern is about potential confusion: it is not clear whether the manuscript focuses on the Knime implementation of a previously devised algorithm (SVM-RCE-R) or whether it introduces both the algorithm and its implementation. I think this aspect should be better stated in both the Abstract and the Introduction. We have updated the abstract as well as the introduction to describe the novelty of the user specific ranking function as well the simplicity of KNIME implementation. The novelty of SVM-RCE-R vs. SVM-RCE, which, to my understanding, is the introduction of the ranking functions, should be pointed out more effectively. In particular, the Methods section opens with the description of the SVM-RCE algorithm: this comes unexpectedly to the reader since the manuscript should focus on describing SVM-RCE-R. The novelty element sneaks into the “Weighted rank function” subsection, while it should be better highlighted. This could be achieved by merely restructuring the Methods section or by adjusting the subsection titles. In response to this suggestion, we have now changed the headings in the methods section. We now we have a specific subsection to bring the reader’s attention to the user specific ranking function. It has to be clearly stated which datasets are used for each experiment discussed in the Results section: as far as I have understood, the experiments concerning the ranking functions (e.g., Figure 4) consider all 12 datasets, while those comparing the three approaches (e.g., Table 3) consider five datasets. Still, Figure 5 presents results for three datasets while the main text states “for each data set”. The beginning sentence of Results can be misleading, as it may seem that all results are obtained on five datasets. We have now made it clear in the results section about which datasets were used for which figures. Moreover, it is clearly stated results that we have used all the datasets for our results and specific datasets were used for comparison. Were the GEO datasets used “as is” or was some preprocessing applied? More details should be included regarding the input format: for example, is it a gene expression table? What should be on the rows and the columns? Should there be row and column names? Based on your comment we have now included a description of the dataset as well. In addition, we have included the input data in our underlying data so that the results can easily be replicated. A comparison with SVM-RCE is only briefly mentioned, while the discussion would benefit from a more extended comparison (as done with SVM-RFE and SVM-IG), also to show SVM-RCE-R improvements over the previous version. We agree, the section comparing SVM-RCE and SVM-RCE-R is not clearly mentioned. We have now included a sub section to indicate to the readers the comparison results. The authors state that “SVM-RFE has a slightly better accuracy although significantly lower specificity than SVM-RCE-R”, but according to Table 3 it appears that average SVM-RFE accuracy is always lower than, or equal to, average SVM-RCE-R accuracy, and the same for SVM-RFE specificity (except for one dataset): please elaborate. Moreover, how was the significance assessed? We have corrected the mistake which in the results for Table 3 and we pointed out that only SVM-RCE-R performs weaker in sensitivity for two of the datasets otherwise it outperforms or is on par with other approaches. For the ranking function, the authors chose accuracy, sensitivity, specificity, F1 score, AUC, and precision. This is quite a comprehensive set of metrics, which the authors could further enrich with the Matthews Correlation Coefficient (MCC), a balanced measure of accuracy and precision that can still be effectively used when sample classes are highly imbalanced. We are planning to use the MCC in our further research where we search for the optimal combination of ranks. As you have mentioned it is a very good metric for highly imbalanced datasets. Thank you very much for the feedback about the metric. The notation used in Table 3 should be improved: it is misleading to indicate the weights using the metric names to which they refer. For example, Acc=0.2 should be replaced by w1=0.2, and so on. Please refer to the rank function notation introduced in “Weighted rank function” and rename the weights accordingly. We have now updated the results in the table so that it is consistent with what is stated in the description of the ranking function. Figure 4: since the goal is to compare different ranking functions, and the ranking function includes F1 among its terms, I suggest adding the average F1 to the figure. The same holds for Figure 5. We were focusing on the more widely used metrics to show the performance of our models, but as you have mentioned also looked into F-measure, so we now have included it in our results. All other typos and errors that you have so helpfully provided have been corrected in the revised manuscript. Kind regards, The authors"
}
]
},
{
"id": "73308",
"date": "02 Dec 2020",
"name": "Osman Ugur Sezerman",
"expertise": [
"Reviewer Expertise Bioinformatics"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this manuscript, the authors describe the implementation of feature selection method, SVM-RCE-R, in KNIME and demonstrate the usefulness of this tool. As they clearly describe, this is an improvement of the previously developed SVM-RCE. The most important novel feature in SVM-RCE-R is the user-specific ranking function, allowing the researcher to select for different performance metrics. As KNIME provides an easy-to-use interface, and the tool requires simple input formats, it will most likely be a valuable tool for biomedical researchers with many different backgrounds.\nMajor issues:\nThe introduction of the user-specific ranking function could be emphasized more clearly, perhaps in the Introduction section.\n\nThe analyzed datasets presented in the results section should be more clearly defined for each result, the number of datasets are confusing.\n\nThe abstract states that “The input for the Knime tool is an EXCEL file (or text or CSV) with a simple structure…”, this structure should be described in the main text.\n\nMinor issues:\nIn Table 3, the metric names should be replaced by the weights of these metrics.\n\nMinor typos:\n\nPage 3, last paragraph: it should be “user-specific”, “ranking functions” should be singular (i.e., “ranking function”). Page 4, in the user-specific ranking function, “pres” should be “prec”. Page 5, “Reserve” should be “Rserve”. Please use \"meta-node\" throughout the manuscript. “are under curve” should be “area under the curve”.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "6194",
"date": "05 Jan 2021",
"name": "Malik Yousef",
"role": "Author Response",
"response": "Dear Prof. Ugur Sezerman, We deeply appreciate your taking the time to provide your valuable feedback to us. We have revised the manuscript based on that feedback and would also like to address the comments: The introduction of the user-specific ranking function could be emphasized more clearly, perhaps in the Introduction section. We have now included a description of the novelty of our ranking function in the abstract to better focus on this topic. Moreover, we have also updated the introduction section with your recommendations. Finally, we have revised the methods section so that it clearly states and describes the user specific ranking function. The analyzed datasets presented in the results section should be more clearly defined for each result, the number of datasets are confusing. We have updated the results section and now clearly mention which datasets were used in each graph or table. We have also included how many datasets were used for comparison results and have clearly stated their names. The abstract states that “The input for the Knime tool is an EXCEL file (or text or CSV) with a simple structure…”, this structure should be described in the main text. In the data section, we now include a description of the input data, and we have also updated the underlying data for the input data files. As for the minor comments, all the relevant mistakes pointed out have been corrected and updated. We once again thank you for your feedback and contribution and we welcome any further feedback. Kind regards, The authors"
}
]
},
{
"id": "73698",
"date": "14 Dec 2020",
"name": "Abhishek Kumar",
"expertise": [
"Reviewer Expertise Multiomics",
"genomics and ML"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nYousef et al. have presented SVM-RCE-R as the improved implementation of the original algorithm SVM-RCE in Knime with additional features. The most important new feature is the ranking function R(). The authors demonstrate R() is important in improving the performance over the original study SVM-RCE. The Knime implementation demonstrates the novelty of the integration of biological information into the machine learning feature selection component. The approach is performing grouping by clustering and ranking by internal cross-validation. The approach is interesting and would also contribute to the new trend of integrative biological knowledge into the process of feature sections.\nMajor issues:\nI have two major concerns:\nThe authors must define the \"user-specific ranking function\" more clearly at the start of the manuscript.\n\nUsage of collected datasets are not clear and authors must provide better examples of results derived from datasets.\n\nMinor issues:\nThe authors must improve carefully figure and table legends plus typos and errors:\nTable 3 title: 'ACC=accuracy', 'Spe=specificity', and so on.\n\nFigure 4: \"The average of 100 iterations if computed” should be is 'computed'.\n\nResults section: \"For the comparison of the three approaches, five datasets are considered for,\" -> the last 'for' should be removed.\n\nFigure 3: It should be 'SVM-RCE-R' and not \"SVM-RCE\".\n\nFigure 1: It should be 'SVM-RCE-R' and not \"SVM-RCE\".\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "6204",
"date": "05 Jan 2021",
"name": "Malik Yousef",
"role": "Author Response",
"response": "Dear Dr. Abhishek Kumar, Thank you very much for your feedback suggestions, which significantly improve the manuscript and enrich the content. We would like to address your comments: The authors must define the \"user-specific ranking function\" more clearly at the start of the manuscript. We have now made changes based on the earlier reviewers as well to clarify and focus the manuscript more on the user-specific ranking function. We have made the relevant changes in the introduction, abstract as well as defining a more focused heading on the novel ranking feature. Usage of collected datasets are not clear and authors must provide better examples of results derived from datasets. We noticed that we were not clear in describing which datasets were used for which figures, therefore now we have mentioned in detail which datasets are used in their relevant tables and figures. Figure 4: \"The average of 100 iterations if computed” should be is 'computed'. We wrote it down as “if” since the user has the option of computing the algorithm to his/her needs and it can differ to their choices. We have also gone through all the typos and spelling mistakes that you have so kindly provided us, thank you very much. Please let us know if you further feedback. Yours truly, The authors."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1255
|
https://f1000research.com/articles/9-1149/v1
|
17 Sep 20
|
{
"type": "Research Article",
"title": "Alcohol consumption patterns among men who have sex with men in major cities of Myanmar: A cross-sectional study",
"authors": [
"Kyaw-Min Htut",
"Chitlada Areesantichai",
"Myo-Myo Mon",
"Chitlada Areesantichai",
"Myo-Myo Mon"
],
"abstract": "Background: Alcohol consumption patterns vary widely across the regions of the world. Although previous studies have focused on the sexual risk behaviours among men who have sex with men (MSM), studies regarding binge alcohol drinking among MSM in Myanmar are scarce. Methods: A cross-sectional study was conducted to identify the alcohol consumption patterns among MSM aged over 18 years in two major cities of Myanmar where the MSM population is higher than other regions. Purposive sampling was applied and sampling was made through Myanmar MSM network. Face-to-face interviews were conducted using a structured questionnaire. Patterns of alcohol consumption were described as frequency/percentage and mean/median as appropriate. Bivariate analysis was also done to find out the association between types of MSM and binge drinking. Results: A total of 256 MSM included in the study (mean age, 27.33±7.7 years). Of 256 participants, 225 MSM had the experience of alcohol consumption in their lifetime (225/256, 87.9%). Among ever drinkers, 152 MSM consumed alcohol within three months (152/225, 67.6%). Regarding beer consumption, the highest proportions of MSM from both groups (42.8%, 36.8%) consumed 1-3 times per week. Overall, 57.2% of young MSM and 41.2% of adult MSM consumed beer together with their friends. Nearly 34% of young MSM and nearly 38% of adult MMS consumed beer at gatherings of friends. At different time periods, higher proportions of Thange (partners of MSM) had experienced of binge drinking than apwint (open) and apone (hidden) (p<0.05). Conclusions: The current study identified the alcohol consumption patterns in terms of type, amount, frequency at different time periods among MSM in major cities of Myanmar. It is suggested to develop and implement alcohol control policy for MSM since the proportion of current drinkers as well as binge drinking higher among these groups.",
"keywords": [
"Men who have sex with men",
"alcohol",
"binge drinking",
"Myanmar"
],
"content": "Introduction\n\nGlobally, alcohol is widely consumed as a beverage and for recreation and socialization. Alcohol consumption patterns vary widely across the regions, ranging from daily heavy drinking to occasional hazardous drinking (Ennett et al., 2016; World_Health_Organization, 2018). Excessive alcohol use and chronic alcohol binging are associated with high morbidity and mortality (Stockings et al., 2016). About 40% of global population aged over 15 years had consumed alcohol while 2.3 billion of them are current drinkers. Worldwide, 44.8% of total recorded alcohol is consumed in the form of spirits. The second most consumed type of beverage is beer (34.3%) followed by wine (11.7%) (World_Health_Organization, 2018).\n\nTypes of alcoholic beverage varied among countries in South-East Asia region. Spirit was most common in Democratic People’s Republic of Korea (97%), India (92%), Sri Lanka (85%), Philippines (72%), Thailand (69%) and Myanmar (68%). Similarly, wine was the commonest type in Indonesia (76%), Nepal (49%), and Maldives (37%) while beer was common in Brunei Darussalam (100%), Vietnam (91%), Cambodia (88%), Singapore (70%), Timor-Leste (68%), and Malaysia (61%) (World_Health_Organization, 2018). The most common type of alcoholic beverage in Myanmar was spirits, which was consumed by 68% of drinkers (World_Health_Organization, 2018).\n\nPrevious studies have highlighted that hazardous alcohol drinking was common among men who have sex with men (MSM), ranging from incidences of 14% to 52%, and it was also associated with unsafe sexual behaviours (Davis et al., 2016; Herrera et al., 2016; Liu et al., 2016; Santos et al., 2018). Furthermore, a significant association was noted between hazardous alcohol drinking and sexually transmitted infections, including HIV. According to a study conducted in China, 14.4% of MSM reported hazardous drinking and 16.8% reported binge drinking. Hazardous or binge drinkers were associated with various risky sexual behaviours such as have multiple partnerships, pay for sex, and have condomless insertive anal intercourse (Liu et al., 2016). In a study in Peru, 45% of MSM and transgender women had an alcohol use disorder. Higher incidence of condomless anal intercourse was seen among participants with alcohol use disorder (AUD). However, AUD positivity was not associated with either condomless anal intercourse or recent STI/HIV infection (Herrera et al., 2016). Another study in US has documented the heavy alcohol use patterns and correlates in a diverse sample of MSM using respondent-driven sampling. It was found that prevalence of RDS adjusted weekly drinking was 24.9% and weekly binge drinking was 19.3%. Independent correlates of hazardous alcohol consumption were identified as being moderately or extremely interested in reducing alcohol use; ever receiving alcohol treatment; using ecstasy; reporting syphilis diagnosis; and having more than five male partners (Santos et al., 2018).\n\nIn Myanmar, HIV is concentrated among key affected populations like MSM, whereas HIV prevalence was over 10% among them (UNAIDS, 2015). Alcohol drinking was common among MSM, as it was with other men. Although previous studies have focused on risky sexual behaviours among MSM, few studies assessed hazardous alcohol drinking among MSM. Similarly, in Myanmar, we know of no studies to have reported on alcohol consumption among MSM. Therefore, current study was conducted to identify the alcohol consumption patterns and binge drinking at different time periods among men who have sex with men in Myanmar.\n\n\nMethods\n\nA cross-sectional study was conducted among MSM those aged more than 18 years in Yangon and Mandalay, major cities of Myanmar where MSM population is higher than other regions during June and July 2020.\n\nInclusion criteria\n\n- Self-identifying MSM;\n\n- MSM who had engaged in insertive or receptive anal sex or both;\n\n- MSM who have at least six months stay in Yangon and Mandalay\n\nExclusion criteria\n\n- MSM who are not mentally sound according to the records from the centre\n\n- MSM who do not understand and communicate in the Myanmar language\n\nPurposive sampling was applied and sampling of the MSMs was made through the Myanmar MSM network. Identification of the places for recruitment of the possible participants was made after discussion with the focal persons from the networks of MSM. There might be bias resulting from applying purposive sampling; however, attempts to reduce the bias were made by providing thorough explanations to the focal person to recruit different type of MSM from different sources, such as drop-in-centre and beauty parlours.\n\nBy considering the estimated proportion of MSM who are current drinkers as 20% (Oo et al., 2015), to achieve a 95% confidence level and an error of 5%, the minimum required sample size becomes 246 (Wayne, 1995).\n\nFirstly, a structured questionnaire was developed in English by reviewing the literature (see Extended data; Htut et al., 2020b). Then, translation was done into Myanmar language and back translation was carried out into English by a translator who was expert in both languages and had experience of translation regarding questionnaire used in MSM related research. Training of the interviewers was done at Department of Medical Research and pre-test was done at a non-study township in Yangon Region. After receiving ethical approval, at the venues where MSMs usually gather like drop-in-centers of International Non-governmental Organizations, beauty parlors, etc., eligible participants were contacted and invited to participate in the study. After getting the informed consent, data collection was done by face-to-face interview. Strict adherence to ethical principles were ensured throughout the data collection period in order to maintain the confidentiality of the information of the study participants.\n\nAccording to the local terminology in Myanmar, three groups of MSM were included in the study: Apwint or open type, Apone or hidden type and Tha Nge (National_AIDS_Program, 2019). Apwint or open MSM are defined as individuals born biological male but who openly express themselves femininely by dress and/or social interactions. Apone or hidden MSM are defined as individuals born biological males who may also want to express themselves femininely but may not disclose this behaviour to all segments of their social networks. Tha Nge are defined as having a masculine outward appearance but have sex with men”.\n\nTypes of alcohol beverages consumed by MSM included Beer, Wine and Rum.\n\nData entry was carried out with EpiData version 3.1 and data analysis was done with SPSS version 21. Exploratory data analysis was done to check the errors, consistencies and missing values. The number of standard drinks was calculated by volume of container in liters multiplied by the percentage of alcohol volume multiplied by 0.789 (the specific gravity volume of ethyl alcohol). Binge drinking was defined as five or more standard drinks for men in a sitting or within two hours. Descriptive statistics were shown according to the data obtained from the assessment. Patterns of alcohol consumption were described as frequency/percentage and mean/median as appropriate. Bivariate analysis was also done to find out the association between types of MSM and binge drinking.\n\nThe proposal was submitted to Institutional Review Board, Department of Medical Research, Myanmar (Ethics/DMR/2020/036). Written informed consent was taken from the participants after thorough explanation about the objectives of the study. Confidentiality and anonymity of the information were strictly ensured. All answer sheets and data reports were kept in locked cabinet.\n\n\nResults\n\nA total of 256 MSM were included in the study, of whom 151 participants were residents of Yangon and 105 participants were residents of Mandalay. Mean age of MSM was 27.33±7.7 years and ranged from 18 to 57 years. Tables were presented according to the age group that was categorized as 15–24 years as young MSM and ≥25 years as adult (older) MSM. As shown in Table 1, 39.1% of 15–24 years age group and 58.9% of ≥25 years age group were “apwint” (open). Regarding their education status, 52.7% of the younger age group and 39% of the older age group had attended education up to high school level. Over 46% of young MSM and over 66% of adult MSM have regular income earning job. Median monthly income was 200,000 MMK in both groups. Over 33% of young MSM were private/government staff while over 35% of adult MSM were running their own business. Individual-level responses from each participant are available as Underlying data (Htut et al., 2020a)\n\nOf the 256 participants, 225 had experience of alcohol consumption in their lifetime (225/256, 87.9%). Among ever drinkers, 152 had consumed alcohol within the past three months (152/225, 67.6%). The amount and frequency of different types of alcohol consumption within three months in terms of amount, frequency, with whom they drink together, time of consumption and reasons are shown in Table 2. Regarding beer consumption, the mean amount consumed was 4.1±2.5 standard drinks by young MSM and 4.5 ± 3.0 standard drinks by adult MSM (ethanol concentration of 41 and 45 grams, respectively), highest proportions of MSM from both groups (42.8%, 36.8%) consumed 1–3 times per week.\n\nTable 3 shows the types of person that MSM drink with, the reasons and timings of different types of alcohol consumption by age group of MSM within three months. Over 57.2% of young MSM and 41.2% of adult MSM consumed beer together with their friends. A majority of younger and older MSM consumed beer in the evening or at night (96% and 94.2%, respectively). Nearly 34% of young MSM and nearly 38% of adult MMS consumed beer for the reason of a friends’ gathering.\n\nTable 4 shows that binge drinking was associated with type of MSM. At different time periods, higher proportions of Thange (partner of MSM) had experienced of binge drinking than apwint (open) and apone (hidden), and the association was statistically significant (p<0.05).\n\n\nDiscussion and recommendation\n\nWorldwide, the prevalence of alcohol consumption was 43% in general population while about half has never consumed alcohol. Of all participants in present study, nearly 90% have ever consumed alcohol in their life time and over 50% were current drinkers within one month. Previous studies have documented the alcohol consumption among general population in Myanmar. They reported the prevalence of 50% lifetime drinkers and 20% current drinkers. The proportion of ever drinkers in current study was much higher than those from previous studies done in two different townships in Myanmar (Oo et al., 2015; Win & Areesantichai, 2014). These studies focused on the general adult population, which might differ from the MSM population. These differences in study population and time periods might contribute to the discrepancy. Regarding the types of alcoholic beverages, as reported in the report of World Health Organization, common types of alcoholic beverages consumed in Myanmar included spirits, beer and wine. Likewise, beer, whisky and wine were the common alcoholic beverages stated by MSM in current study.\n\nStudies in China and other countries also highlighted alcohol consumption among MSM and general population but using different screening tools. Additionally, previous studies have documented the prevalence of alcohol consumption at different time periods (Liu et al., 2016; Lu et al., 2019). In present study, nearly 70% of MSM had consumed alcohol within the past 3 months and 36.2% of Apwint (open type), 35.3% of Apone (hidden type) and 62.8% of Thange (sexual partners of MSM) had experience of binge drinking. In a large-scale study in China, over 56% of 3,588 MSM had consumed alcohol in the past 3 months and 17% were binge drinkers (Liu et al., 2016). Though recent alcohol consumption within 3 months among MSM was similar between the studies, the proportion of binge drinkers was higher in present study.\n\nIn Peru, 45% of MSM had an alcohol use disorder and over 90% were hazardous drinkers (Herrera et al., 2016). Another study in China among a general adult population aged 18 to 34 years also identified the prevalence of alcohol consumption as about 45% (Lu et al., 2019). Similarly, another study in US among MSM documented that nearly 14% and 25% were monthly and weekly binge drinkers (Santos et al., 2018). Unlike other studies in China, Fan et al. (2016) documented that 23% of MSM had consumed a drink containing alcohol in the previous year while 7% were heavy alcohol drinkers.\n\nThe current study also identified the proportion of drinkers and binge drinkers in different time periods (one month, three months, six months, one year and lifetime). and over one third to nearly two third among different types of MSM reported binge drinking within three months. A higher proportion of participants drinking alcohol was noted in current study than in previous studies among MSM population. Differences in background sociocultural conditions may contribute to this discrepancy. (Liu et al., 2016; Lu et al., 2019).\n\nThere was a limitation in the present study that should be acknowledged. Alcohol consumption patterns were self-reported and their behaviours could not be validated with other methods, such as observation. However, we tried to overcome this limitation by carefully explaining the objectives of the study to allow participants to answer with accurate responses.\n\nThe development and implementation of an alcohol control policy for MSM should be considered, since over half of them were current drinkers (within one month). Between one-third and three-quarters of them had binge drinking at different time periods, which could lead to adverse health and social consequences.\n\n\nData availability\n\nFigshare: msm_data_alcohol.sav. https://doi.org/10.6084/m9.figshare.12911510.v1 (Htut et al., 2020a).\n\nThis project contains the de-identified underlying data from each participant for the present study.\n\nFigshare: Questionnaire_Alcohol consumption patterns among men who have sex with men in major cities of Myanmar: A cross-sectional study. https://doi.org/10.6084/m9.figshare.12926636.v1 (Htut et al., 2020b).\n\nThis project contains the questionnaire used in the present study.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "References\n\nDavis A, Kaighobadi F, Stephenson R, et al.: Associations Between Alcohol Use and Intimate Partner Violence Among Men Who Have Sex with Men. LGBT health. 2016; 3(6): 400–406. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEnnett ST, Jackson C, Cole VT, et al.: A multidimensional model of mothers' perceptions of parent alcohol socialization and adolescent alcohol misuse. Psychol Addict Behav. 2016; 30(1): 18–28. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFan W, Lu R, Wu G, et al.: Alcohol drinking and HIV-related risk among men who have sex with men in Chongqing, China. Alcohol. 2016; 50: 1–7. PubMed Abstract | Publisher Full Text\n\nHerrera M, Konda K, Leon S, et al.: Impact of alcohol use on sexual behavior among men who have sex with men and transgender women in Lima, Peru. Drug Alcohol Depend. 2016; 161: 147–154. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHtut KM, Areesantichai C, Mon MM: msm_data_alcohol.sav. figshare. Dataset. 2020a. http://www.doi.org/10.6084/m9.figshare.12911510.v1\n\nHtut KM, Areesantichai C, Mon MM: Questionnaire_Alcohol consumption patterns among men who have sex with men in major cities of Myanmar: A cross-sectional study. figshare. Journal contribution. 2020b. http://www.doi.org/10.6084/m9.figshare.12926636.v1\n\nLiu Y, Ruan Y, Strauss SM, et al.: Alcohol misuse, risky sexual behaviors, and HIV or syphilis infections among Chinese men who have sex with men. Drug Alcohol Depend. 2016; 168: 239–246. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLu W, Xu J, Taylor AW, et al.: Analysis of the alcohol drinking behavior and influencing factors among emerging adults and young adults: a cross-sectional study in Wuhan, China. BMC Public Health. 2019; 19(1): 458. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNational_AIDS_Program: Myanmar Integrated Biological and Behavioral Surveillance Survey & Population Size Estimates among MSM 2015. 2019. Reference Source\n\nOo WM, Aung MS, Soe PP, et al.: Alcohol consumption among adult males in urban area of Thanlyin Township, Yangon Region, Myanmar. Int J Res Med Sci. 2015; 3(11): 3192–3196. Publisher Full Text\n\nSantos GM, Rowe C, Hern J, et al.: Prevalence and correlates of hazardous alcohol consumption and binge drinking among men who have sex with men (MSM) in San Francisco. PLoS One. 2018; 13(8): e0202170. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStockings E, Hall WD, Lynskey M, et al.: Prevention, early intervention, harm reduction, and treatment of substance use in young people. Lancet Psychiatry. 2016; 3(3): 280–296. PubMed Abstract | Publisher Full Text\n\nUNAIDS: Situation Analysis of the HIV Response among Men who have Sex with Men and Transgender Persons in Myanmar. 2015. Reference Source\n\nWayne W: Biostatistics: A Foundation of Analysis in the Health Sciences. John Wiley&Sons. In: Inc. 1995. Reference Source\n\nWin SMS, Areesantichai C: Pattern of Alcohol Drinking among Adults in Pha-An Township, Myanmar. J Health Res. 2014; 28(Suppl): S41–S45. Reference Source\n\nWorld_Health_Organization: Global status report on alcohol and health 2018: World Health Organization. 2018. Reference Source"
}
|
[
{
"id": "71570",
"date": "23 Sep 2020",
"name": "May Soe Aung",
"expertise": [
"Reviewer Expertise Public Health",
"Epidemiology",
"Maternal and Reproductive Health",
"Infection Control",
"NCDs",
"Biostatistics",
"Environmental Health."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIt is an interesting article which highlighted the alcohol consumption patterns among men who have sex with men. It was accurately worked by current literature to fulfill the study objectives. Suitable statistical methods are applied and interpretations are appropriate. For more clarification, reviewer’s comments are provided as follows: Abstract\nRegarding beer consumption, type of grouping should be mentioned precisely instead of “both groups”.\n\nClear description of the variables: “Together with their friends” and “at gathering of friends” is required at the start of article because repeated use of same word may cause a bit confusing.\n\nIn this study, only amount of alcohol consumption was found out by different time periods according to Table 4. However, it was mentioned that type and frequency were also identified by different time periods. Therefore, conclusion is necessary to be coincided with the finding.\n\nThere was no comparison of MSM with other specific groups in this article. Therefore, the proportion of current drinkers as well as binge drinking should not be concluded by the term, “higher”.\nIntroduction\nDescription of RDS in the second last paragraph is more appropriate.\nMethods\nAs the pretesting was done in non-selected township, number of townships for purposive sampling from Yangon and Mandalay Cities should be mentioned.\n\nAre there any other places of recruiting participants? Because “etc.” was added after drop-in-centers of International Non-governmental Organizations and beauty parlors in data collection.\n\nRegarding three groups of MSM, similar word formatting should be used according to the literature because there is no consistency between abstract and operational definitions.\n\nTypes of alcoholic beverages mentioned in operational definitions section included “RUM” but following tables in results section were shown by “Whisky”. Therefore, it is better to explained about the discrepancy of usage if possible.\n\nReferences should be cited for calculation of the number of standard drinks. Is it similar to amount of alcohol use?\n\nIn data management and analysis, statistical test for association and level of statistical significance are needed to be mentioned.\nResults\nIn second paragraph, there was mentioning about the types of person that MSM drink with, the reasons and timings of consumption although Table 2 was shown amount and frequency of alcohol use.\n\nThe calculation of ethanol concentration by mean of grams in relation to mean amount of standard drinks should be explained.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "5971",
"date": "08 Jan 2021",
"name": "Kyaw-Min Htut",
"role": "Author Response",
"response": "Thank you for your expert opinions and approval on our manuscript. Comments from reviewer Abstract 1. Regarding beer consumption, type of grouping should be mentioned precisely instead of “both groups”. Response: Thanks for your comments and “both groups” means young age group (15-24 yrs) and adult age group (>24 yrs) 2. Clear description of the variables: “Together with their friends” and “at gathering of friends” is required at the start of article because repeated use of same word may cause a bit confusing. Response: Sorry for making you confusing. “Together with their friends” is one of the responses for the question \"With whom they consumed alcohol\". Similarly, “at gathering of friends” is the response for the reason of alcohol consumption. 3. There was no comparison of MSM with other specific groups in this article. Therefore, the proportion of current drinkers as well as binge drinking should not be concluded by the term, “higher”. Response: Proportion of current drinkers among MSM from current study was higher than the previous studies conducted among general population in Myanmar Methods 1. As the pretesting was done in non-selected township, number of townships for purposive sampling from Yangon and Mandalay Cities should be mentioned. Response: The participants were recruited through Myanmar MSM network and it is not possible to mention the number of townships. 2. Are there any other places of recruiting participants? Because “etc.” was added after drop-in-centers of International Non-governmental Organizations and beauty parlors in data collection. Response: Others included the hotspot places of MSM were they usually gather such as home of famous beautician, moat and office of community based organization. 3. Regarding three groups of MSM, similar word formatting should be used according to the literature because there is no consistency between abstract and operational definitions. Response: I think three groups of MSM (Apwint, Apone and Thange) were used. 4. Types of alcoholic beverages mentioned in operational definitions section included “RUM” but following tables in results section were shown by “Whisky”. Therefore, it is better to explained about the discrepancy of usage if possible. Response: Thank you for your comment. Whisky was missed to mention in the operational definition."
}
]
},
{
"id": "75393",
"date": "11 Dec 2020",
"name": "Inn Kynn Khaing",
"expertise": [
"Reviewer Expertise Diabetes",
"Aging."
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIt is good idea to study alcohol consumption which is very popular among men who have sex with men (MSM) but it is not clearly described why the authors want to choose target population of this group as well as aim of this study. We suggest that the prevalence of alcohol drinking among men in Myanmar should be stated in general so the prevalence in this group will be higher or not can be determined.\n\nThe study design is suitable and sample size is adequate. We recommend that age range should be included in abstract. (18-57 yrs) We aware that the authors used 15-24 yrs in (Table -1 ) but young MSM should be 18-24yrs (Table -1) because the study was done over 18 yrs of age in this study.\n\nThe authors tried to explore the data as much as possible but it is difficult to get exact data of alcohol consumption among MSM. Out of 256 participants, 225 participants consumed alcohol in their lifetime, 89.7%, but the authors did not specify about others who were more than 18 yrs of age or not. (Never drinking alcohol in their lifetime?) These data only indicates alcohol drinking but no impact. We recommend that if the authors can correlate the data with some points like the prevalence of sexually transmitted disease among them, it will be more scientific and reproducible. We also doubtful about the lowest salary among MSM is only 1,5000 MMK per month. If the salary is too low, it will affect the alcohol consumption.\n\nJustify for using bivariate analysis of Binge drinking according to the category of MSM at different time periods. It is not relevant with Table -4 results in discussion: “over one third to nearly two third among different types of MSM reported binge drinking within three months”.\n\nAs we mentioned above, it is difficult to guarantee for full reproducibility. It will need to be more scientific study and outcome for reliable data. But, we understood that there would be some constraints because of COVID-19 pandemic.\n\nWe suggest that there should be some outcome for conclusion for alcohol consumption among MSM. After describing the results, the authors could not discuss properly about binge drinking among different category of MSM in different time periods and it does not highlight how different time periods of binge drinking have an influence on different category of MSM. So it was difficult to get conclusion from the results. It is not a good conclusion.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "6258",
"date": "11 Jan 2021",
"name": "Kyaw-Min Htut",
"role": "Author Response",
"response": "According to comment 1, the prevalence of alcohol drinking among men in Myanmar had described as \"According to a previous study in Myanmar, there was 50% of lifetime drinkers and 20% of current drinker among general men.\" According to the comment 2, I have edited 18-24 yrs instead of 15-24 yrs in every sentence and tables. Comment number 3: “These data only indicates alcohol drinking but no impact” Response: It does not include any impact of alcohol consumption as the study does not focus and include in our study objectives. Comment number 3: “We recommend that if the authors can correlate the data with some points like the prevalence of sexually transmitted disease among them, it will be more scientific and reproducible.” Response: Thanks so much for your suggestion and we agree that these points are very important to discuss. But we don’t have such kind of data because it is out of our study scope. Comment number 3 of reviewer 2: “We also doubtful about the lowest salary among MSM is only 1,5000 MMK per month. If the salary is too low, it will affect the alcohol consumption.” Response: It is not the salary. We just ask about their income in terms of money they received in anyway because some of them did not have regular job. According to comment 4, we did bivariate analysis between binge drinking and type of MSM since there would be difference in behaviors according to the types of MSM \"It was detected that binge drinking was more common among Thange than other two types of MSM at different time periods showing higher risk among that particular group of MSM\" and \"Emphasis should be done more on the partners of MSM as higher proportion of them had practice of binge drinking\" were added according to your 5th and 6th comments."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1149
|
https://f1000research.com/articles/8-701/v1
|
21 May 19
|
{
"type": "Study Protocol",
"title": "A recall-by-genotype study on polymorphisms in the TMPRSS6 gene and oral iron absorption: a study protocol",
"authors": [
"Momodou W. Jallow",
"Susana Campino",
"Andrew M. Prentice",
"Carla Cerami",
"Momodou W. Jallow",
"Susana Campino",
"Andrew M. Prentice"
],
"abstract": "Background: Oral iron supplementation is commonly used to treat and prevent anaemia. The transmembrane protease serine 6 gene (TMPRSS6), which encodes matriptase 2, is a negative regulator of hepcidin, the key controller of iron homeostasis. Genome-wide association studies (GWAS) have identified several single nucleotide polymorphisms (SNPs) in the TMPRSS6 gene that are associated with an increased risk of iron-deficiency anaemia. We will investigate the in vivo effects of three previously reported TMPRSS6 variants (rs855791, rs4820268 and rs2235321) on oral iron absorption in non-anaemic volunteers in The Gambia. Methods: A recall-by-genotype study design will be employed. Pre-genotyped participants will be recruited from the West African BioResouce (WABR), which currently contains over 3000 genotyped individuals. Male and female volunteers will be selected based on polymorphisms (rs855791, rs4820268 and rs2235321) in the TMPRSS6 gene in the Gambian population. The effects of a single variant allele at one SNP and the additive effect of two or three variant alleles from either two or all three SNPs will be investigated. Study participants will be given a single oral dose of 400mg ferrous sulfate, and blood samples will be collected at baseline, two hours and five hours post supplementation. Differences in iron absorption between genotype groups will be assessed by measuring the increase in serum iron concentration at five hours post iron ingestion. Discussion: This study will increase understanding of the role of genetic variations in TMPRSS6 on oral iron absorption in subjects of West African origin. This will test for the biological basis for the association of each of the three TMPRSS6 variants with iron absorption. This may help in guiding future iron intervention strategies, particularly in populations with a high frequency of these SNPs and a high frequency of anaemia. Study registration: ClinicalTrials.gov NCT03341338 14/11/17.",
"keywords": [
"recall-by-genotype",
"iron supplementation",
"anaemia",
"TMPRSS6",
"hepcidin regulatory genes",
"genetic variants."
],
"content": "Abbreviations\n\nAGP: alpha-1-acid glycoprotein, CRP: c-reactive protein, EDTA: ethelenediametelenetatraacetic acid, FBC: full blood count, G6PD: glucose-6-phosphate dehydrogenase, GWAS: genome-wide association study, Hb: haemoglobin, IRIDA: iron-refractory iron deficiency anaemia, KWLPS: Kiang West Longitudinal Population Study; LSHTM: London School of Hygiene & Tropical Medicine, MAF: minor allele frequency, MRCG: Medical Research Council The Gambia, SNP: single nucleotide polymorphism, sTfR: soluble transferrin receptor, TMPRSS6: transmembrane protease serine 6, TSAT: transferrin saturation, UIBC: unsaturated iron binding capacity, WABR: West Africa BioResource, WK: West Kiang\n\n\nIntroduction\n\nDespite aggressive implementation of iron supplementation programs, either alone or in combination with food-based supplementation, the prevalence of anaemia remains high in low- and middle-income countries1,2. The World Health Organisation (WHO) has set 2050 as a target date by which the current anaemia burden will be reduced by half. In order to achieve this goal, it will be important to identify the major drivers of anaemia.\n\nThe transmembrane protease serine 6 gene (TMPRSS6), which encodes for matriptase-2, is one of the negative regulators of hepcidin3, the key iron homeostasis regulator4. When serum iron levels are low, matripase-2 suppresses hepcidin expression, allowing more iron from the diet to be absorbed through the intestines into the bloodstream5,6. A single nucleotide polymorphism (SNP) in the TMPRSS6 gene can lead to decreased expression or inactivation of matripase-27, which would then lead to inappropriately elevated hepcidin levels, inhibited iron absorption and would thereby result in an increased risk of anaemia5.\n\nMultiple SNPs in the TMPRSS6 gene have been linked to iron-refractory iron deficiency anaemia (IRIDA), a hereditary anaemia that is not responsive to oral iron supplementation8. In addition, many SNPs in TMPRSS6 (including rs855791, rs4820268 and rs3345321) have been linked to an increased risk of iron deficiency anaemia (IDA) in genome-wide association studies (GWAS)9–11. In Caucasian populations, rs855791 has been reported to be in strong linkage disequilibrium (LD) with rs4820268 (r2=0.83) and rs2235321 (r2=0.44)12. Similarly, in Asian populations, rs855791 is reported to be in high LD with rs4820268 (r2=0.65)12.\n\nThe minor allele frequency (MAF) of these SNPs varies between racial and ethnic groups. In African populations, the MAF of rs855791 is lower (10%) than in East Asians (57%), South Asians (54%) and Europeans (39%)13. Similarly, the MAF of rs4820268 is lower in Africans (28%) compared to Europeans (42%), whereas, the MAF of rs2235321 in Africans (41%) is similar to that of the European population (42%)13. The effects of these SNPs (rs855791, rs4820268 and r2235321) on iron absorption and hepcidin levels in Subsaharan African populations has not been studied.\n\nWe hypothesize that the variant alleles at these SNPs may impair iron absorption and may be partially responsible for the disproportionately high anaemia prevalence in sub-Saharan Africa. Here, we propose to investigate effects of these three TMPRSS6 SNPs on oral iron absorption in Gambian adults.\n\nWe anticipate that this study will provide a biological insight into the association of these three TMPRSS6 variants with anaemia.\n\n\nProtocol\n\nThe primary objective of this study is to assess the impact of single and multiple copies of variant alleles of the TMPRSS6 SNPs (rs855791, rs4820268 and rs2235321) on oral iron absorption. The primary outcome measure will be the change in serum iron concentration before and five hours after a single 400 mg dose of ferrous sulfate iron given orally (Figure 1).\n\nSecondary endpoints related to the primary objective are:\n\n(1) Increase in transferrin saturation (TSAT) above baseline after a single oral 400 mg dose of ferrous sulfate iron.\n\n(2) Increase in serum unbound iron binding capacity (UIBC) above baseline after a single oral 400 mg dose of ferrous sulfate iron.\n\n(3) Increase in serum hepcidin levels above baseline after a single oral 400 mg dose of ferrous sulfate iron.\n\n(4) Ferritin, haemoglobin, mean corpuscular volume (MCV) and soluble transferrin receptor (sTFR) at baseline, as measures of iron status.\n\n(5) White blood cell count (WBC), granulocyte count, C-reactive protein (CRP) and alpha-1-acid glycoprotein (AGP) at baseline, as measures of the inflammatory state.\n\n(6) Sickle cell haemoglobin and glucose 6-phosphatase deficiency (G6PD) status at baseline to assess potential confounding effects of these two genetic conditions, which are common in this population.\n\nWe will employ a recall-by-genotype study design, in which participant selection will be based on TMPRSS6 SNPs reported to be associated with the risk of iron-deficiency anaemia: rs855791, rs4820268 and rs223532110,14,15. We will utilize the West African BioResouce (WABR), which contains the Kiang West Longitudinal Population Study (KWLPS) as the basis for selection of pre-genotyped participants16.\n\nThe proposed study will be conducted within the population of West Kiang (WK) District, in the Lower River Region of The Gambia, and study procedures will be conducted at the Medical Research Council The Gambia (MRCG) at London School of Hygiene & Tropical Medicine (LSHTM), Keneba Field Station16. Individuals that are eligible for the study but have moved to the coastal region of The Gambia will be followed-up by a fieldworker and study procedures will be conducted at the MRCG Fajara site. Participants currently residing in WK will be prioritised.\n\nA total of 300 participants (male and female) will be recruited. Participants will be chosen based on three TMPRSS6 SNPs (rs855791, rs4820268 and rs2235321), from which we will generate nine genotype combinations, as detailed in Table 1. This will allow the investigation of the effect of each SNP individually and in combination. Composite genotype group 3 is the control group with no variant alleles. Due to the low MAF of rs855791 in our study population, we are unable to include homozygotes for the variant allele. This limited the selection of genotype combinations, and only nine combinations had sufficient participants to include in the study.\n\nFor inclusion, participants must be 18 years and above, in good physical health, have available genotype data, be able to fast overnight prior to the study visit and be able to give informed consent. Individuals will be excluded from the study if they have any signs of infection at the time of enrolment, are severely anaemic (Hb <7 g/dl), pregnant or breastfeeding, or have a positive malaria test at screening.\n\nThe total sample size will be 300. This will include approximately 62 wild type subjects and an average of 31 in each of the eight variant genotype groups. This study size will be able to detect a 12% mean decrease in serum iron at five hours after oral iron supplementation between the wild type and the variant genotype groups with 90% power and a type 1 error of 0 in this study.\n\nPotential participants with the candidate composite genotypes of interest will be selected from the study database by the principal investigator, and contact details (including address and phone number) will be extracted from the WK Demographic Surveillance System16 by the study data manager. Participants will be contacted either in person or by telephone. Participants who provide informed consent will be invited to the study site where the rest of the study procedures will be conducted, as summarised in Figure 2.\n\nWABR = West Africa Bioresource.\n\nEach participant will be given a single dose of 400mg ferrous sulfate oral iron (2x 200mg ferrous sulfate tablets), equivalent to 130mg elemental iron. To ensure that the iron tablets are taken, a nurse will observe and record the time injestion. Participants will be asked to stay at the study site until the study is completed, which is after collecting the five hour post supplementation blood sample (Figure 1).\n\nAll data generated from this study will be anonymised by allocating a unique study ID to each participant. Screening, enrolment and sample collection details will be collected in standard study forms and entered into the study database. Data will be double-entered by two data entry clerks and verified by a data supervisor.\n\nIn order to prevent bias in treatment, the composite genotype of individuals will not be disclosed to the study team (data management, field and clinical staff). In addition, participants will be recruited in groups at random, and individuals with different composite genotype groups will be mixed during study visits.\n\nA 3ml whole blood sample will be collected at baseline. 2.5ml will be collected in lithium heparin tubes. 500µl will be collected in EDTA (ethylenediaminetetraacetic acid) micro tubes to be used for full blood count (FBC), malaria rapid testing and sickle screening.\n\nPost supplementation blood samples (3ml blood sample in lithium heparin tube) will be collected at two hours and five hours following iron ingestion. Pre- and post-supplementation blood samples in lithium heparin tubes will be spun and the plasma aliquoted in barcode-labelled tubes and stored at -20°C for iron biomarker analysis.\n\nFBC will be analysed using a 3-part haematology analyser (Medonic M-series, Boule Medical, Sweden). Iron biomarkers [serum iron, unsaturated iron binding capacity (UIBC), ferritin, soluble transferrin receptor (sTfR), haptoglobin (HP)] and inflammatory markers [C-reactive protein (CRP) and alpha-1-acid glycoprotein (AGP)] will be measured using a Cobas Integra 400 plus biochemistry analyser (Roche Diagnostics). Total iron binding capacity and transferrin saturation of iron (TSAT) will be calculated from serum iron and UIBC. Plasma hepcidin levels will be measured using a commercially available ELISA (DRG Instruments GmbH, Germany). The sickle rapid test will be analysed using the sodium metabisulphide method and positive samples will be genotyped by Hb electrophoresis. G6PD deficiency will be assessed using a qualitative enzyme assay (G6PD Hb+ R&D Diagnostics).\n\nPrimary analysis will be to assess the change in serum iron between the composite genotype groups at the five hours post-supplementation time point. A linear model will be fitted with genotype group as the independent variable and serum iron or TSAT as response variables and genotype group as the main predictor, with the inclusion of age, sex and inflammation status (CRP and AGP levels) as covariates. Using the same approach, we will also examine the effect of genotype on secondary outcome measures. The baseline iron level of the participants may vary. All secondary analysis are exploratory.\n\nIn order to remove this potential source of bias, we will adjust for baseline serum iron in the regression analysis. If the missing data rate is more than 5%, we will consider imputation. The follow-up duration is short; thus, we expect little bias from loss to follow-up. We will also consider sensitivity analysis, fitting a multivariate regression model where the main outcomes of interest (including TSAT, iron and hepcidin) will be jointly regressed to the same set of predictors.\n\nThis study has been approved by the MRC Unit The Gambia at the LSHTM Scientific Coordinating Committee, MRC Unit The Gambia at the LSHTM / Gambia Government Joint Ethics Committee (SCC1429), and the LSHTM Ethics Committee (LSHTM Ethics reference number 11679). A trained field worker will visit each potential study participant to issue an information sheet detailing the purpose and nature of the study (see Extended data)17. Individuals who cannot read will have the information sheet translated into a language they understand by the fieldworker, in presence of an independent witness. Furthermore, participants will be given the opportunity to ask questions to the investigators that they deem important. Participants will be informed that they are free to withdraw from the study anytime, and they can further raise any question about the study with the investigators.\n\nParticipants will provide written informed consent, and those who cannot write will provide a thumbprint prior to enrolling into the study. Confidentiality of study participants will be protected by anonymising all study samples and forms by allocating a study number to each participant.\n\nThis study was retrospectively registered with ClinicalTrials.gov (NCT03341338) on 14th November 2017.\n\nThe study results will be published in relevant peer-reviewed journals and key findings will be presented at international scientific meetings. Data sharing will be in agreement with the MRC policy on research data sharing.\n\nThe study is in the data collection phase at the time of publication.\n\n\nDiscussion\n\nGWAS has identified several genetic variants associated with iron status3,11,15,18–20. However, detailed understanding of genotype-phenotype relationships is required to identify their effects on iron absorption. The recall-by-genotype (RbG) study design is an efficient tool for detailed investigations of genotype-phenotype relationships because it minimizes confounders and improves statistical power while reducing sample size21. In this study, we will use the RbG study design to assess the functional effects of the three common TMPRSS6 variants on iron absorption. We expect that this study will provide new insights into the association between these TMPRSS6 gene variants and oral iron absorption in a population where anaemia prevalence is high.\n\n\nData availability\n\nNo underlying data are associated with this article\n\nFigshare: Jallow et al. Patient Information sheet and consent form.docx. https://doi.org/10.6084/m9.figshare.8058959.v217\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).",
"appendix": "Grant information\n\nThe work was undertaken with support from core funding obtained for the MRC International Nutrition Group from the UK Medical Research Council (MRC) and Department for International Development (DFID) under the MRC/DFID Concordat (PI: Andrew M Prentice).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe authors wish to acknowledge Dr. Branwen Hennig for her prior work on the KSWLPS and mentorship, and Dr. Laura Corbin for critically reading the manuscript.\n\n\nReferences\n\nKassebaum NJ, GBD 2013 Anemia Collaborators: The Global Burden of Anemia. Hematol Oncol Clin North Am. 2016; 30(2): 247–308. PubMed Abstract | Publisher Full Text\n\nPasricha SR, Drakesmith H: Iron Deficiency Anemia: Problems in Diagnosis and Prevention at the Population Level. Hematol Oncol Clin North Am. 2016; 30(2): 309–25. PubMed Abstract | Publisher Full Text\n\nDe Falco L, Silvestri L, Kannengiesser C, et al.: Functional and clinical impact of novel TMPRSS6 variants in iron-refractory iron-deficiency anemia patients and genotype-phenotype studies. Hum Mutat. 2014; 35(11): 1321–9. PubMed Abstract | Publisher Full Text\n\nGanz T: Systemic iron homeostasis. Physiol Rev. 2013; 93(4) 1721–1741. PubMed Abstract | Publisher Full Text\n\nFinberg KE, Whittlesey RL, Fleming MD, et al.: Down-regulation of Bmp/Smad signaling by Tmprss6 is required for maintenance of systemic iron homeostasis. Blood. 2010; 115(18): 3817–3826. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDu X, She E, Gelbart T, et al.: The serine protease TMPRSS6 is required to sense iron deficiency. Science. 2008; 320(5879): 1088–92. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSilvestri L, Pagani A, Nai A, et al.: The serine protease matriptase-2 (TMPRSS6) inhibits hepcidin activation by cleaving membrane hemojuvelin. Cell Metab. 2008; 8(6): 502–511. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDe Falco L, Sanchez M, Silvestri L, et al.: Iron refractory iron deficiency anemia. Haematologica. 2013; 98(6): 845–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSal E, Keskin EY, Yenicesu I, et al.: Iron-refractory iron deficiency anemia (IRIDA) cases with 2 novel TMPRSS6 mutations. Pediatr Hematol Oncol. 2016; 33(3): 226–32. PubMed Abstract | Publisher Full Text\n\nPelusi S, Girelli D, Rametta R, et al.: The A736V TMPRSS6 polymorphism influences hepcidin and iron metabolism in chronic hemodialysis patients: TMPRSS6 and hepcidin in hemodialysis. BMC Nephrol. 2013; 14: 48. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDelbini P, Vaja V, Graziadei G, et al.: Genetic variability of TMPRSS6 and its association with iron deficiency anaemia. Br J Haematol. 2010; 151(3): 281–284. PubMed Abstract | Publisher Full Text\n\nChambers JC, Zhang W, Li Y, et al.: Genome-wide association study identifies variants in TMPRSS6 associated with hemoglobin levels. Nat Genet. 2009; 41(11): 1170–2. PubMed Abstract | Publisher Full Text | Free Full Text\n\n1000 Genomes Project Consortium, Auton A, Brooks LD, et al.: A global reference for human genetic variation. Nature. 2015; 526(7571): 68–74. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGichohi-Wainaina WN, Tanaka T, Towers GW, et al.: Associations between Common Variants in Iron-Related Genes with Haematological Traits in Populations of African Ancestry. PLoS One. 2016; 11(6): e0157996. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAn P, Wu Q, Wang H, et al.: TMPRSS6, but not TF, TFR2 or BMP2 variants are associated with increased risk of iron-deficiency anemia. Hum Mol Genet. 2012; 21(9): 2124–2131. PubMed Abstract | Publisher Full Text\n\nHennig BJ, Unger SA, Dondeh BL, et al.: Cohort Profile: The Kiang West Longitudinal Population Study (KWLPS)-a platform for integrated research and health care provision in rural Gambia. Int J Epidemiol. 2017; 46(2): e13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJallow MW, Campino S, Cerami C, et al.: Jallow et al Patient Information sheet and consent form.docx. 2019. http://www.doi.org/10.6084/m9.figshare.8058959.v2\n\nBenyamin B, Ferreira MA, Willemsen G, et al.: Common variants in TMPRSS6 are associated with iron status and erythrocyte volume. Nat Genet. 2009; 41(11): 1173–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcLaren CE, Garner CP, Constantine CC, et al.: Genome-wide association study identifies genetic loci associated with iron deficiency. PLoS One. 2011; 6(3): e17390. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNai A, Pagani A, Silvestri L, et al.: TMPRSS6 rs855791 modulates hepcidin transcription in vitro and serum hepcidin levels in normal individuals. Blood. 2011; 118(16): 4459–4462. PubMed Abstract | Publisher Full Text\n\nCorbin LJ, Tan VY, Hughes DA, et al.: Formalising recall by genotype as an efficient approach to detailed phenotyping and causal inference. Nat Commun. 2018; 9(1): 711. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "48820",
"date": "21 Jun 2019",
"name": "Dale R. Nyholt",
"expertise": [
"Reviewer Expertise Genetics"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study protocol aims to test associations between three variants (rs855791, rs4820268 and rs2235321) in TMPRSS6 gene and serum iron concentration after an oral iron supplementation. While this study is interesting and the findings are of potential importance, there are some concern/suggestions highly recommended to be addressed by the authors.\nIt has been suggested that a linear model will be fit with genotypes as the outcome, serum iron as the predictor, and age, sex, and inflammation status biomarkers as covariates. As it has been well established that serum iron levels can be influenced by environmental factors such as smoking (Ghio et al. 2008)1, alcohol consumption (Whitfield et al. 2001)2 and also BMI (Shattnawi et al. 2018)3, the authors should consider collecting these measures and incorporating them into a more complex regression model.\nBlood collection can be conducted in a detailed plan considering participants’ diet, time of day, and so forth. More importantly, the authors could improve serum iron measurement by considering a fasting blood test with the pre-defined fasting period such as 12 hours.\nIn ‘Sample size calculation’ section author claimed that ‘This study size will be able to detect a 12% mean decrease in serum iron at five hours after oral iron supplementation between the wild type and the variant genotype groups with 90% power and a type 1 error of 0 in this study.’ Please elaborate on this sentence and explain how a 12% mean decrease in serum iron was estimated (e.g. please state the assumed parameters such as mean and standard deviation of serum iron).\nTo further understand causal connection between rs855791, rs4820268 and rs2235321 SNPs and iron serum level, or iron-deficiency anaemia, the authors will need to conduct molecular experiments on mRNAs and proteins to experimentally identify direct effect of the listed SNPs on gene expression and then relate the expression level and corresponding genotypes to the trait, in this case serum iron level. In addition, a huge accessibility of GWAS, eQTL and pQTL studies enables the authors to perform in-silico analysis to verify their protocol and also hypothesise new ideas. Herein, I summarised some of GWAS results relevant to these three SNPs which can be furthered investigated in this proposed study.\nWhile rs855791 SNP is a missense variant at TMPRSS6 gene, it is associated with protein levels of TFRC (Sun et al. 2018)4 and transcript expression of ALAS2 in blood tissue (Westra et al. 2013)5. rs4820268 SNP is a synonymous coding variant of TMPRSS6 gene, but again it is recognised as trans-eQTL of ALAS2 in blood tissue (Westra et al. 2013)5. rs2235321 SNP is another synonymous coding variant of TMPRSS6 gene and the SNiPA tool reports that it is associated with neither complex traits (e.g. iron status biomarkers) nor transcript/protein expression (Arnold et al. 2014)6. Altogether, it seems that rs855791 and rs4820268 variants have an impact on ALAS2 expression that is most highly expressed in bone marrow tissue (Fagerberg et al. 2014)7 and contributes in heme metabolism and iron homeostasis (Barman-Aksözen et al. 2015)8.\nLast, it is necessary to make sure that this study will potentially add novel findings into the literature in which several studies focused on TMPRSS6 polymorphisms and iron related traits to date (Nalado et al. 20199; Sørensen et al. 201910). The authors may consider to focus on knowledge gaps and aim to comprehensively relate gene variants to gene expression/activity and then serum iron level. Notably, as a functional analysis, population-based differences are not much interesting because the functional effects of variants less likely vary from population to population.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? No\n\nAre the datasets clearly presented in a useable and accessible format? Yes",
"responses": []
},
{
"id": "48819",
"date": "02 Jul 2019",
"name": "Alida Melse-Boonstra",
"expertise": [
"Reviewer Expertise Nutritional science",
"nutrient bioavailability"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nMajor comments:\nIn the introduction the following is stated: ‘We hypothesize that the variant alleles at these SNPs may impair iron absorption and may be partially responsible for the disproportionately high anaemia prevalence in sub-Saharan Africa.’ This, however, contradicts the lower or at most equal MAF of the TMPRSS6 variants in African populations as compared to Caucasian populations. Especially the low MAF of rs855791 is important since it has so far shown the strongest inverse association with haemoglobin and iron concentrations in Caucasian and Asian populations. It seems to me that the conclusion then should be that TMPRSS6 variants are less likely to play a major role in the development of anaemia in African populations, unless other variants than the ones under study are responsible for such an effect.\nTable 1: Since authors already have selected the gene combinations based on numbers available in their database of 3000 genotyped individuals, they might as well already indicate the number of subjects available for each gene variant combination.\nExclusion criteria: It would seem more appropriate to exclude participants with sickle cell anaemia or G6PD deficiency instead of using it as a co-variate in the analysis. However, if this reduces the numbers of each genotype combination too much, authors may as well include them while doing a retrospective sensitivity analysis.\nThe primary outcome measure will be the change in serum iron concentration before and five hours after a single 400 mg dose of ferrous sulfate iron given orally. This is not the best way to measure iron absorption, which would ideally be assessed with a stable isotope method. Past studies have shown that change in serum iron concentration cannot be used as a measure of iron bioavailability at the individual level. Authors should provide references that back up the validity of their approach.\nMinor comments:\nUnder study procedures, second paragraph: ‘injestion’ is miss-spelled. Sample size calculation: A type 1 error of 0 seems to be ideal, yet unrealistic. Is this a typo?\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Partly\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Not applicable",
"responses": [
{
"c_id": "6033",
"date": "04 Jan 2021",
"name": "Carla Cerami",
"role": "Author Response",
"response": "Comments from reviewer 2: Comment 1: In the introduction the following is stated: ‘We hypothesize that the variant alleles at these SNPs may impair iron absorption and may be partially responsible for the disproportionately high anaemia prevalence in sub-Saharan Africa.’ This, however, contradicts the lower or at most equal MAF of the TMPRSS6 variants in African populations as compared to Caucasian populations. Especially the low MAF of rs855791 is important since it has so far shown the strongest inverse association with haemoglobin and iron concentrations in Caucasian and Asian populations. It seems to me that the conclusion then should be that TMPRSS6 variants are less likely to play a major role in the development of anaemia in African populations, unless other variants than the ones under study are responsible for such an effect. Response 1: Thank you for this comment. The statement “and may partially be responsible for disproportionately high anaemia prevalence in sub-Saharan Africa” has now been changed to “these and other genetic variations may contribute to the high anaemia prevalence in sub-Saharan Africa”. We agree that the MAF of rs855791 is low in African populations (MAF=0.1). In fact, we will not be able to address the impact of rs855791 in this study precisely because of this low MAF. However, the other two SNPs in this study, rs4820268 and rs3345321, each have a MAF>0.3 in this population. Comment 2: Table 1: Since authors already have selected the gene combinations based on numbers available in their database of 3000 genotyped individuals, they might as well already indicate the number of subjects available for each gene variant combination. Response 2: A column has been added that shows the number of individuals for each genotype group, see the table below. Table 1. Description of the genotype groups as the bases for participant selection SNP (effect allele) rs2235321 (A) rs855791 (A) rs4820268 (A) Number of effect alleles Number of individuals available for each combination Major/Minor allele G/A G/A A/G AA/GG/AA A/A G/G A/A 4 570 AG/GG/GA A/G G/G G/A 2 687 GG/GG/AA G/G G/G A/A 2 199 GG/GG/GA G/G G/G G/A 1 386 GG/GG/GG G/G G/G G/G 0 161 AG/AG/AA A/G A/G A/A 4 132 AG/GG/AA A/G G/G A/A 3 632 GG/AG/AA G/G A/G A/A 3 104 GG/AG/GA G/G A/G G/A 2 98 Comment 3: Exclusion criteria: It would seem more appropriate to exclude participants with sickle cell anaemia or G6PD deficiency instead of using it as a co-variate in the analysis. However, if this reduces the numbers of each genotype combination too much, authors may as well include them while doing a retrospective sensitivity analysis. Response 3: We do not currently have the G6PD and sick cell carriage status of the participants. Both will be determined during this study. The section describing this now reads “G6PD and sick cell carriage status will be determined and individuals carrying these variants will be excluded from the analysis if this will not significantly reduce the sample size. Otherwise, a retrospective sensitivity analysis will be done to assess the impact of these variants.” (line 206) Comment 4: The primary outcome measure will be the change in serum iron concentration before and five hours after a single 400 mg dose of ferrous sulfate iron given orally. This is not the best way to measure iron absorption, which would ideally be assessed with a stable isotope method. Past studies have shown that change in serum iron concentration cannot be used as a measure of iron bioavailability at the individual level. Authors should provide references that back up the validity of their approach. Response 4: We acknowledge that the stable isotope method is heavily favored as a methodology for measuring iron absorption. However, we have chosen to use serum iron increase as a proxy as a or oral iron absorption based on the work of Nai et al (TMPRSS6 rs855791 modulates hepcidin transcription in vitro and serum hepcidin levels in normal individuals. Nai A, Pagani A, Silvestri L, Campostrini N, Corbella M, Girelli D, Traglia M, Toniolo D, Camaschella C. Blood. 2011 Oct 20;118(16):4459-62. PMID: 21873547) and our own previously published work (Cross J, Bradbury R, Fulford A, Jallow AT, Prentice AM, Cerami C. Oral iron acutely elevates bacterial growth in human serum. Scientific Reports, 2015 Nov 23;5:16670 PMID: 26593732 PMCID: PMC4655407). Additionally, we conducted a pilot study in which we 400 mg ferrous sulfate iron orally to determine the frequency and best time point to have the highest serum iron levels. From this pilot study, serum iron levels at 5 hours peaked at 5 hours. Minor comments: Under study procedures, second paragraph: ‘injestion’ is miss-spelled. Sample size calculation: A type 1 error of 0 seems to be ideal, yet unrealistic. Is this a typo? Response: ‘Injestion’ has been corrected to ‘ingestion’ “type 1 error of 0’ has been corrected to ‘type 1 error of 0.01’."
}
]
}
] | 1
|
https://f1000research.com/articles/8-701
|
https://f1000research.com/articles/9-581/v1
|
09 Jun 20
|
{
"type": "Software Tool Article",
"title": "Seqpare: a self-consistent metric of similarity between genomic interval sets",
"authors": [
"Selena C. Feng",
"Nathan C. Sheffield",
"Jianglin Feng",
"Selena C. Feng",
"Nathan C. Sheffield"
],
"abstract": "Searching genomic interval sets produced by sequencing methods has been widely and routinely performed; however, existing metrics for quantifying similarities among interval sets are inconsistent. Here we introduce Seqpare, a self-consistent and effective metric of similarity and tool for comparing sequences based on their interval sets. With this metric, the similarity of two interval sets is quantified by a single index, the ratio of their effective overlap over the union: an index of zero indicates unrelated interval sets, and an index of one means that the interval sets are identical. Analysis and tests confirm the effectiveness and self-consistency of the Seqpare metric.",
"keywords": [
"Genome analysis",
"interval set",
"similarity metric",
"sequence comparison",
"algorithm"
],
"content": "Introduction\n\nFunctional genomic data are often summarized as interval sets and deposited in public repositories (e.g., UCSC, ENCODE, Roadmap, GEO, SRA etc.). Identifying relationships among sequences and searching through widely available sequence data are routine tasks in genomic research. A fundamental operation in genomic/epigenomic analysis is comparing two interval sets, and many algorithms and tools have been developed for this purpose (Alekseyenko & Lee, 2007; Cormen et al. 2001; Feng et al., 2019; Giardine et al., 2005; Jalili et al., 2019; Kent et al., 2002; Li, 2011; Neph et al., 2012; Quinlan & Hall 2010; Richardson, 2006). These methods are based on computing the total number of intersections (overlaps) between the two interval sets. To compare a query interval set with multiple interval sets in a genomic sequence database, searching tools LOLA (Sheffield & Bock, 2016) and GIGGLE (Layer et al., 2018) calculate two values — Fisher’s exact p-value and the odds-ratio based on the total number of intersections — and use them as the similarity score to rank the search results. These similarity metrics have proven useful for determining relationships among interval sets, but also have some flaws. First, calculating the Fisher’s exact test results requires building a contingency table, but determining its values is not straightforward. The p-value and odds-ratio for two intervals sets (with number of intervals N1 and N2) are calculated from four numbers, namely, the number of intersections between the two sets, n, the number of intervals in set 1 that do not overlap an interval in set 2, N1 - n, the number of intervals in set 2 that do not overlap an interval in set 1, N2 - n, and the number of intervals that are not present in either set, m. Determining the value of the fourth number m is not straightforward; in LOLA, it depends on the definition of a “universe set” that is not objectively defined, whereas GIGGLE estimates m from the two interval sets. Second, the total number of overlaps n does not necessarily reflect similarity since intervals can have very different lengths (often in the range of 1 to 105 base pairs) and two very different intervals may intersect by only a few base pairs. This can result in inconsistency of the metrics: a comparison between two identical interval sets may have a larger p-value or smaller odds-ratio than a comparison between two different interval sets (see example cases and analysis in the next section). More strikingly, since one interval may contain or cover other intervals in an interval set, depending on how the overlaps are computed, n can be larger than N1 and/or N2, i.e., N1-n and/or N2-n can be negative, which leads to both the p-value and odds-ratio being undefined—another potential source of inconsistency. Third, the Fisher’s exact-based metrics require two values (p-value and odds-ratio) but neither is a direct measurement of the similarity: p-values are sensitive to the total number of regions and can range as low as 10-200 for large genomic interval sets, and odds-ratios are sensitive to small numbers; and neither metric directly informs on how similar the two sets are. Last, the p-value calculation is computationally expensive for genomic interval sets, particularly when the number of intervals is large (up to 109). To overcome these weaknesses of the Fisher’s exact-based metrics, we developed Seqpare, a self-consistent metric for quantifying the similarity among genomic interval sets.\n\n\nMethods\n\nThe Seqpare metric uses a single index to quantify the degree of similarity S of two interval sets with number of intervals N1 and N2. Similar to the Jaccard index, the Seqpare metric is directly defined as the ratio of the total effective overlap O of the two interval sets over the union N1+N2 - O:\n\n\n\nFor two intervals v1 in set 1 and v2 in set 2, the similarity s is defined as:\n\n\n\nwhere o is the length of the intersection and l1 and l2 are the lengths of v1 and v2 respectively. Definition 2 is the Jaccard index for individual intervals: o represents the effective overlap of the two intervals and s takes values in the range of [0, 1]: s = 0 indicates that there is no overlap between the two intervals, and s = 1 means that the union equals the overlap so v1 and v2 are identical. Then the total effective overlap O for the two interval sets can be calculated by adding up the similarities of all mutual best matching (MBM) pairs:\n\n\n\nA MBM pair is defined as a pair of intervals v1 and v2 that fulfill the following conditions: among all intervals in set 2 that intersect v1, v2 matches v1 the best, i.e., the similarity s between v1 and v2 is the highest among those intersections; and among all intervals in set 1 that intersect v2, v1 matches v2 the best. Clearly, if two intervals only intersect each other, then they form an MBM pair. In Figure 1a, the two long intervals (the 1st in set 1 and the 1st in set 2) only intersect each other (intersection pair ip1,1), so they form an MBM pair; similarly, the two short intervals ip2,2 form another MBM pair. For intervals that involve multiple intersections, we define a relatively simple and strict rule to find the MBM pairs: find and choose the first MBM pair that has the highest s among all involved intersection pairs, then find and choose the next MBM pair that has the highest s from the rest of the intersection pairs (excluding all pairs that involve the intervals that are already chosen), and so on until there are no more intersection pairs. In Figure 1b, there are 3 intersection pairs: ip1,2 with s=1/5, ip1,3 with s=1/9, and ip2,3 with s=1/5. So the first MBM pair is either ip1,2 or ip2,3 depending on which one is found first. If ip1,2 is chosen as the first MBM pair, then ip1,3 will not be considered since interval 1 in set 1 is already chosen, and then we have only ip2,3 left, which is the second MBM pair. We get the same result if ip2,3 is chosen as the first MBM pair. In Figure 1d, there are 6 ips and two MBM pairs: ip1,1 and ip3,2 both with s=1, where interval 2 in set 1 (i2,1) does not have a match. Note that interval i2,1 matches best with interval i1,2, but i1,2 does not match best with i2,1, so they are not the mutual best matching pair.\n\nThe length ratio of the short interval to the longer intervals are 1: 5 in a), 1: 3: 5 in b), and 2: 3 :4 in d). The total number of overlaps n is 2 for a), 3 for b) (interval 1 in Set 1 intersects two intervals in Set 2), 4 for c) and 6 for d). The p-value in case b is smaller than that in case a. N1 - n and N2 - n are both negative in cases c and d.\n\nSince the number of total matching pairs = Min(N1, N2) — the minimum of N1 and N2 — and s is in range of [0, 1], we obtain O = Min(N1, N2), and S takes a value in the range of [0, 1]. If S is zero, then there is no matching pair, and vice versa; if S = 1, then N1 = N2 = O (the two sets are equivalent), and vice versa. And, because each s is the mutual best match, O is symmetric (the amount of overlap between set 1 and set 2 is the same as that between set 2 to set 1) and so is S.\n\nIn Figure 1a, Oa=2 and Sa=1, which is correct because the two sets are identical. In Figure 1b, Ob=2/5 and Sb=1/14, which is expected since the two sets are very different. The Fisher’s exact approach is inconsistent here: the p-value in 1b is smaller than that in 1a although the two sets in 1b are very different while those in 1a are equivalent. Assuming that the number of intervals N in the ‘universe set’ is 100, then Fisher’s exact test contingency table is [(2, 0), (0, 98)] in 1a and [(3, 0), (0, 97)] in 1b, which gives pa = 2.02x10-4 and pb = 6.18x10-6 respectively. The odds-ratio is 8 in both cases. In cases c and d, N1- n and N2 - n are all negative, so it is not conceptually appropriate to use Fisher’s exact test to calculate the p-value and odds-ratio.\n\nThe implementation of the Seqpare metric is simple. The searching for MBM pairs is deterministic and it can be implemented by directly following the description in the above section. The Seqpare code is built on top of the AIList v0.0.1 (Feng et al., 2019) software written in C.\n\nThe Seqpare software (Feng & Feng, 2020) was tested on Linux machines and the minimum required memory is 8GB. The interval set file should be in the format of bed or bed.gz.\n\n\nResults\n\nTo test Seqpare and compare it with the Fisher’s exact-based metrics, we took 100 interval sets from a UCSC database and used one interval set, affyGnf1h, as a query to search over the database. Because the database contains the query set, affyGnf1h should have the highest similarity score. Table 1 (Feng & Feng, 2020) shows part of the result. Interval set affyGnf1h indeed ranks first with maximum similarity 1 when using Seqpare, but it ranks 94th out of 100 when using the p-value and ranks last when using the odds-ratio. This happens because N1-n and N2-n are both negative (n=16686, N1=N2=12158). Given this inconsistency, GIGGLE sets negative N1-n and N2-n to zero to calculate the p-value, and to one to calculate the odds-ratio. The Seqpare indices for other interval sets are all small (<0.03) because the average effective overlap of an intersection pair in those sets is about 0.1 or less, i.e., they are very different from the query set affyGnf1h; however, all of the p-values are so small (e-200), which suggests that the p-value is not a meaningful similarity index for these genomic interval sets. This search takes 6m30s for Seqpare and 15m32s for GIGGLE. All computations were carried out on a computer with a 2.8GHz CPU, 16GB memory, and an external SSD hard disk. The complete results can be found at the same site as the software.\n\n\nConclusion\n\nWe have shown that the Fisher’s exact test approach may be not the most appropriate test statistic for comparing similarity among interval sets. While the approach has been shown to be successful for many questions, we have demonstrated how it can break down for a variety of reasons, such as very similar interval sets, within-set containment, widely varying interval lengths among sets, or small effective overlaps. In contrast, Seqpare is a self-consistent metric for quantifying the similarity of two interval sets that addresses these concerns. Seqpare is the first rigorously defined metric for comparing two sequences based on their interval sets. In addition to the metric itself, our Seqpare software tool provides functions for both searching and mapping large-scale interval datasets. We anticipate that this approach will contribute to novel results for interval set searching.\n\n\nData availability\n\nTest data of interval sets are from http://hgdownload.cse.ucsc.edu/goldenPath/hg19/database\n\nZenodo: deepstanding/seqpare: First release of Seqpare. http://doi.org/10.5281/zenodo.3840051 (Feng & Feng, 2020)\n\nThis project contains the following underlying data:\n\n- AffyGnf1h_ucsc100_seqpare (Seqpare similarity result)\n\n- AffyGnf1h_ucsc100_giggle (GIGGLE p-value and odds-ratio result)\n\nData is available alongside the source code under the terms of the MIT license.\n\n\nSoftware availability\n\nSource code available from: https://github.com/deepstanding/seqpare\n\nArchived source code at time of publication: http://doi.org/10.5281/zenodo.3840051 (Feng & Feng, 2020)\n\nLicense: MIT",
"appendix": "References\n\nAlekseyenko AV, Lee CJ: Nested containment list (NCList): A new algorithm for accelerating interval query of genome alignment and interval databases. Bioinformatics. 2007; 23(11): 1386–93. PubMed Abstract | Publisher Full Text\n\nCormen TH, Leiserson CE, Rivest RL, et al.: Introduction to algorithms second edition. 2001. Reference Source\n\nFeng J, Ratan A, Sheffield NC: Augmented interval list: A novel data structure for efficient genomic interval search. Bioinfomatics. 2019; 35(23): 4907–4911. Publisher Full Text\n\nGiardine B, Riemer C, Hardison RC, et al.: Galaxy: A platform for interactive large-scale genome analysis. Genome Res. 2005; 15(10): 1451–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJalili V, Matteucci M, Goecks J, et al.: Next generation indexing for genomic intervals. IEEE T KNOWL DATA EN. 2019; 31(10): 2008–2021. Publisher Full Text\n\nKent WJ, Sugnet CW, Furey TS, et al.: The human genome browser at ucsc. Genome Res. 2002; 12(6): 996–1006. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLayer RM, Pedersen BS, DiSera T, et al.: GIGGLE: A Search Engine for Large-Scale Integrated Genome Analysis. Nat Methods. 2018; 15(2): 123–126. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi H: Tabix: Fast Retrieval of Sequence Features From Generic TAB-delimited Files. Bioinformatics. 2011; 27(5): 718–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNeph S, Kuehn MS, Reynolds AP, et al.: BEDOPS: high-performance genomic feature operations. Bioinformatics. 2012; 28(14): 1919–1920. PubMed Abstract | Publisher Full Text | Free Full Text\n\nQuinlan AR, Hall IM: BEDTools: A flexible suite of utilities for comparing genomic features. Bioinformatics. 2010; 26(6): 841–842. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRichardson JE: Fjoin: Simple and efficient computation of feature overlaps. J Comput Biol. 2006; 13(8): 1457–1464. PubMed Abstract | Publisher Full Text\n\nFeng S, Feng J: deepstanding/seqpare: First release of Seqpare (Version v1.0.0). Zenodo. 2020. http://www.doi.org/10.5281/zenodo.3840051\n\nSheffield NC, Bock C: LOLA: Enrichment analysis for genomic region sets and regulatory elements in R and bioconductor. Bioinformatics. 2016; 32(4): 587–589. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "66632",
"date": "30 Jul 2020",
"name": "Zhaohui Steve Qin",
"expertise": [
"Reviewer Expertise Bioinformatics."
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nFeng et al. described a novel method named seqpare, to measure the similarity between two sets of genomic intervals. I think the authors picked an important problem to work on and they presented a nice solution. The performance, from their studies, seem quite impressive. It is a nice new tool for solving such problems. but I do have a few issues with the paper.\nI never get what exactly does “self-consistent” really mean. I do not think this is the same as the consistent used in statistics which is an asymptotic property. I think the authors should provide a better explanation of this term and why this is important for such a problem.\n\nGenomic intervals, in my opinion, should come with genomic coordinates, to indicate where does it come from in terms of the reference genome. Of course, this will not be the case if the data does not come from a model organism. I am surprised that such information is not used in their method. I think it is a disadvantage if this is not used. Since the problem can be much easier if order is being considered. I am interested to know what do the authors think about this.\n\nFigure 1C is not mentioned in the paper. Not sure why.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly",
"responses": [
{
"c_id": "6185",
"date": "04 Jan 2021",
"name": "John Feng",
"role": "Author Response",
"response": "Response to Reviewer 1 Comment: Feng et al. described a novel method named seqpare, to measure the similarity between two sets of genomic intervals. I think the authors picked an important problem to work on and they presented a nice solution. The performance, from their studies, seem quite impressive. It is a nice new tool for solving such problems. but I do have a few issues with the paper. Response: Thank you for your helpful comments and suggestions. We have revised the paper according to the specific comments, and we have also added a subset of the test data and instructions to the software page to ease the verification of the result. Comment: I never get what exactly does “self-consistent” really mean. I do not think this is the same as the consistent used in statistics which is an asymptotic property. I think the authors should provide a better explanation of this term and why this is important for such a problem. Response: The word \"self-consistent\" is used here with its literal meaning of \"not having parts or aspects which are in conflict or contradiction with each other\" (google dictionary). Being self-consistent is necessary for a definition, concept, metric or theory. We think it's a useful word and we have explained it in the Introduction. Nevertheless, we did remove it from the title since it may cause confusion without context. Comment: Genomic intervals, in my opinion, should come with genomic coordinates, to indicate where does it come from in terms of the reference genome. Of course, this will not be the case if the data does not come from a model organism. I am surprised that such information is not used in their method. I think it is a disadvantage if this is not used. Since the problem can be much easier if order is being considered. I am interested to know what do the authors think about this. Response: Genomic intervals in the standard BED file format contain essential genomic information such as Chromosome, start coordinate, end coordinate and strand, etc. Like all other tools mentioned in the paper, Seqpare uses BED file format. Seqpare does not require the intervals to be sorted, which is an advantage over tools that do require intervals to be sorted. Comment: Figure 1C is not mentioned in the paper. Not sure why. Response: Figure 1 shows four cases and Figure 1c is mentioned as case c in the last paragraph of the Methods section \"Seqpare metric\". We replaced ‘case c’ with ‘Figure 1c’."
}
]
},
{
"id": "71030",
"date": "11 Nov 2020",
"name": "Burcak Otlu",
"expertise": [
"Reviewer Expertise Computational biology",
"cancer genomics",
"bioinformatics."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors have defined a similarity metric for defining similarity between two interval sets. They have justified well why this metric is better than Fisher's exact p-value and odds-ratio.\n\nI wonder what is the time complexity of MBM calculation? Does this metric work when there are self-contained intervals (overlapping intervals) in each interval set?\n\nWhy you choose this metric instead of simply using jaccard index?\nIn the text, \"6 ips\" is written. It would be better to explain \"ips\" before its usage.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "6186",
"date": "04 Jan 2021",
"name": "John Feng",
"role": "Author Response",
"response": "Comment: The authors have defined a similarity metric for defining similarity between two interval sets. They have justified well why this metric is better than Fisher's exact p-value and odds-ratio. Response: Thank you for this positive comment. Comment: I wonder what is the time complexity of MBM calculation? Does this metric work when there are self-contained intervals (overlapping intervals) in each interval set? Why you choose this metric instead of simply using jaccard index? Response: Thanks for the insightful comments. 1. Seqpare finds all intersections (total m) by using AIList algorithm with time complexity of O(n log n), where n is the number of intervals in the datasets. The MBM algorithm calculates the similarities of these intersection pairs and then finds the MBM pairs one by one. The time complexity of MBM depends on the cases. If all intersection pairs are simple (Figure 1a), then the complexity is O(m); in the worst case (Figure 1c or Figure 1d), all intersection pairs are correlated and the complexity is O(m2). For most real genomic interval sets, the time complexity is close to O(m). 2. Seqpare is a general metric of similarity and it works when there are self-contained intervals. Figure 1c and 1d are two cases of self-contained interval sets, where seqpare works well but the p-value based metric breaks down. Also, most of the UCSC interval sets used in this paper contain large number of overlapping intervals. 3. Jaccard index is defined as the ratio of Intersection over Union and it takes value in the range of [0, 1], i.e., the Intersection is never larger than the Union and the Union is never zero or negative. Jaccard index can be strictly used for two individual intervals, as explained in the Metric section of the paper. However, it cannot be used directly as similarity metric for two interval sets. For two interval sets, the simple Jaccard metric would be J = N/(N1+N2-N), where N1 and N2 are the number of intervals in the two interval sets, and N is the number of intersections between them. However, the ‘Intersection’ N can be larger than the ‘Union’ (N1+N2-N), and (N1+N2-N) can also be zero or negative: in Figure 1c, N=4, N1=2, N2=2, and (N1+N2-N) is zero; in Figure 1d, N=6, N1=2, N2=3, (N1+N2-N) is -1. Therefore, just like the p-value, simple Jaccard index is not a rigorous similarity metric for interval sets. This also suggests that using p-value or simple Jaccard index to compare interval sets can be misleading and defining a rigorous similarity metric for interval sets is challenging. Comment: In the text, \"6 ips\" is written. It would be better to explain \"ips\" before its usage. Response: We replaced it as ‘6 intersection pairs’."
}
]
}
] | 1
|
https://f1000research.com/articles/9-581
|
https://f1000research.com/articles/9-1500/v1
|
23 Dec 20
|
{
"type": "Research Article",
"title": "Corneal laser procedure for vision improvement in patients with late stage dry age-related macular degeneration - a retrospective observational cohort study",
"authors": [
"Raymond M. Stein",
"Samuel N. Markowitz",
"Michael J. Berry II",
"Michael J. Berry",
"Raymond M. Stein",
"Samuel N. Markowitz",
"Michael J. Berry II"
],
"abstract": "Purpose: To determine the safety and efficacy of corneal photovitrification (CPV), a new corneal laser procedure, for vision improvement in patients with late stage dry age-related macular degeneration (AMD). Design: Retrospective observational cohort study Participants: 32 eyes of 17 patients with late stage dry AMD; each eye received a single CPV corneal laser procedure and had 12 months (12m) post-treatment (Tx) follow-up. Methods: Pre- and post-Tx examinations included slit-lamp biomicroscopy, subjective manifest refraction, best corrected distance and near visual acuity (BCDVA and BCNVA), and potential visual acuity (PVA). Additional examinations including contrast sensitivity (CS), corneal topography (CT), ray tracing aberrometry (RTA) and microperimetry (MP) were obtained for a subgroup (n=12) of eyes. Main Outcome Measures: BCDVA, BCNVA, PVA, CS, CT, RTA and MP measurements Results: Safety – There were no complications or adverse events.\n\nEfficacy – Mean (± SD) BCDVA improved significantly (p < 0.004) from 20/238 (1.08 logMAR) at baseline to 20/144 (0.86 logMAR) at 12m post-Tx corresponding to 11.0 (± 13.1) letters gained. Mean contrast sensitivity improved significantly (p < 0.05) by a factor of 1.86 from baseline at 12m post-Tx. Conclusions: Subject to the limitation of a small sample size, this pilot study indicates that the CPV corneal laser procedure is safe and efficacious for vision improvement in patients with late stage dry AMD. The CPV Tx mechanism of action involves retinal irradiance distribution modifications that may stimulate patient use of functional, rather than atrophied, retinal regions.",
"keywords": [
"cornea",
"laser",
"retina",
"vision improvement",
"age-related macular degeneration"
],
"content": "Introduction\n\nAge-related macular degeneration (AMD) is a leading cause of vision impairment globally1. Several FDA-approved pharmacologic treatments are available for the neovascular form of AMD (also termed “wet” AMD); in particular, anti-vascular endothelial growth factor (anti-VEGF) injections are broadly used to manage the disease by reducing its progression and, in many cases, by providing some vision improvement2. However, management of dry AMD and especially geographic atrophy (GA), late stage AMD, has proven to be more challenging. There are no FDA-approved pharmacologic treatments3 and only limited success in providing vision improvement has been obtained by use of intraocular implants4, of which only the implantable miniature telescope (IMT)5 is FDA-approved. Dry AMD procedures such as subthreshold laser therapy6 and photobiomodulation7 are directed primarily toward improving retinal function for early to intermediate stage AMD patients. Low vision aids such as prism spectacles8 and electronic glasses9 are also available but have not been broadly used by late stage dry AMD patients. The purpose of the present study is to describe a new corneal laser procedure that offers significant vision improvement in patients with late stage dry AMD.\n\n\nMethods\n\nThis retrospective observational cohort study (registered with ClinicalTrials.gov NCT04349254 on 16 April 2020) was completed in conformance with ethical principles of the World Medical Association Declaration of Helsinki. The study protocol (Pro00044329) was approved on 4 August 2020 by an institutional review board (Advarra, Aurora, Ontario, Canada). Written Informed consent, with a provision for release of medical records, was obtained from each patient prior to treatment. Patients were primarily referred by optometrists and treated in the clinic during the period February through November 2018. Clinic records of 32 eyes of 17 patients [8F, 9M; mean (± SD) age: 81.7 (± 8.8) y; range: 58–96 y; all of Caucasian ancestry] with late stage dry age-related macular degeneration who had received one treatment in each eye using the same device and protocol and with 12-month (12m) follow-up examinations were identified and patient records were examined during the period August through October 2020. Eyes were either pseudophakic or phakic with no visually significant cataract. All patient eyes had vision impairment, with mean ± SD best spectacle-corrected distance and near visual acuities (BCDVA and BCNVA) of 20/238 (1.08 ± 0.37 logMAR; 31 letters) and 20/199 (1.00 ± 0.29 logMAR; 35 letters), respectively. The study size was limited by the availability of records but the size was confirmed to provide statistical significance of outcome measures.\n\nInclusion criteria included age of ≥ 55 y and, in the eye to be treated, diagnosed late stage dry AMD, moderate to profound BCDVA impairment (in the range of 20/44 to 20/1000), normal corneal topography (i.e., without distorted or unclear mires) and examinations extending to at least 12 months post-Tx. Exclusion criteria included previous corneal surgery and visually significant ocular disease other than AMD.\n\nExaminations included slit-lamp biomicroscopy; optical coherence tomography; subjective manifest refraction (SMR); BCDVA and BCNVA; potential visual acuity (PVA)10 using Gonzalez-Markowitz charts (Precision Vision, Woodstock, IL) at 50 cm examination distance. For a subset of eyes (n=12), additional examinations included: 1) contrast sensitivity (CS) using Pelli-Robson charts (Precision Vision, Woodstock, IL), 2) corneal topography (CT) and ray tracing aberrometry (RTA) using an iTrace analyzer (Tracey Technologies, Houston, TX) and 3) retinal sensitivity, fixation stability and preferred retinal locus using a Macular Integrity Assessment (MAIA) microperimeter (MP; Centervue, Fremont, CA). SMR, BCDVA and BCNVA examinations were completed pre-Tx for 32 eyes and at 1 month (1m), 3m, 6m, and 12m post-Tx for 29, 28, 28 and 32 eyes, respectively; SMR measurements were recorded for all eyes pre-Tx but only for 12 eyes post-Tx. PVA examinations were completed pre-Tx for 20 eyes. Other examinations (CS, CT, RTA and MP) were completed pre-Tx and at post-Tx times extending to 12m post-Tx for 12 eyes (CS, CT and RTA) and 8 eyes (MP).\n\nTreatments were completed using a Clear-K® Low Vision Aid System (Optimal Acuity Corporation, Austin, TX) to deliver pulsed laser energy simultaneously to the cornea in 4 spots of 0.5 mm diameter arranged symmetrically 90° apart and located on a 6.0 mm diameter ring centered on the pupillary centroid as shown in Figure 1. Laser parameters included 2 µm wavelength, 150 ms pulse duration and 48 to 50 mJ energy per spot. Laser light was transmitted from the console through an optical fiber array terminated by a handpiece that docks onto a sapphire applanation window/suction ring (SAWSR) assembly mounted on the eye. Laser energy was delivered through the SAWSR onto the eye in order to provide a fixed location of treatment spots with epithelial protection (by the sapphire window acting as a heat sink) from thermal damage. Treatments produced small corneal changes in shape that acted to redirect light onto functional regions of the retina. Patients were reclined to a supine position, given a drop of topical anaesthetic in the eye to be treated, and then treated.\n\nActual Tx spots are barely visible under room lighting conditions.\n\nStatistical significances of paired outcomes were assessed by Wilcoxon signed rank tests. Intereye correlations in bilateral treatments and correlations between potential visual acuity test measurements and visual acuity changes were assessed by Pearson correlation coefficients; statistical significances of the correlations were evaluated by bootstrap resampling. OD and OS logMAR values for correlated bilateral treatments were averaged for each outcome (BCDVA and BCNVA at baseline and at each follow-up time) in order to calculate statistical significances of post- vs. pre-Tx differences11. Many outcomes are reported as mean (± standard deviation) values. Microsoft Excel software (2010) functions were used for statistical analysis.\n\n\nResults\n\nNo complications or significant adverse events occurred.\n\nThe mean BCDVA and BCNVA of the treated eyes increased from baseline at each follow-up (f/u) time, as shown in Figure 212 in terms of mean letters of vision gained on standard eye charts and in Figure 312 in terms of a histogram of percentages of changes in lines of vision at 12m post-Tx. Table 112 summarizes descriptive statistics of outcomes for bilateral treatments [calculated with intereye correlation, since Pearson correlation coefficients are positive and large (mean: 0.6) for all outcomes], for unilateral treatments, and for all (unilateral plus uncorrelated bilateral) treatments. For all treatments (and for bilateral Txs only), all outcomes are statistically significant at the p<0.05 level. The largest mean (± SD) gain of 14.6 (± 11.1) letters in BCDVA was achieved at 1m post-Tx; the mean BCDVA gain was stable at ca. 11 to 12 letters from 3m to 12m post-Tx. 58.6% (17 of 29) of treated eyes gained 15 or more letters (3 or more lines) of BCDVA at 1m post-Tx compared to baseline; this success percentage decreased to 43.8% (14 of 32 eyes) at 12m post-Tx, possibly due to regression of treatment effect and/or progressive dry AMD vision loss. Figure 212 also shows the expected mean BCDVA loss for untreated late stage AMD eyes for a similar cohort from another study13, amounting to 4.1 letters lost at 12m. So, the 11.9 letter mean gain in BCDVA at 12m post-Tx for CPV treated eyes is actually 15 letters mean better vision than expected for untreated eyes.\n\nBottom timeline – Mean BCDVA letters lost for a similar cohort of untreated eyes (from Ref. 13).\n\nThe histogram shows the percentage of eyes that lost 1, 2 or 3 or more lines of vision, were unchanged, or gained 1, 2, or 3 or more lines of best corrected distance visual acuity (BCDVA) and best corrected near visual acuity (BCNVA).\n\nEntries contain the mean and standard deviation of letters gained, sample size (n) and p-value (where appropriate).\n\nPre-Tx potential visual acuity (PVA) measurements10 demonstrated variable improvements compared to pre-Tx BCDVA measurements: 6 eyes had less than 10 letters (2 lines) improvement, 6 eyes had between 10 to 14 letters (2 to 2.8 lines) improvement and 8 eyes had 15 or more letters (3 or more lines) improvement. 1m post-Tx mean BCDVA improvements correlated moderately well (Pearson correlation coefficient = 0.44; p < 0.03) with pre-Tx mean PVA improvements. The Pearson correlation coefficient = 0.32 (p < 0.12) for 3m post-Tx mean BCDVA improvements with pre-Tx mean PVA improvements was also moderately good. Pearson correlation coefficients for 6m and 12m post-Tx mean BCDVAs compared to pre-Tx mean PVAs were near-zero and were not statistically significant. The decrease in correlation may be caused by partial loss of treatment effect and, in part, by the progressive loss of BCDVA in dry AMD eyes. In the “best” mean PVA improvement group (with 15 or more letters improvement relative to baseline BCDVA), the mean (± SD) post-Tx BCDVA improvements were 21.0 (± 12.1) and 15.2 (± 13.6) letters at 1m and 3m, respectively, compared to lesser BCDVA improvements of 8.8 (± 12.0) and 5.4 (± 8.7) letters at 1m and 3m, respectively, for the PVA improvement group of 14 or less letters improvement relative to baseline BCDVA.\n\nThe mean binocular BCDVA and BCNVA values also increased significantly compared to baseline values. No symptoms of aniseikonia, polyplopia or dysphotopsia were found in all or most patients although one patient had aniseikonia that was probably due to a large difference between spectacle lenses. Amsler grid tests typically demonstrated line straightening and reduction and/or relocation of dark and missing areas.\n\nSlit-lamp biomicroscope examination of treated corneas showed that treated spots were indented and lightly opacified; corneal epithelia were intact. Goldmann applanation tonometry measurements showed that mean intraocular pressures were unchanged at each post-Tx time.\n\nMean monocular contrast sensitivity (CS; at ca. 1 cycle/degree) of the treated eyes, as measured under photopic conditions without glare, increased as a function of follow-up time [significantly (p < 0.05) at all f/u times] – e.g., from log CS (mean ± SD) = 0.73 ± 0.40 pre-Tx to 1.00 ± 0.33 at 12m post-Tx, representing a mean CS increase by a factor of 1.86 from baseline. Mean binocular CS also increased as a function of f/u time by amounts similar to monocular increases.\n\nPre-Tx, the mean subjective manifest refraction (SMR) of the subset (n=8) of treated eyes with both SMR and aberrometry measurements was -0.41 D sphere - 0.81 D cylinder X 113°, 2.50 D add; -0.81 (± 0.46) D spherical equivalent (SE). At 1m post-Tx, the mean SMR was -0.31 D sphere - 0.53 D cylinder X 90°, 2.50 D add; -0.58 (± 0.43) D SE corresponding to a mean (± SD) hyperopic SE shift of +0.23 (± 0.33) D from baseline. The changes in SMR values from pre-Tx to 1m post-Tx were small, as were SMR changes from baseline to 3m and longer post-Tx. At 12m post-Tx, the mean (± SD) SMR SE change from baseline was -0.98 (± 0.76) D. None of the SE changes were statistically significant.\n\nRay tracing aberrometry (RTA) measurements provided information on retinal irradiance distribution modifications (IDMs) and objective refraction (OR) changes. Pre-Tx, rays of light incident on the cornea produced a “tight” pattern of retinal irradiation. Post-Tx, rays of light incident on the 3 mm optical zone (OZ) of the cornea were redistributed outward from the pre-Tx pattern center on the retina by a maximum mean (± SD) IDM value of 37 ± 20 µm. Pre-Tx, the mean OR of the treated eyes with aberrometry measurements was 0.16 D sphere - 1.65 D cylinder X 85°; -0.67 (±0.93) D SE. At 1m post-Tx, the mean OR was -0.22 D sphere - 1.92 D cylinder X 101°; -1.18 (± 0.77) D SE corresponding to a mean (± SD) myopic SE shift of -0.51 (± 0.52) D from baseline by aberrometry (in contrast to the SMR mean hyperopic SE shift of 0.23 D). At 12m post-Tx, the mean (± SD) aberrometry myopic SE shift from baseline was -0.12 (± 0.86) D, a smaller shift than the SMR myopic SE shift. Mean (± SD) total aberrations increased from 0.42 (± 0.21) µm at baseline to 0.53 (±0.33) µm and 0.44 (± 0.27) µm at 1m and 12m post-Tx, respectively. Most of the aberrometric changes were due to increased lower order (defocus and astigmatism) changes; both defocus and astigmatism increased. None of the total aberration changes were statistically significant.\n\nCorneal topography (CT) measurements provided information on corneal refractive changes, averaged within the 3 mm optical zone to calculate an Effective Refractive Power (Eff RP) and Astigmatism (Astig). Pre-Tx, mean (± SD) Eff RP and Astig values were 44.22 (± 1.09) D and 1.51 (± 1.21) D, respectively. At 1m and 12m post-Tx, mean (± SD) Eff RP changes were 0.39 (± 0.32) D and 0.22 (± 0.45) D, respectively, while mean (± SD) Astig changes were 0.04 (± 0.67) D and 0.06 (± 0.45) D. None of the Eff RP or Astig changes are statistically significant.\n\nDetailed (non-averaged) CT changes were obtained from CT difference maps. Figure 412 shows sample CT difference (1d post-Tx minus pre-Tx) maps in terms of Z Elevation and Refraction changes. In the left panel of Figure 412, depressions are evident in the Z Elevation map with maximum depressions centered on treatment spots that are located on the 6 mm optical zone (OZ). Maximum depressions vary from spot to spot in the range of -10 to -22 µm, with a mean (± SD) of -15.5 (± 5.2) µm. In the right panel of Figure 412, refractive increases are evident in the Refractive map with maximum increases on ca. the 4 mm OZ. Maximum refraction increases vary from spot to spot in the range of 1.0 to 2.1 D, with a mean (± SD) of 1.44 (± 0.51) D. At 1m post-Tx, Z Elevation and maximum Refraction changes decreased to mean (± SD) values of -10.6 (± 3.7) µm and 0.90 (± 0.38) D, respectively. At 12m post-Tx, the Z Elevation and maximum Refraction changes decreased further to mean (± SD) values of -7.4 (± 4.1) µm and 0.71 (± 0.46) D, respectively. For all eyes with CT measurements, the mean Z Elevation changes from baseline were approximately -18 µm at 1d post-Tx, decreasing to -12 µm and -8 µm at 1m and 12m post-Tx, respectively. The corresponding mean Refraction changes from baseline were approximately 2.1D, 1.1 D and 0.7 D at 1d, 1m and 12m post-Tx, respectively.\n\nGrids display 1 mm increments. Left map: Tx spots are located at 6.0 mm optical zone (OZ). Right map: maximum refractive changes are located at ca. 4 mm OZ. Tx spots were located as shown in Figure 1.\n\nMicroperimetry exams showed that fixation stability (FS) improved post-Tx, as determined by 95% bivariate contour ellipse area (BCEA) measurements; the 95% BCEA is the area containing 95% of the fixation points during a microperimeter scan of ca. 6 minutes duration. The mean (± SD) 95% BCEA decreased from 68.6°² (± 57.9°²) pre-Tx to 42.7°² (± 30.3°²) at 1m post-Tx and 31.5°² (± 25.1°²) at 4m post-Tx, with near statistical significance (p = 0.09 and 0.05 at 1m and 4m, respectively). The Pearson correlation coefficient between 95% BCEA and BCDVA (in logMAR units) was large (0.63) for all 30 FS data points. Four of the eight eyes with MP exams had large shifts (5.3° ± 1.9°) in preferred retinal loci (PRLs) that typically reduced the overlap of fixation points with atrophied regions of the retina. (Two of the eyes did not have usable 1m exams but did have usable 3m exams that showed large PRL shifts.) Figure 512 shows an example of the PRL change in one eye due to CPV Tx. Two of the eight eyes had almost no post-Tx shifts in PRLs but these were eyes with good pre-Tx FS and almost no post-Tx FS change.\n\nPre-Tx: the distribution of fixation points (blue dots) was centered at bottom left with most of the points within the atrophic retina (in white). 3m Post-Tx: the distribution of fixation points was centered superiorly with respect to the atrophic retina. Retinal sensitivity measurements (shown in color coding at 37 stimulus points over 10° diameter) – the black (zero sensitivity) stimulus points overlap the atrophic retina. The distribution of fixation points moved from overlap with the atrophic area (pre-Tx) superiorly onto a more functional retina (3m post-Tx). The distribution of fixation points also decreased from a 95% Bivariate Contour Ellipse Area (BCEA containing 95% of the fixation points) of 171.6°² pre-Tx to 46.2°² at 3m post-Tx.\n\n\nDiscussion\n\nLaser irradiation of the cornea by CPV Tx is a new corneal laser procedure that differs from invasive corneal procedures such as conductive keratoplasty (CK)14, laser-assisted in situ keratomileusis (LASIK)15, laser thermal keratoplasty (LTK)16 and small incision lenticule extraction (SMILE)17 in many ways including that, for CPV Tx, no corneal tissue is punctured, cut or removed and the epithelium is protected from thermal damage. The laser is “eye safe” meaning that the laser light is completely absorbed in the cornea; none of the laser light propagates through the cornea to irradiate the lens or the retina. The regions of corneal refraction change shown in Figure 4 act as aspheric multifocal lenses to redistribute light entering the cornea in order to produce a retinal irradiance distribution modification (IDM) for vision improvement. Laser Raman spectromicroscopy measurements18 indicated that tissue in CPV treated spots has reduced water content, probably leading to tissue compaction. Atomic force microscopy measurements18 indicated that tissue in CPV treated spots has increased modulus, leading to a change of treated anterior corneal stroma from a gel-like state with viscoelastic character to a more glass-like state with elastic character. Tx spots are lightly opacified and are confined to the anterior stroma upon slit-lamp biomicroscopy examination; Tx spots are not cosmetically significant under normal room lighting. Since Z Elevation and Refraction changes persist, although decreasing as a function of follow-up time, it is likely that CPV effects of reduction of water content and increase of modulus also persist. None of the eyes in this pilot study have received more than one CPV Tx so the possible benefit of multiple CPV Txs on eyes (treated simultaneously and/or sequentially) to maintain or even increase vision improvement remains to be investigated.\n\nThe corneal shape and refraction changes described above produce prompt redistribution of visible light similar to “prismatic effects” produced by prism spectacles8. Unlike prism spectacles, however, the four treatment spots create a quatrefoil pattern of corneal shape and refractive change (Figure 4) for 360 degrees that resembles four lenses, which each redistributes light rays. Pre-treatment, a location in visual space that maps onto an area of dysfunctional retina in an AMD patient is perceived as part of a blind spot. However, after CPV treatment, light rays from this location are mapped onto four different regions of the retina. If any of these regions are functional, then visual information about that location is transmitted to the central brain. The success of a “one size fits all” procedure for a patient population with different geometries of dysfunctional regions on their retinas results from the relatively high probability that at least one of the four lenses will enable the retina to transmit information about any location in visual space with compromised vision.\n\nRay tracing measurements indicate that the displacement of images is ca. 40 µm. This is not a large enough shift to move images from the fovea all the way to functional areas of the retina in most of our patients. However, this shift can allow images near the boundary of dysfunctional areas to move onto functional areas, and thereby restore some visual function. In addition, CPV treatment produces refractive changes over a large portion of the central cornea (cf. Figure 4). These large refractive changes distribute blurry versions of images in blind spots over a larger region of the retina, thus reaching more functional areas of the retina. In addition, ray tracing measurements were made for on-axis light beams; off-axis light beams are likely to experience a larger shift in their location on the retina.\n\nIt is notable that the eye movement strategy of patients changed. Six of the eight patients with microperimeter (MP) exams experienced large shifts in their preferred retinal locus (PRL) of fixations. Typically, the new PRL was displaced away from atrophied areas of the retina, and thus allowed the patient to gain more central visual information with each saccade. (An example of this pattern is shown for one eye in Figure 5). It is unclear how the changes in retinal irradiance distribution produced by CPV Tx helped eyes to find a better PRL. One possibility is that before treatment, the motor learning system of patients was in a “false optimum”, in which small deviations from the fixation location resulted in decreased visual function. Then, following treatment, visual information was spread over a larger region of the retina, helping the motor learning system to find a “true optimum”. Further study is needed to determine patterns (and causes) of PRL changes following CPV treatment.\n\nAnother significant change in eye movements following treatment was an improvement in fixation stability, which may make it easier for the brain to perform spatial integration of visual information. These changes may also be related to a shift in the PRL, as patients making saccades into a damaged area of the retina may partially compensate by making larger fixational eye movements to attempt to scan images over functional regions of the retina. Therefore, the change in eye movements can be thought of as helping to restore more normal visual function. [It should be noted that the fixation stability observed in patient eyes with AMD in this study is much worse than in normal eyes (mean 95% BCEA = ca. 3°² for patients in their 80s19), so the treated eyes are still in need of further functional improvement.] Further study is needed to determine if and how CPV treatment improves fixation stability.\n\nA potential concern is that the quatrefoil pattern of refractive changes produced by CPV Tx could result in multiple or fragmented visual perception. However, patients experienced unified vision. The visual information transmitted via multiple lenses combines, as each lens transmits the same information about a single location in visual space onto multiple regions of the retina. Functional combination of this information is akin to spatial integration, which the brain routinely carries out and which can result in acuity far greater than the spacing between cone photoreceptors20,21. This combination can also be thought of as a form of vision multiplexing22, which is an example of the wider phenomenon of sensory cue combination23–25. While the mechanism of optimal sensory cue combination remains an area of active research, the phenomenon is well-established.\n\nSince potential visual acuity (PVA) measurements10 correlate well with near-term (1m post-Tx) BCDVA improvements, PVA screening may be very useful as predicted previously26. Patient expectations may also be guided by PVA measurements.\n\nThere were no safety problems associated with the minimalist corneal laser procedure used in this study. CPV Tx produces minimal corneal changes without risks of intraocular surgery.\n\nThe only FDA-approved device for vision improvement in dry AMD patients is the implantable miniature telescope (IMT)5. At 12m post-implantation, IMT patients achieved mean improvements of 3.47 and 3.18 ETDRS lines for BCDVA and BCNVA, respectively, from baseline5, mostly because of telescopic magnification inherent in the procedure. Since all the IMT patients had cataracts, the IMT study authors attributed ca. 1 line of BCDVA improvement to cataract removal27, so the net BCDVA vision improvement in IMT patients was ca. 2.5 lines at 12m post-implantation. Patients with CPV Txs achieved mean improvements of 2.2 and 2.7 lines of BCDVA and BCNVA, respectively, at 12m post-Tx without the risks and safety problems associated with the very invasive IMT surgery and without added magnification. It should also be noted that the IMT produces “tunnel vision” in the implanted eye, while CPV treatment does not.\n\nFor CPV Tx, the “best” mean PVA improvement group (with 15 or more letters PVA improvement relative to baseline BCDVA) achieved a mean post-Tx BCDVA improvement of 21.0 letters at 1m post-Tx. The large standard deviations associated with each group and follow-up time (cf. Table 1) may be due, at least in part, to the range of accuracy with which light rays are redirected onto functional retinal areas. In this pilot study, each eye received the same CPV Tx. The possible benefit of custom CPV Txs (for example, by changing the pattern and energy density of Tx spots) that increase the accuracy of light ray redistribution onto the most functional retinal areas in each eye remains to be investigated.\n\nCPV treated patients typically experienced rapid and comfortable Txs with no post-Tx requirements for new medications or visual rehabilitation training (as is the case, for example, for IMT patients5). The present CPV study involved both unilateral and bilateral Txs, depending on whether one or both eyes needed vision improvement. A CPV bilateral Tx regimen contrasts with the IMT unilateral Tx regimen that is required because the IMT device produces “tunnel vision” and patients need an untreated fellow eye for peripheral vision and ambulation5.\n\nAnother CPV study28 demonstrated that bilateral Txs of both wet and dry AMD eyes produced similar vision improvements as in the present study. There is considerable merit to using combination therapy for wet AMD eyes in which anti-VEGF injections reduce the progression of the disease and CPV Tx provides significant vision and vision-related quality of life improvements.\n\nLimitations of the present pilot study are:\n\n1 – small sample size,\n\n2 – retrospective analysis of outcomes and\n\n3 – follow-up of only 12 months post-Tx.\n\nA prospective clinical study on a larger patient cohort with inclusion of additional measurements over a period extending to 24 months post-Tx is planned.\n\nThe CPV procedure for vision improvement is a new modality that may be broadly useful to benefit patients affected by late stage AMD.\n\n\nData availability\n\nVision improvement outcomes plus additional measurements\n\nDRYAD: Data from: Corneal laser procedure for vision improvement in patients with late stage dry age-related macular degeneration. https://doi.org/10.5061/dryad.sn02v6x2x12\n\nThis project contains the following underlying data:\n\n- F1000Research_Dataset_1_-_Corneal_laser_procedure_-_with_age_ranges.xls (Vision improvement outcomes (BCDVA, BCNVA, PVA and CS) and additional measurements (SMR, CT, RTA and MP))\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).",
"appendix": "Acknowledgements\n\nThe authors gratefully acknowledge Robert Devenyi, MD, FRCSC, Ghani Salim, MD and Chiara Cinco for fine discussions, documentation help and supplementary information.\n\n\nReferences\n\nWong WL, Su X, Li X, et al.: Global prevalence of age-related macular degeneration and disease burden projection for 2020 and 2040: a systematic review and meta-analysis. Lancet Glob Health. 2014; 2(2): e106–16. PubMed Abstract | Publisher Full Text\n\nHolekamp NM: Review of neovascular age-related macular degeneration treatment options. Am J Manag Care. 2019; 25(10 Suppl): S172–S181. PubMed Abstract\n\nBandello F, Sacconi R, Querques L, et al.: Recent advances in the management of dry age-related macular degeneration: A review [version 1; peer review: 2 approved]. F1000Res. 2017; 6: 245. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDunbar HMP, Dhawahir-Scala FE: A discussion of commercially available intra-ocular telescopic implants for patients with age-related macular degeneration. Ophthalmol Ther. 2018; 7(1): 33–48. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHudson HL, Lane SS, Heier JS, et al.: Implantable miniature telescope for the treatment of visual acuity loss resulting from end-stage age-related macular degeneration: 1-year results. Ophthalmology. 2006; 113(11): 1987–2001. PubMed Abstract | Publisher Full Text\n\nEng VA, Wood EH, Boddu S, et al.: Preventing progression in nonexudative age- related macular degeneration with subthreshold laser therapy: a systematic review. Ophthalmic Surg Lasers Imaging Retina. 2019; 50(3): e61–e70. PubMed Abstract | Publisher Full Text\n\nMarkowitz SN, Devenyi RG, Munk MR, et al.: A double-masked, randomized, sham-controlled single center study with photobiomodulation for the treatment of dry age-related macular degeneration. Retina. 2020; 40(8): 1471–1482. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarkowitz SN, Reyes SV, Sheng L: The use of prisms for vision rehabilitation after macular function loss: an evidence-based review. Acta Ophthalmol. 2013; 91(3): 207–11. PubMed Abstract | Publisher Full Text\n\nWittich W, Lorenzini MC, Markowitz SN, et al.: The effect of a head-mounted low vision device on visual function. Optom Vis Sci. 2018; 95(9): 774–784. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGonzalez EG, Tarita-Nistor L, Markowitz SN, et al.: Computer-based test to measure optimal visual acuity in age-related macular degeneration. Invest Ophthalmol Vis Sci. 2007; 48(10): 4838–45. PubMed Abstract | Publisher Full Text\n\nMurdoch IE, Morris SS, Cousens SN: People and eyes: statistical approaches in ophthalmology. Br J Ophthalmol. 1998; 82(8): 971–3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBerry M: Data from: Corneal laser procedure for vision improvement in patients with late stage dry age-related macular degeneration, Dryad, Dataset. 2020. http://www.doi.org/10.5061/dryad.sn02v6x2x\n\nWong WT, Kam W, Cunningham D, et al.: Treatment of geographic atrophy by the topical administration of OT-551: results of a phase II clinical trial. Invest Ophthalmol Vis Sci. 2010; 51(12): 6131–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcDonald MB: Conductive keratoplasty: a radiofrequency-based technique for the correction of hyperopia. Trans Am Ophthalmol Soc. 2005; 103: 512–36. PubMed Abstract | Free Full Text\n\nAzar DT, Koch DD: LASIK: fundamentals, surgical techniques and complications. Marcel Dekker, Inc., New York. 2003. Reference Source\n\nKoch DD, Kohnen T, McDonnell PJ, et al.: Hyperopia correction by noncontact holmium:YAG laser thermal keratoplasty: U.S. phase IIa clinical study with 2-year follow-up. Ophthalmology. 1997; 104(11): 1938–47. PubMed Abstract | Publisher Full Text\n\nSekundo W: Small incision lenticule extraction (SMILE): principles, techniques, complication management and future concepts. Springer Verlag, New York. 2015. Publisher Full Text\n\nSerdarevic O, Heller D, Berry M: Corneal photovitrification–basic science experiments. Lasers Surg Med. 2017; 49: 466–467.\n\nMorales MU, Saker S, Wilde C, et al.: Reference clinical database for fixation stability metrics in normal subjects measured with the MAIA microperimeter. Transl Vis Sci Technol. 2016; 5(6): 6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWestheimer G, Hauske G: Temporal and spatial interference with vernier acuity. Vision Res. 1975; 15: 1137–41. PubMed Abstract | Publisher Full Text\n\nLee BB, Wehrhahn D, Westheimer, et al.: The spatial precision of macaque ganglion cell responses in relation to vernier acuity of human observers. Vision Res. 1995; 35(19): 2743–58. PubMed Abstract | Publisher Full Text\n\nPeli E: Vision multiplexing: an engineering approach to vision rehabilitation device development. Optom Vis Sci. 2001; 78(5): 304–15. PubMed Abstract | Publisher Full Text\n\nHillis JM, Ernst MO, Banks MS, et al.: Combining sensory information: mandatory fusion within, but not between, senses. Science. 2002; 298(5598): 1627–30. PubMed Abstract | Publisher Full Text\n\nKording KP, Wolpert DM: Bayesian integration in sensorimotor learning. Nature. 2004; 427(6971): 244–7. PubMed Abstract | Publisher Full Text\n\nDoya K, Ishii S, Pouget A, et al.: Bayesian Brain: Probabilistic Approaches to Neural Coding. Cambridge, MA: MIT Press. 2007. Reference Source\n\nHarris MJ, Robbins D, Dieter JM, et al.: Eccentric visual acuity in patients with macular disease. Ophthalmology. 1985; 92(11): 1550–3. PubMed Abstract | Publisher Full Text\n\nHudson HL, Stulting RD, Heier JS, et al.: Implantable telescope for end-stage age-related macular degeneration: long-term visual acuity and safety outcomes. Am J Ophthalmol. 2008; 146(5): 664–673. PubMed Abstract | Publisher Full Text\n\nSerdarevic O, Tasindi E, Dekaris I, et al.: Vision improvement in dry and wet age-related macular degeneration (AMD) patients after treatment with a new corneal CPV procedure for light redirections onto the retina. Acta Ophthalmol. 2017; 95(S259). Publisher Full Text"
}
|
[
{
"id": "87602",
"date": "01 Jul 2021",
"name": "Uwe Oberheide",
"expertise": [
"Reviewer Expertise laser surgery of the eye"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nTo start with the positive: The idea, to multiply and redirect the incoming light by corneal laser treatment is very appealing. Unfortunately, no evidence has been produced here. Comments to be made: The term “vitrification” as used here is not known as a distinct result from any known laser-tissue-interaction. The given wavelength, pulse duration and energy suggest that there is a thermal effect, maybe subthreshold? Since there is no histological proof of how the corneal tissue has interacted with the treatment it remains speculative. Maybe the authors have performed a pre-clinical in-vitro investigation? If so, it would be very helpful to fully understand the character of laser-tissue interaction. If not, it should be mentioned and it should be explained in which way the laser-tissue interaction is different from laser thermokeratoplasty, a very well known treatment that was used for hyperopic and presbyopic treatments. The principle of operation is not comprehensible for the reader. The localization of the laser spots do not explain or justify postulated multiplaction and redirection of light. With symmetrical application, one would not expect an upward deflection (as shown with the fixation spots), but a symmetrical redistribution/distortion. However, the wavefront measurements made pre/post that could show this are not presented. Were the points applied as a function of the steep/flat axis positions? Always at the points given as examples? Then one would also have to compare the effect in relation to the previous refractive forces. The discussion with generation of four lenses, which overlap individually, is theory only. No proof has been made. The light still passes the central corneal area, while the treatment is placed peripherally at four points (4-6mm zone) probably leading to a refractive power increase. This increase of central corneal refractive power could as well and primarily have contributed to the higher near visual acuity in the eyes. The number of cases is in deed very small, each eye accounts for 3 percent. In addition, the results are mixed with different subgroups and the statistics are not very meaningful - standard deviations are sometimes larger than the actual measured value. Here, it would be better to indicate the range. The comparison of 15 gained lines in comparison to untreated eyes is meaningless here - 11 gained (standard deviation is 13!!). The other eyes would have lost 4 lines, because the disease pattern certainly does not allow statistics here and one would have to take the second eye (with bilateral disease) as reference.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": [
{
"c_id": "6971",
"date": "30 Jul 2021",
"name": "Michael Berry",
"role": "Author Response",
"response": "We thank the reviewers for taking the time to read and comment on our manuscript. However, we take issue with many of their comments and are concerned that some comments do not conform to widely accepted standards for clinical trials in the ophthalmology community. Detailed responses follow: “To start with the positive: The idea, to multiply and redirect the incoming light by corneal laser treatment is very appealing. Unfortunately, no evidence has been produced here.” This comment is puzzling. Our paper provides extensive evidence in the form of corneal topography demonstrating depressions in the surface of the eye and refractive changes following laser treatment (Fig. 4) and microperimetry exams that clearly show a shift to a new preferred retinal locus (PRL) along with increased fixation stability (Fig. 5). Thus, our laser treatment produced demonstrable changes in the optics of the eye, along with demonstrable changes in patient eye movements. Most importantly, our core result is that visual acuity improved at 12 months post-treatment (p < 0.002 for BCNVA and p < 0.004 for BCDVA) as did contrast sensitivity (p < 0.05). This is the bottom line for patients. “The comparison of 15 gained lines [sic, we think the reviewers meant “letters”] in comparison to untreated eyes is meaningless here - 11 gained (standard deviation is 13!!).” This comment does not appear to conform to accepted standards for clinical trials in ophthalmology: 1. It is common to use statistics to evaluate whether an effect present in data with scatter is significant or not. We found a very high level of significance for all conditions tested (p < 0.01). This analysis refutes the claim that the “comparison ... is meaningless...” 2. Many FDA-approved treatments that are part of the ophthalmology community’s standard-of-care are based on studies where the standard deviation exceeds the mean effect. For example, the effectiveness of aflibercept (Eylea) was reported to be a gain of 5.9 ± 13.8 (mean ± SD) letters of visual acuity at 12 months post-treatment (Eleftheriadou et al., Ophthalmol. Ther. 2018;7:361-368). Similar results are seen for 0.5 mg ranibizumab (Lucentis), which reported a gain of 7.2 letters with a standard error of ~1 letter for 240 patients (Rosenfeld et al., N. Engl. J. Med. 2006;355:1419-1431). The corresponding standard deviation is SD = 1 * sqrt(240) = 15.5 letters. Thus, according the reviewer’s standards, anti-VEGF treatments for wet AMD would be dismissed as “meaningless”, and papers describing the clinical results would not be publishable. 3. Another example is the Implantable Miniature Telescope (IMT), the FDA-approved device for treatment of dry AMD. In the initial IMT study (Lane SS, et al. Am J Ophthalmol 2004;137:993-1001), the mean (± SD) values of BCNVA letters gained were 11.5 (± 15.8) at 3m, 7.1 (± 17.4) at 6m and 8.5 (± 9.8) at 12m post-implantation. The BCDVA letters gained were 11.1 (± 11.5) at 3m, 12.1 (± 9.4) at 6m and 12.0 ± (8.1) at 12m post-implantation. These standard deviations are very large and the ratios of standard deviations/means are larger for most (4 of 6) of the IMT outcomes compared to the outcomes we cite in this paper. “In addition, the results are mixed with different subgroups and the statistics are not very meaningful - standard deviations are sometimes larger than the actual measured value. Here, it would be better to indicate the range.” We direct the reader to Figure 3, which shows the full distribution and range of treatment efficacies, for the change in visual acuity 12 months post-treatment. “The other eyes would have lost 4 lines [sic, we think the reviewers meant “letters”], because the disease pattern certainly does not allow statistics here and one would have to take the second eye (with bilateral disease) as reference.” The disease does not progress at the same rate in both eyes, so it is not clear how this control would be an improvement over citing results from the literature with many more subjects. “The number of cases is indeed very small, each eye accounts for 3 percent.” The paper clearly states, in Discussion, that one of the limitations of the study is a small sample size. That said, the improvement in visual acuity is large enough that the data provide high statistical certainty (p < 0.002 for BCNVA improvement at 12 months post-treatment), even for our “small” sample size. “The principle of operation is not comprehensible for the reader.” Laser irradiation caused a reduction in water content and an increase in modulus of the cornea (reference 18). In addition, we used corneal topography to directly observe a depression on the surface of the eye in the laser-treated regions (Fig. 4a). These depressions then changed the refractive power of the optics of the eye (Fig. 4b). By changing the eye’s refraction, light rays were deflected onto different locations on the retina. These are all experimentally-determined facts. Ultimately, we agree that there are aspects of how our treatment restores vision that are not completely understood and require further study. But many elements of the “principle of operation” underlying this treatment have been determined by data presented or cited in this paper. “The discussion with generation of four lenses, which overlap individually, is theory only. No proof has been made.” We presented measurements using corneal topography that showed that our laser treatment produced four regions of depression of the surface of the eye (Fig. 4). This is experimental data, not a theory. As far as the impact that these depressions will have on the refraction of light, the paper claims in Discussion that this effect “resembles four lenses”. This qualified statement, made in Discussion, is entirely appropriate. We stated in the previous manuscript that the full effects of this perturbation are not fully understood and will require further investigation. What is certain, however, is that these depressions will refract light differently than in the untreated eye. This statement is not “theory only”; it relies on Snell’s law, which has been known since at least the early 1600s. “The localization of the laser spots do not explain or justify postulated multiplication and redirection of light.” We have noted in the paper that “Further study is needed to determine patterns (and causes) of PRL changes following CPV treatment”. It is often the case that a discovery of a new effect precedes an understanding of the mechanism of action. To cite one example, the clinical efficacy of penicillin was well-established in the early 1940s but a complete understanding of its mechanism of action was not determined until the 1960s."
}
]
},
{
"id": "95172",
"date": "30 Sep 2021",
"name": "Mukharram M Bikbov",
"expertise": [
"Reviewer Expertise ophthalmology",
"corneal surgery"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe idea of laser remodeling of the cornea to create a new focus on the retina is very interesting, but what exactly this laser affects the cornea is not entirely clear.\nIf such studies had been conducted before, it would be interesting to make a link to them. It also remains unclear how the laser has a universal effect on corneas with different topography with different indicators of spherical and cylindrical components in all patients.\nThe technique is theoretically more similar to thermokeratoplasty and should have an effect similar to treating presbyopia, with an increase in the central refractive power of the cornea and an increase in near visual acuity, providing a kind of telescopic effect similar to intraocular lens implantation in AMD. It would be interesting to compare the effectiveness of treatment in one eye with the second eye of an intact patient.\nIt is also necessary to clarify the selection criterion for patients with macular lesion volume in dry AMD.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1500
|
https://f1000research.com/articles/9-1498/v1
|
23 Dec 20
|
{
"type": "Research Article",
"title": "Cloning of human ABCB11 gene in E. coli required the removal of an intragenic Pribnow-Schaller Box before it’s Insertion into genomic safe harbor AAVS1 site using CRISPR–Cas9",
"authors": [
"Nisha Vats",
"Madhusudana Girija Sanal",
"Senthil Kumar Venugopal",
"Pankaj Taneja",
"Shiv Kumar Sarin",
"Nisha Vats",
"Senthil Kumar Venugopal",
"Pankaj Taneja",
"Shiv Kumar Sarin"
],
"abstract": "Background: Genomic safe harbors are sites in the genome which are safe for gene insertion such that the inserted gene will function properly, and the disruption of the genomic location doesn’t cause any foreseeable risk to the host. The AAVS1 site is the genetic location which is disrupted upon integration of adeno associated virus (AAV) and is considered a ‘safe-harbor’ in human genome because about one-third of humans are infected with AAV and so far there is no apodictic evidence that AAV is pathogenic or disruption of AAVS1 causes any disease in man. Therefore, we chose to target the AAVS1 site for the insertion of ABCB11, a bile acid transporter which is defective in progressive familial intra hepatic cholestasis type-2 (PFIC-2), a lethal disease of children where cytotoxic bile salts accumulate inside hepatocytes killing them and eventually the patient. Methods: We used the CRISPR Cas9 a genome editing system to insert the ABCB11 gene at AAVS1 site in human cell-lines. Results: We found that human ABCB11 sequence has a “Pribnow- Schaller Box” which allows its expression in bacteria and expression of ABCB11 protein which is toxic to E. coli; the removal of this was required for successful cloning. We inserted ABCB11 at AAVS1 site in HEK 293T using CRISPR-Cas9 tool. We also found that the ABCB11 protein has similarity with E. coli endotoxin (lipid A) transporter MsbA. Conclusions: We inserted ABCB11 at AAVS1 site using CRISPR-Cas9; however, the frequency of homologous recombination was very low for this approach to be successful in vivo.",
"keywords": [
"Progressive Familial Intra Hepatic Cholestasis Type-2",
"CRISPR-Cas9",
"AAVS1 site",
"human ABCB11/BSEP",
"MsbA",
"E. Coli",
"Endotoxin",
"Lipid A transporter",
"Pribnow-Schaller Box",
"cloning"
],
"content": "Introduction\n\nProgressive familial intrahepatic cholestasis type-2 (PFIC2), a severe liver disease which is familial, neonatal, progressive and often fatal which results from a mutation of ATP binding cassette subfamily B member 11 (ABCB11) gene which codes for an ABC transporter bile salt export pump (BSEP)1,2. Mutations in ABCB11 gene results in accumulation of cytotoxic taurocholate and other cholate conjugates leading to progressive hepatocyte destruction1. Currently the definitive cure for PFIC2 is liver transplantation, which is limited by suitable donor organs. Gene therapy, allogenic hepatocyte transplantation3 and autologous transplantation of hepatocytes/liver organoids differentiated from ‘gene corrected’ induced pluripotent cells could be future options4,5. Adeno-associated virus (AAV) is so ubiquitous in man and animals that about 30% of the world population are positive for this virus and to date no disease is proven to be associated with this virus6–9. In our study we inserted ABCB11 gene at the AAVS1 site using CRISPR-Cas9 tool (Figure 1) in HEK293T cells and a fibroblast line.\n\nThe gene of interest (ABCB11) was flanked by 5’ and 3’ overhangs which are homologous to the AAVS1 site in chromosome 19. ABCB11 gene was driven by human promoter phospho-glycerol kinase.\n\n\nMethods\n\nWe PCR-amplified the 3966 bp ABCB11 (from cDNA prepared from total RNA of human liver tissue) using multiple overlapping primers (Extended data, Supp. Table 1)10 which were assembled by overlap extension PCR11. Phusion DNA polymerase (NEB, US Cat. #M0530L) was used as per the manufacturers protocol. Annealing temperate for all PCRs unless otherwise stated was 60°C C for 20s and an extension time 30s/kb at 72°C. Initial denaturation was carried out at 98°C for 30s and 5s in subsequent steps. In general, we used 24 cycles to amplify PCR products for cloning (and 32 cycles for other PCRs). We used 1.0 unit of the enzyme per 50 µl PCR reaction. The product was cloned in ‘donor’ vector having AAVS1 recombination overhangs on both ends and ampicillin resistance for selection (Extended data, Supp. Figure 1)10. The upstream (5’) overhang sequence (803 bp) in the vector was homologous to a sequence (NCBI Sequence ID: NC_000019.10, 55115768 to 55116570) inside the PPP1R12C gene (Figure 1), which would be in-frame with a 2A ribosome skip sequence and puromycin resistance gene if insertion of the construct happens by homologous recombination with the target site. This is region was followed by a poly adenylation signal and the downstream (3’) overhang sequence which was the continuation of the 5’ overhang (837 bp). We inserted the ABCB11 sequence driven by human phospho-glycerol kinase (hPGK) promoter between these overhang sequences. DH5α cells, JM 109 and One Shot™ Stbl3™ Chemically Competent E. coli cells (ThermoFisher Catalog#:C737303) were used for transformation and cloning.\n\nBacterial promoter prediction was done using BPROM, a prediction tool for bacterial promoters12. DNA or protein sequence comparisons were done using the appropriate tool from NCBI BLAST platform13. Primers were designed manually or using NCBI Primer Blast14.\n\nWe designed two guide RNAs targeting the AAVS1 site (Figure 1) and cloned in two vectors expressing SPCas9 and SACas9 at BsaI sites (Extended data, Supp. Figure 2)10 following the standard gRNA design, cloning protocols and resources15. Off-target analysis was done using the tool Custom Alt-R® CRISPR-Cas9 guide RNA16.\n\nHEK293T, HepG2 and FS1 (fibroblast) cells were grown in high glucose DMEM (Hi-Media Lab, Mumbai, Cat.# AL111-500ML) supplemented with 10% fetal bovine serum (CellClone, Genetix Biotech Asia, New Delhi, Cat.# CCS-500-SA-U), 1x penicillin (100U/ml) and streptomycin (100 µg/ml) (Hi-Media, Mumbai Cat. # A018-5X100ML). When 80% confluent, the cell lines were transfected with Cas9-sgRNA vectors (without the donor vector). At 48 h post-transfection, about 10000 of these cells were used for comet assay17 and genomic DNA (gDNA) isolated from the remaining cells was used for T7-endonuclease assay18 to evaluate the in vitro ‘DNA cutting’ activity of Cas9-sgRNA construct. Subsequently, we transfected these cells with the donor vector, Cas9-sgRNA vector and a control vector (pEGFPN1) in the ratio: 2:1:1 using PEI19 from Sigma Aldrich, Inc. (CAS #9002-98-6). After 48 h the cells were imaged and used for downstream applications. Half of the transfected dishes were serially passaged without puromycin selection for two weeks and DNA and protein were isolated. On the remaining dishes puromycin selection was started following the manufacturer’s protocol20 after 36 h of transfection and puromycin resistant colonies at 8 µg/mL were further cultured in puromycin containing media and gDNA was isolated.\n\nWhole-cell extracts (see Cell culture), scraped out and extracted using RIPA Lysis and Extraction Buffer, were run on 10% SDS-PAGE and transferred to a polyvinylidene difluoride membrane using a transfer apparatus following the standard protocols (Bio-Rad). After incubation with 5% nonfat milk in TBST (10 mM Tris, pH 8.0, 150 mM NaCl, 0.5% Tween 20) for 60 min, the membrane was washed once with TBST and incubated with rabbit antibodies against human ABCB11 (Affinity, Catalog #DF 9278) 1: 2000 dilution; human β-actin (Santa Cruz Cat.# SC4778), dilution 1:1000; 4°C overnight. The membrane was washed with TBST buffer and incubated with a 1:5000 dilution of horseradish peroxidase-conjugated anti-rabbit (Santa Cruz Cat# SC-2004)/anti-mouse antibodies (Cat.#SC-2005) for 2 h at room temperature. Blots were washed with TBST four times and developed with the ECL system (Bio-Rad, US Cat.#170-5060) according to the manufacturer’s protocols. Raw, uncropped images from western blotting are available as Underlying data21.\n\nT7 endo I assay detects heteroduplex DNA that results from annealing DNA stands that have been modified after a sgRNA/Cas9 mediated cut to DNA strands without modifications. T7 Endonuclease-1 was purchased from NEB, US (Cat. #NEB #E3321) and was used to digest the PCR products amplified from gDNA extracted from Cas9-sgRNA transfected (test) and un-transfected cells (control) using primers flanking the expected Cas9-sgRNA cut sites following the manufactures protocol18. PCR gel images are available as Underlying data21.\n\nA total of 50–100 cells treated as described in Cell culture were embedded in 0.7% low-melting agarose and mounted on a precoated slide and was immersed in alkaline 0.1% SDS solution overnight, neutralized and electrophoresis was done in an alkaline buffer (pH 10) at 0.74 V/ cm for 30 minutes17. Comet assay images are available as Underlying data21.\n\nSequencing of PCR products and plasmids were conducted by Invitrogen Bioservices India Pvt. Ltd., a part of Thermo Fisher Scientific, Whitefield, Bengaluru, PIN 560 066, India and Medauxin, Bengaluru, AMCO Colony, Koti Hosahalli, Bengaluru, PIN 560 092, India.\n\n\nResults\n\nFew ampicillin resistant DH5α E. coli colonies which we got after transformation were screened for the insert by colony PCR. One colony was positive for all the fragments of the ABCB11 gene. Sequencing revealed that mutations in ABCB11 (Figure 2). Repeated attempts failed and we considered the possibility of unstable DNA sequences. Therefore, we tried JM109 which gave one positive colony and plasmid was isolated. However, after we soon found the bacteria failing to grow or losing the plasmid on subsequent cultures. Therefore, we transformed One Shot™ Stbl3™ Chemically Competent E. coli cells, which are suitable for cloning unstable DNA segments. We got many positive colonies, however, upon overnight culture the bacteria formed a big pellet (partly lysed bacteria) which cannot be resuspended in phosphate buffered saline. Therefore, we concluded that the ABCB11 gene/gene product is toxic to bacterial cells. It is possible the ABCB11, being a membrane transporter, may be toxic to bacteria. Sequencing files are available as Underlying data21.\n\n(a) G to A mutation is marked for example. (b) In another example, sequence of another clone, mutation was in a different site.\n\nWe conducted PAGE followed by Coomassie staining to see differential protein expression between ABCB11 donor vector transformed bacteria versus untransformed bacteria (Figure 3a; Extended Data)21. We found differential expression of a few proteins. We subsequently performed a western blot and interestingly antibody against human ABCB11 identified a specific protein over expressed in the transformed bacterial cells (Figure 4b). However, in our construct ABCB11 gene was under a eukaryotic promoter. Considering the possibility of some DNA elements which have similarity to bacterial promoters inside the ABCB11 sequence we performed a bioinformatic analysis using BPROM to predict hidden bacterial promoters (Extended data, Supp. Table 2)10. The promoter-site (Pribnow-Schaller box tcatataat) containing sequence (ggttttgagtcagataaatcatataataat) which we identified was modified to (ggtTTCGAAtcagataaatcaTACAACaat) by PCR using an oligonucleotide primer sequence incorporating the modified sequence and subsequent overlap extension PCR amplification of the entire gene fragment. With this modification, we were able to clone ABCB11 coding sequence which was not toxic to bacteria.\n\n(a) Total protein extract from E. coli transformed the donor vector on Poly Acryl amide Gel Electrophoresis followed by Coomassie blue staining showed differential expression multiple proteins. (b) Western Blot with anti-human ABCB11 antibody shows multiple bands including one probably corresponding to MsbA- a bacterial Lipid A transporter. (c) A bioinformatic analysis (Protein BLAST) revealed ABCB11 and MsbA are sharing conserved domains. (d) Sequence alignment of ABCB11 and MsbA showing conserved domains.\n\n4a The PCR product (even without T7 digestion) shows a distinct band pattern resulting from the formation of heteroduplex. 4b. PCR amplified product after digestion with T7 endonuclease shows faint bands resulting from the digestion of heteroduplexes.\n\nA protein BLAST search identified MsbA (UniProtKB - P60752), a member of the ABC transporter superfamily. MsbA, a 64.46 kD protein, has an important role in E. coli Lipid A (endotoxin or LPS) transport (Figure 3c, d). This protein flips core endotoxin from its site of synthesis on the inner leaflet of the inner membrane to the outer leaflet of the inner membrane. western blot showed identified a unique band in the donor vector transformed E. coli while the untransformed E. coli also showed a faint band but specific band at the same position (Figure 3b).\n\nT7 Endonuclease Assay and Comet Assay were used to evaluate the gDNA cutting activity of Cas9-sgRNA. We observed digestion of heteroduplexes at the CRISPR-Cas9 cut sites which were sensitive to T7 endonuclease (Figure 4a). These heteroduplexes were observed on the agarose gel electrophoresis of PCR products as well (Figure 4b). The Cas9-sgRNA damaged the genome of the transfected cells leading to the formation of comet shaped nuclear material upon electrophoresis (Figure 5).\n\nThe untreated cells have intact round/oval nuclei while the Cas9-sgRNA treated cells shows a comet shaped nucleus because of DNA damage.\n\nOligonucleotide PCR primers were designed for bioinformatically predicted off-targets (Extended data, Supp. Tables 3a, b)10. We amplified these segments using genomic DNA extracted from the treated cells (48 h post-transfection with SPCas9-sgRNA vector) as template. The PCR products were sequenced and analyzed for sequence disruption (Table 1).\n\nOligonucleotide primers designed to amplify these off-target sites to verify off target disruptions.\n\nWestern blotting done with total protein extract of fibroblasts 48 h post-transfection with the ABCB11 donor vector showed the expression of ABCB11 protein (Figure 6a). A fibroblast line was used in these experiments because they don’t express ABCB11, naturally while cell lines such as HEK293T and liver cell lines such as HepG2 do express ABCB11.\n\n(a) Western Blot with anti-human ABCB11 antibody confirmed the expression after transient transfection with the donor vector having ABCB11. (b) Western Blot was done using total protein isolate from fibroblasts (which does not naturally express ABCB11) after four passages post-co-transfection of the donor vector with ABCB11 gene and the CRISPR-Cas9-sgRNA vector.\n\nWestern blotting was repeated with total protein extract from fibroblasts after 2 weeks (fourth passage) post-transfection with SPCas9-sgRNA vector together with the donor vector containing ABCB11. This blot also showed the expression of ABCB11 protein (Figure 6b) suggesting the integration of ABCB11 into the host genome.\n\nWe obtained only three puromycin resistant colonies upon transfecting about 20 million HEK cells with a transfection efficiency of 70 to 80% in four 6 cm dishes. The gDNA isolated from transfected cells (Cas9-sgRNA plasmid alone) after 72h, (Cas9-sgRNA plasmid plus donor vector) after 21 days of puromycin selection were used as PCR templates with a forward primer complementary to a region upstream of the 5’ recombination overhang of the vector and a reverse primer complementary to a sequence in the puromycin resistance gene to amplify a segment spanning from a site in the host cell genomic DNA slightly upstream of the genomic integration site to a segment donated by the donor vector (puromycin resistance gene). This PCR product was used as a template for a nested PCR and product was confirmed by restriction enzyme digestion (BamH1) and sequencing (Primers: Extended data, Supp. Table 410, Figure 7). We also PCR-amplified parts of ABCB11 using primers (Extended data, Supp. Table 1)10 which would give specific PCR products from the inserted cassette, to make sure the gene is not deleted from the cassette integrated to the host cell. PCR products showed the expected sizes confirming that the amplified products are from the cassette and not from the native ABCB11 gene present in the cell line (Figure 8).\n\n(a) The gDNA isolated from transfected cells (Cas9-sgRNA plasmid alone) after 72h, (Cas9-sgRNA plasmid plus donor vector) after 21 days of puromycin selection were used as PCR templates (lane 1, 2) with a forward primer complementary to a region upstream of the 5’ recombination overhang of the vector and a reverse primer complementary to a sequence in the puromycin resistance gene to amplify a segment spanning from a site in the host cell genomic DNA slightly upstream of the genomic integration site to a segment donated by the donor vector (puromycin resistance gene). Note that genomic DNA from untreated HEK293T cells did not give any products in the expected range (lane 3, 4). (b) The PCR product mentioned in (a) was used as a template for a nested PCR. (c) The PCR product mentioned in (b) was confirmed by restriction enzyme digestion (BamH1) and sequencing (Primers: Extended data, Supp.Table 4)10.\n\nABCB11-specific primers were used, which would give specific PCR products from the inserted cassette. This was done to make sure the ABCB11 gene was not deleted from the cassette integrated to the host cell genome. The PCR products showed the expected sizes confirming that the amplified products originated from the integrated cassette and not from the native ABCB11 gene present in the cell line.\n\n\nDiscussion\n\nTo our knowledge, this is the first time the ABCB11 gene was inserted into the AAVS1 safe-harbor using CRISPR-Cas9 technology in human cells. Liver directed gene therapy is another approach and was successful in rodents22. Adeno associated vectors do not integrate and therefore the effect of gene therapy many not last in human beings, especially in infants, as the viral vector dilutes out as the cells proliferate in a growing liver23. Another approach is transplantation of hepatocytes differentiated from gene corrected patient iPSC4,24,25. AAVS1 site is considered as a ‘safe haven’ in human genome26 where we chose to insert the gene mutated in PFIC. We could not find any other study which attempted to insert the ABCB11 gene at the AAVS1 site. AAV is a common virus and it is considered non-pathogenic because the seroprevalence of wild-type AAV in humans ranges from 40% for AAV8 to 70% for AAV1 and AAV2, yet; we are not aware of any disease caused by AAV6,7,27. AAV integration into AAVS1 site causes disruption of PPP1R12C (protein phosphatase 1 regulatory subunit 12C). However, this gene is not associated with any disease27. A puromycin gene was placed in the donor cassette such that the puromycin gene will be transcribed only if homologous recombination happens. We obtained only a few puromycin resistant colonies suggesting that homologous recombination was a rare event (~10-7). This suggests that in vivo gene therapy using CRISPR-Cas9 technology making use of homology directed gene repair could be difficult. “Targeted Integration and high transgene expression in AAVS1 Transgenic mice after in-vivo hematopoietic stem cell transduction with HDAd5/35++ Vectors”28 is reported; however, to achieve this they used an adenoviral gene delivery system with AAV5 ITRs and AAV35 helper. Integration of AAV/ Cas9 into Cas9 mediated cut sites is a potentially hazardous consequence of this approach29,30.\n\nWe found that the human ABCB11 donor vector transformed bacteria either died or the ABCB11 gene sequence got mutated meaning either the DNA sequence or the ABCB11 protein has some untoward effects on bacteria. We performed a western blot and found that in ABCB11-transformed bacterial clones giving a band around 45 kD and HepG2 cells/liver tissue is giving a band at around 140 kD which corresponds to ABCB11. It was interesting to note that untransformed E. coli cells are also showing a band although very faint around 45 kD which suggested the possibility of a bacterial protein which might have structural similarity to human ABCB11. We performed a bioinformatic search and identified MsbA an E. coli protein which functions as a lipid transporter (~64 kD). MsbA is involved in the transport of bacterial endotoxin-a function like the ABCB11 which transports the bile salts which are lipid derivatives31,32. Another interesting observation was the identification of a bacterial promoter sequence (Pribnow-Schaller Box) in human ABCB11 causing unexpected expression of ABCB11 protein and bacterial toxicity (Figure 9). This was an important lesson for us because we spent a lot of time trying to clone ABCB11. It is therefore important to search for and eliminate if any bacterial promoter sequences or similar elements are identified, for successful cloning of eukaryotic/toxic genes in E. coli. We do not know how exactly ABCB11 caused bacterial toxicity. Possibly, expression of ABCB11 in E. coli might be destabilizing the E. coli membrane, since ABCB11, being a lipid transporter, is a membrane-spanning protein.\n\nA Pribnow box was detected inside ABCB11, which allowed the gene to transcribe in E. coli, causing bacterial lysis, probably through competitive replacement of a homologous transporter protein in E. coli (E. coli Endotoxin (Lipid A) Transporter) MsbA, resulting in Lipid A (L) accumulation inside the bacteria.\n\nAlternatively, the human protein, which is similar to the E. coli protein might have caused a competition between the bacterial transporter MsbA and human protein for membrane incorporation resulting in the accumulation of endotoxin within E. coli cells because unlike the bacterial transporter the membrane incorporated ABCB11 might not be able to transport endotoxin out. This raises the possibility that endotoxin is toxic to E. coli itself if it is not exported and MsbA can therefore be considered as a drug target. This point required further validation (Figure 9). The advantage of such an antibiotic is that it will be selective to endotoxin producing microbes. More research is required in this direction. It may be noted that endotoxin producing microbes play an important role in sepsis33 and diseases such as non-cirrhotic portal hypertension34–36.\n\nTo conclude, we successfully inserted ABCB11 at the AAVS1 site using CRISPR-Cas9, however, the frequency of homologous recombination was very low as evident from the number of puromycin resistant colonies. With this low efficiency, with the current technology it is unlikely that this approach would be successful in in-vivo gene editing. It is worth, exploring MsBA as a novel antibiotic target for LPS producing bacteria, although our data in this direction is primitive and requires further validation.\n\n\nData availability\n\nHarvard Dataverse: Sequencing Data, supplementary data to Cloning of Human ABCB11 Gene in E. coli required the removal of an Intragenic Pribnow-Schaller Box before it’s Insertion into Genomic Safe Harbor AAVS1 Site using CRISPR Cas9. https://doi.org/10.7910/DVN/32TXCD21.\n\nThis project contains the following underlying data:\n\n2020_06_20_211607 actin.jpg. (Uncropped western blot image.)\n\n4a .T7 endonuclease digestion.jpg. (PCR gel image.)\n\n4b. PCR without T7 endonuclease.jpg (PCR gel image.)\n\n4c... Cas9-sgRNA treated cells comet assay.jpg. (Image taken from Comet assay, treated cells.)\n\n4c...... untreated cells comet-1.jpg. (Image taken from Comet assay, untreated cells.)\n\nABCB11_fibroblasts.jpg. (Uncropped western blot image.)\n\nABCB11_fragments.tif. (PCR gel image.)\n\nABCB11_WB_P4_2020_06_19_183004-1.tif. (Uncropped western blot image.)\n\nDH5a WB-ABCB11-long_Exposure.jpg. (Uncropped western blot image.)\n\nDH5a WB-ABCB11.jpg. (Uncropped western blot image.)\n\nE.Coli_PAGE_Coumasse.jpg. (Uncropped PAGE gel.)\n\nJM109_ABCB11 WB.tif. (Uncropped western blot image.)\n\nnested PCR product BamH1 digest.jpg. (PCR gel image.)\n\nNested PCR secondary.jpg. (PCR gel image.)\n\nNested_Primary pcr.jpg. (PCR gel image.)\n\nRepeat_WB_2020_07_11_181831.jpg. (Uncropped western blot image.)\n\nSequencing data.7z. (Sequencing data produced in the present study.)\n\nT7endo 30420202.jpg. (PCR gel image.)\n\nHarvard Dataverse: Supplementary Tables to Cloning of Human ABCB11 Gene in E. coli required the removal of an Intragenic Pribnow-Schaller Box before it’s Insertion into Genomic Safe Harbor AAVS1 Site using CRISPR Cas9. https://doi.org/10.7910/DVN/NTUOXM10.\n\nThis project contains the following extended data:\n\n- Supp-Tables-revised-HepInt.docx. (Supp. Tables 1–4.)\n\n- SuppFigures.pptx. (Supp. Figure 1a, b.)",
"appendix": "Acknowledgements\n\nThe corresponding author thanks Mr. Rahul Saha for his assistance in western blot.\n\n\nReferences\n\nAmer S, Hajira A: A comprehensive review of progressive familial intrahepatic cholestasis (PFIC): genetic disorders of hepatocanalicular transporters. Gastroenterology Res. 2014; 7(2): 39–43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGerloff T, Stieger B, Hagenbuch B, et al.: The sister of P-glycoprotein represents the canalicular bile salt export pump of mammalian liver. J Biol Chem. 1998; 273(16): 10046–10050. PubMed Abstract | Publisher Full Text\n\nSanal MG: Cell therapy from bench to bedside: Hepatocytes from fibroblasts - the truth and myth of transdifferentiation. World J Gastroenterol. 2015; 21(21): 6427–33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSanal MG: Future of liver transplantation: non-human primates for patient-specific organs from induced pluripotent stem cells. World J Gastroenterol. 2011; 17(32): 3684–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSanal MG: A highly efficient method for generation of therapeutic quality human pluripotent stem cells by using naive induced pluripotent stem cells nucleus for nuclear transfer. SAGE Open Med. 2014; 2: 2050312114550375. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKaplitt MG, Feigin A, Tang C, et al.: Safety and tolerability of gene therapy with an adeno-associated virus (AAV) borne GAD gene for Parkinson’s disease: an open label, phase I trial. Lancet. 2007; 369(9579): 2097–2105. PubMed Abstract | Publisher Full Text\n\nVandamme C, Adjali O, Mingozzi F: Unraveling the complex story of immune responses to AAV vectors trial after trial. Hum Gene Ther. 2017; 28(11): 1061–1074. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHüser D, Khalid D, Lutter T, et al.: High Prevalence of Infectious Adeno-associated Virus (AAV) in Human Peripheral Blood Mononuclear Cells Indicative of T Lymphocytes as Sites of AAV Persistence. J Virol. 2017; 91(4): e02137–16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmith JR, Maguire S, Davis LA, et al.: Robust, persistent transgene expression in human embryonic stem cells is achieved with AAVS1-targeted integration. Stem Cells. 2008; 26(2): 496–504. PubMed Abstract | Publisher Full Text\n\nSanal MG: Supplementary Tables to Cloning of Human ABCB11 Gene in E. coli required the removal of an Intragenic Pribnow-Schaller Box before it’ s Insertion into Genomic Safe Harbor AAVS1 Site using CRISPR Cas9. Harvard Dataverse, V1. 2020. http://www.doi.org/10.7910/DVN/NTUOXM\n\nLee J, Shin MK, Ryu DK, et al.: Insertion and deletion mutagenesis by overlap extension PCR. Methods Mol Biol. 2010; 634: 137–146. PubMed Abstract | Publisher Full Text\n\nSalamov VSA, Solovyevand A: Automatic annotation of microbial genomes and metagenomic sequences. Metagenomics and its …. 2011. Reference Source\n\nMadden T: The BLAST sequence analysis tool. The NCBI Handbook 2nd edition. 2013. Reference Source\n\nYe J, Coulouris G, Zaretskaya I, et al.: Primer-BLAST: a tool to design target-specific primers for polymerase chain reaction. BMC Bioinformatics. 2012; 13: 134. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAddgene: CRISPR Plasmids and Resources . [cited 2020 Jul 8]. Reference Source\n\nwww.idtdna.com/site/order/designtool/index/CRISPR_CUSTOM; [cited 2020 Jul 8]. Reference Source\n\nNandhakumar S, Parasuraman S, Shanmugam MM, et al.: Evaluation of DNA damage using single-cell gel electrophoresis (Comet Assay). J Pharmacol Pharmacother. 2011; 2(2): 107–111. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDetermining Genome Targeting Efficiency using T7 Endonuclease I NEB. [cited 2020 Jul 8]. Reference Source\n\nGoula D, Remy JS, Erbacher P, et al.: Size, diffusibility and transfection performance of linear PEI/DNA complexes in the mouse central nervous system. Gene Ther. 1998; 5(5): 712–717. PubMed Abstract | Publisher Full Text\n\nAntibiotic Kill Curve | Sigma-Aldrich. [cited 2020 Jul 9]. Reference Source\n\nSanal MG: Sequencing Data, supplementary data to Cloning of Human ABCB11 Gene in E. coli required the removal of an Intragenic Pribnow-Schaller Box before it’ s Insertion into Genomic Safe Harbor AAVS1 Site using CRISPR Cas9. Harvard Dataverse, V1. 2020. http://www.doi.org/10.7910/DVN/32TXCD\n\nAronson SJ, Bakker RS, Shi X, et al.: Liver-directed gene therapy results in long-term correction of progressive familial intrahepatic cholestasis type 3 in mice. J Hepatol. 2019; 71(1): 153–162. PubMed Abstract | Publisher Full Text\n\nCoppoletta JM, Wolbach SB: Body Length and Organ Weights of Infants and Children: A Study of the Body Length and Normal Weights of the More Important Vital Organs of the Body between Birth and Twelve Years of Age. Am J Pathol. 1933; 9(1): 55–70. PubMed Abstract | Free Full Text\n\nNagamoto Y, Takayama K, Ohashi K, et al.: Transplantation of a human iPSC-derived hepatocyte sheet increases survival in mice with acute liver failure. J Hepatol. 2016; 64(5): 1068–1075. PubMed Abstract | Publisher Full Text\n\nSanal MG: Personalized Medicine in Cell Therapy and Transplantation. In: Barh D., Dhawan D., Ganguly N. (eds) Omics for Personalized Medicine. Springer. 2013; 775–799. Publisher Full Text\n\nPapapetrou EP, Schambach A: Gene insertion into genomic safe harbors for human gene therapy. Mol Ther. 2016; 24(4): 678–684. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKay MA, Nakai H: Looking into the safety of AAV vectors. Nature. 2003; 424(6946): 251–251. PubMed Abstract | Publisher Full Text\n\nLi C, Psatha N, Wang H, et al.: Integrating hdad5/35++ vectors as a new platform for HSC gene therapy of hemoglobinopathies. Mol Ther Methods Clin Dev. 2018; 9: 142–152. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHanlon KS, Kleinstiver BP, Garcia SP, et al.: High levels of AAV vector integration into CRISPR-induced DNA breaks. Nat Commun. 2019; 10(1): 4439. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChakraborty S: Sequencing data from Massachusetts General Hospital shows Cas9 integration into the genome, highlighting a serious hazard in gene-editing therapeutics [version 1; peer review: 1 approved with reservations]. F1000Res. 2019; 8: 1846.\n\nChang G, Roth CB: Structure of MsbA from E. coli: a homolog of the multidrug resistance ATP binding cassette (ABC) transporters. Science. 2001; 293(5536): 1793–1800. PubMed Abstract | Publisher Full Text\n\nWard A, Reyes CL, Yu J, et al.: Flexibility in the ABC transporter MsbA: Alternating access with a twist. Proc Natl Acad Sci USA. 2007; 104(48): 19005–19010. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOpal SM: Endotoxins and other sepsis triggers. Contrib Nephrol. 2010; 167: 14–24. PubMed Abstract | Publisher Full Text\n\nKhanna R, Sarin SK: Non-cirrhotic portal hypertension - diagnosis and management. J Hepatol. 2014; 60(2): 421–441. PubMed Abstract | Publisher Full Text\n\nBi XJ, Chen MH, Wang JH, et al.: Effect of endotoxin on portal hemodynamic in rats. World J Gastroenterol. 2002; 8(3): 528–530. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKamath PS, Tyce GM, Miller VM, et al.: Endothelin-1 modulates intrahepatic resistance in a rat model of noncirrhotic portal hypertension. Hepatology. 1999; 30(2): 401–407. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "81026",
"date": "06 Apr 2021",
"name": "Mitradas Panicker",
"expertise": [
"Reviewer Expertise Molecular Neurobiology",
"Stem Cells",
"Neuroscience",
"Cell Biology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting and useful observation regarding the cloning of the human ABCB11 gene in E.coli prior to its original goal i.e. inserting ABCB11 into the AASV1 safe harbor.\nThe aim was to insert the ABCB11 gene into human cells to establish the initial steps for gene therapy in liver. The authors found that they could not clone their construct of the gene in E. coli. Careful observation and analysis led them to a Pribnow-Schaller Box within the sequence. They proved their hypothesis by modifying the Box and successfully cloning the gene in E. Coli. The work is detailed and carefully done and also suggests further areas such as the potential use of MsbA as a drug target.\nComments:\nThe title could be corrected to reflect the result 'proper' since the removal of the Pribnow-Schaller box is not a necessary requirement for insertion into the AASV1 locus. It became a necessity due the cloning method that had to be adopted.\nThe manuscript requires some additional editing for clarity.\n\nFigure 1 would benefit by indicating the site of the 2A insertion, the hPGK sequence should be marked as a promoter. It is not clear what is the ABCB11 codon optimization for and also what it was optimized for as noted in Fig.1.\nIt would be more accessible if the Pribnow-Schaller Box and the sequences changed are depicted as a figure than as a supplementary table and a few lines of text. If text is preferred then the sequence numbers of the nucleotides involved and the reference sequence should be provided in the text.\nThe observation that homologous insertion was obtained at very low frequency does not necessarily preclude the strategy from being used 'in vivo' since the frequency would be highly dependent on the strategy and the sequences chosen. For e.g. a homology-independent strategy with same construct might prove fruitful. Perhaps the discussion could reflect this.\nThe supplementary Figure of the Donor Vector does not show the ABCB11 sequence. This should also be addressed.\nThe observation by Chang and Gray should be more clearly emphasized in the Discussion. MsbA as a drug target should also refer Zhang et al. (2018)1.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "81735",
"date": "07 Apr 2021",
"name": "Nagendra Chaturvedi",
"expertise": [
"Reviewer Expertise Molecular Cell Biology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript is well-designed with a valuable contribution to the successful cloning of ABCB11, a key defective gene in PFIC2.\nI only have a few minor comments:\nIs that common to see mutations in ABCB11 gene while expressing it in the prokaryotic or eukaryotic expression system?\n\nDo authors know the functional consequences if this gene is not mutated in the expression system?\n\nTitle of manuscript is too long.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1498
|
https://f1000research.com/articles/9-1017/v1
|
21 Aug 20
|
{
"type": "Opinion Article",
"title": "Effects of rapid recruitment and dissemination on Covid-19 mortality: the RECOVERY trial",
"authors": [
"Catherine Knowlson",
"David J. Torgerson",
"Catherine Knowlson"
],
"abstract": "The RECOVERY trial is a large multi-armed, adaptive randomised controlled trial of treatments for Covid-19. It has rapidly recruited and demonstrated that hydroxychloroquine is ineffective in reducing mortality for hospitalised patients, whilst dexamethasone significantly reduces mortality among those patients using supplemental oxygen or on a ventilator. We estimate that the speed of recruitment and dissemination has probably decreased mortality in the UK by at least 200 hospitalised patients in the first month since the British Prime Minister announced the results. Despite its impressive speed, the trial only recruited about 15% of eligible patients, with recruitment rates ranging between 3% to 80% at participating hospitals. Had the trial recruited 50% of the eligible patients then our analysis suggests that more than 2,000 additional lives could have been saved. In a pandemic, rapid recruitment with high centre recruitment is absolutely essential to reduce deaths. Methods of improving site specific recruitment rates need investigating urgently.",
"keywords": [
"Covid-19",
"RECOVERY trial",
"recruitment"
],
"content": "Introduction\n\nThe RECOVERY randomised controlled trial (RCT) is a world-leading study of potential treatments for Covid-19 patients. It is large, simple, adaptive and multi-armed, allowing the investigators to test several treatments at the same time and quickly close trial arms if one of the treatments were found to be effective or ineffective. Importantly, this means the results can be rapidly disseminated to patients, clinicians and policy makers. The trial initially randomised Covid-19 patients admitted to hospital to one of five treatments: lopinavir-ritonavir (an HIV treatment); hydroxychloroquine; dexamethasone; azithromycin or usual care. The protocol was kept simple and flexible to allow “a broad range of patients to be enrolled in large numbers” (RECOVERY study protocol1). Uniquely, the trial started recruiting nine days after submission of its protocol2.\n\nThe trial has rapidly produced some key findings on the effectiveness and ineffectiveness of potential Covid-19 treatments. Its earliest, and somewhat overlooked, finding that hydroxychloroquine was ineffective (and probably harmful) to Covid-19 patients was important given that it has been widely promoted and used3. If the point estimate, of harm, in the hydroxychloroquine comparison is correct, then many lives will be saved worldwide by its, hopeful, reduction in use. Most recently the trial has demonstrated that lopinavir-ritonavir is also ineffective4. The most widely publicised finding from the RECOVERY trial, however, was that of the dexamethasone arm, which statistically significantly reduced mortality among Covid-19 patients at 28 days after randomisation3. This important result was demonstrated in less than three months after the trial was set-up. Within 81 days the trialists recruited 175 hospitals and enrolled 11,303 participants with 9,355 eligible for the dexamethasone comparison3. This remarkable trial will lead to many hundreds of lives saved across the UK and the world and it is a tribute to the investigators and all those who took part either as participants, clinicians or researchers.\n\nThe RECOVERY trial is unique in its rapid recruitment and the speed at which it reported its first findings. Most trials, however, recruit relatively poorly and slowly, and therefore do not report their results in a timely fashion. RECOVERY did not recruit slowly but arguably it did recruit poorly. It has been reported that the overall recruitment rate to RECOVERY was 15%2 of Covid-19 inpatients, with participating hospital recruitment ranging between 3% and 80% of eligible patients being recruited2. In this respect, RECOVERY exhibited similar characteristics of the ‘typical’ non-Covid trial undertaken within an NHS setting: some recruitment sites enrol a very high proportion of eligible patients while others recruit relatively low numbers. Indeed, it is rare for all sites, or the majority, to consistently recruit a high proportion of eligible participants5,6.\n\nFor the ‘standard’ trial (and for RECOVERY) to ensure rapid recruitment in the presence of poor average site recruitment, many more sites have to be enrolled in the study than would be required if there was high recruitment in all clinical sites, or recruitment of the target number takes longer than expected. However, if all sites could recruit the same proportion of eligible participants as the best recruiting site then trials would be finished more rapidly. This would have the benefits of reducing the overall cost of the trial and, most importantly, would improve patients’ health and save more lives. In ‘normal’ times this trade-off in lost lives and reduced quality of life, due to low recruitment, is not identified because either the data are not routinely collected (e.g., quality of life) or it is not collated so that it can be quantified. Slow or poor recruitment is even more catastrophic during a pandemic as there is a brief window of opportunity to recruit and complete a trial to enable infected patients to benefit from novel treatments. Therefore, whilst the clinical trials community in the UK has led the world in rapid, large and effective clinical trials to identify new treatments for Covid-19 there is still room for improvement. In this paper we look at the potential impact of the RECOVERY results on the numbers of patients surviving since the dexamethasone results were reported and then discuss the likely consequences of the RECOVERY trial’s ability to recruit only 15%2 or less of the UK’s hospitalised Covid-19 patients into the trial.\n\n\nMethods\n\nTo examine the actual impact of the RECOVERY trial on lives saved and its ‘potential’ impact had recruitment been even more swift, we used UK estimates of hospital admissions due to Covid-197–10. We assumed that the proportion of patients that were eligible was the same as described in the RECOVERY trial, as well as the proportions on oxygen and ventilation. We used admissions data from the 16th June 2020 (date of the release of the trial results) until 15th July 202: we also assumed that 83% of admitted patients had no contraindications to dexamethasone. However, in line with the RECOVERY results we assumed that 24% of admitted patients did not need either oxygen or ventilator support so would not be offered the dexamethasone.\n\nThe RECOVERY trial recruited 15% of patients with Covid-19 in UK hospitals. There was a huge variation in recruitment rates across the trial, which ranged from 3% to 80% of eligible participants. Recruitment started on the 19th March 2020 with rapid accrual of hospitals (132 participating hospitals by 3rd April) and by the 8th of June 2020 (with 175 hospitals open to recruitment), 11,303 patients had been randomised in total. Of these, 9,355 were randomised into the steroid comparison so this part of the study closed to recruitment3. Assuming an overall 15% recruitment rate, then this implies there were 75,353 patients with Covid-19 in UK hospitals during the recruitment period (although routine statistics suggest that there had been 114,935 Covid-19 admissions across the UK by this date7–10). Making the following assumptions we can estimate the possible loss of life by not recruiting a greater proportion of Covid-19 patients. In our following calculations we assume that on average 50% of eligible patients would take part in the RECOVERY trial if asked. Therefore, to enrol 11,303 patients then we would have to identify 22,606 patients admitted to NHS hospitals with COVID-19. We estimate this target would have been reached by the 1st April (as 24,978 COVID-19 patients had been admitted by this point7–10). The RECOVERY trial’s preliminary results were released by the British Prime Minister eight days after recruitment was completed, which would have taken us to the 9th April 2020 (by which time 48,075 patients had been admitted to hospital in the UK). Between the 9th April 2020, when the results could have been available, and the 15th July 2020 there were 77,310 patients admitted with Covid-197–10. To estimated the number of lives which could have been saved by the earlier completion of the dexamethasone arm, we made the following assumptions base on the RECOVERY trial results: that 83% of admitted patients had no contraindications to dexamethasone, and that 24% of admitted patients did not need either oxygen or ventilator support so would not be offered the dexamethasone.\n\n\nResults\n\nIn Table 1 we show the estimated lives saved in this first month of dexamethasone being made available to all eligible patients (assuming that all hospitals implemented the guidelines without delay). In this month there were approximately 6,980 patients admitted to hospital with Covid-19, which equates to an estimated 5,793 patients who had no contraindications for dexamethasone treatment. Table 1 shows that in just over a month more than 200 extra patients in the UK survived in hospital due to wider use of dexamethasone.\n\n*Assumes steroids are not given to hospitalised but not oxygenated patients as per the results from the RECOVERY trial.\n\n**Adjusted rather than observed differences between groups are used, which are 12.3 and 4.2% reduction in 28-day mortality for ventilated and oxygen supported patients, respectively.\n\nIn Table 2, using the RECOVERY data we have estimated the potential benefit had all the participating hospitals recruited 50% of their eligible patients to RECOVERY (which should be achievable as clinical experience suggests that the vast majority of patients were happy to be included in the trial2, although we are assuming there are no other large Covid-19 studies which would have caused competition for participants) and the dexamethasone recruitment was halted at 9,355 patients and the results were available by the 9th April.\n\n*Assumes steroids are not given to hospitalised but not oxygenated patients as per the results from the RECOVERY trial.\n\n**Adjusted rather than observed differences between groups are used, which are 12.3 and 4.2% point reduction in 28-day mortality for ventilated and oxygen supported patients, respectively.\n\nThe table shows that by not achieving the best recruitment which some UK hospitals are capable of means around 2,880 patients died unnecessarily.\n\n\nDiscussion\n\nThere is a need to complete and report all trials more quickly. This is especially the case in a pandemic. A reason why the RECOVERY trial could be done in the UK is due to the strong research infrastructure and having a national health service. However, we could do better. During the height of the pandemic, government advisors in the daily briefing encouraged patients and their doctors to take part in clinical trials. Whilst some hospitals recruited a remarkable 80% of eligible patients many did less well or did not take part2. If some hospitals can recruit such high proportions of participants, then the majority should be able to do so. We understand that hospitals will be under more pressure than normal, especially when the number of cases are high, which may reduce their ability to recruit. However, if there are no proven treatments available yet, we would argue that the best care for affected patients would be to offer participation in a study to help identify an effective treatment. If there is a second wave of the disease over the winter then measures need to be put into place to ensure that all eligible patients are offered the chance to take part in a clinical trial: swift action in recruitment will save more lives.\n\nThere has been some criticism of the RECOVERY trialists for reporting their results by press conference rather than in a peer reviewed journal2. The peer-reviewed paper published in the New England Journal of Medicine3 on July 17th 2020 had only trivial differences from the basic data released on the 16th June 2020. Had the trialists waited for the peer reviewed paper to be published before having a press conference then it is likely over 200 patients in the UK would have died, plus many more internationally. Consequently, the rapid dissemination of results, in our view, was justified.\n\n\nConclusions\n\nRapid recruitment and dissemination in the RECOVERY trial has, we estimate, saved at least 200 lives in the UK in first month since the trial’s results were released. However, we have estimated that the number lives saved, had the recruitment rate been at least 50% of eligible patients, would have been an order of magnitude greater.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.",
"appendix": "Contributors\n\nDT conceived the paper and wrote the first draft and CK revised and expanded the paper and identified data sources. Both authors contributed to the final approved manuscript. DT acts as the guarantor and affirms that the manuscript is an honest and transparent account of the study.\n\n\nReferences\n\nRandomised evaluation of COVID-19 Therapy (RECOVERY). [Accessed 20.07.2020]. Reference Source\n\nWise J, Coombes R: Covid-19: The inside story of the RECOVERY trial. BMJ. 2020; 370: m2670. PubMed Abstract | Publisher Full Text\n\nRECOVERY Collaborative Group, Horby P, Lim WS, et al.: Dexamethasone in Hospitalized Patients with Covid-19 - Preliminary Report. N Engl J Med. 2020; NEJMoa2021436. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNo clinical benefit from use of lopinavir-ritonavir in hospitalised COVID-19 patients studied in RECOVERY. [Accessed 20.07.2020]. Reference Source\n\nBruhn H, Treweek S, Duncan A, et al.: Estimating Site Performance (ESP): can trial managers predict recruitment success at trial sites? An exploratory study. Trials. 2019; 20(1): 192. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHurtado-Chong A, Joeris A, Hess D, et al.: Improving site selection in clinical studies: a standardised, objective, multistep method and first experience results. BMJ Open. 2017; 7(7): e014796. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCOVID-19 Admissions to Hospital (England). [Accessed 17.07.2020]. Reference Source\n\nCOVID-19 Statistical Report (Public Health Scotland). [Accessed 17.07.2020]. Reference Source\n\nNHS activity and capacity during the coronavirus (COVID-19) pandemic: 16 July 2020 (Wales). [Accessed 17.07.2020]. Reference Source\n\nConfirmed COVID-19 Daily Admissions by HSC Trust (Northern Ireland). [Accessed 17.07.2020]. Reference Source"
}
|
[
{
"id": "74833",
"date": "27 Nov 2020",
"name": "Carlos J. Chaccour",
"expertise": [
"Reviewer Expertise Internal medicine",
"Infectious diseases",
"epidemiology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this opinion piece, Knowlson and Torgerson analyze the recruitment rate and recruitment success of the RECOVERY trial. They point to a large variation between sites ranging from 15 to 80% of all eligible patients and argue that a much faster completion of the trial was feasible if all sites had recruited at least 50% of eligible patients. The authors then proceed to estimate the lives lost due to poor recruitment delaying trial completion. This analysis is based on the proven efficacy of dexamethasone.\n\nThe opinion piece is indeed provoking and well thought. The introduction is well written and sets the stage appropriately. The methods and assumptions are described at large and the results mostly support the conclusion.\nAs an opinion piece, I think it could further improve by mentioning other potential causes for slow/poor recruitment, such as whether hydroxychloroquine, azithromycin or any of the RECOVERY interventions were used outside the trial.\n\nThe authors make clear that their conclusion is based on the effect of dexamethasone and I think this sufficiently illustrates the point. However, as much as RECOVERY served to support the scale-up of dexamethasone, it served to reduce the compassionate use of hydroxychloroquine which often comes with additional risk of harms. Additional consideration could be given to the lives saved by a reduction in the use of hydroxychloroquine.\n\nMinor:\nThere is a zero missing in the phrase: “from the 16th June 2020 (date of the release of the trial results) until 15th July 202:”. The assumption: \"83% of patients without contraindications and 24% not needing oxygen or ICU\" is mentioned twice.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly",
"responses": [
{
"c_id": "6202",
"date": "23 Dec 2020",
"name": "David Torgerson",
"role": "Author Response",
"response": "With respect to the first reviewer (Carlos Chaccour) we have undertaken the minor corrections he suggested in his review. We have also added an additional paragraph in the discussion regarding the point he made about the impact of the RECOVERY findings on off label hydroxychloroquine use, which, worldwide, would also have an impact on reducing harm. We include this in our updated version of the paper."
}
]
},
{
"id": "74832",
"date": "09 Dec 2020",
"name": "Heidi R. Gardner",
"expertise": [
"Reviewer Expertise Clinical trials methodology",
"participant recruitment",
"mixed methods research",
"inclusivity in trials",
"trial efficiency"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThank you for requesting that I review this interesting and very timely article. I enjoyed reading this, as usual from these authors the manuscript is well written, robustly reported, and refreshingly concise.\nThe range of recruitment rates from 3% to 80% was shocking to me, and I'd like to see more detail in this if possible. Are the authors able to break these rates down to individual hospitals or health boards? I'd be interested to know if the low recruiters had high rates of COVID and vice versa to see if there's a correlation. More broadly this information could provide insights into how various sites view trials - are they a priority? If not, why not? If so, what's contributing to that culture and how can we replicate it in other sites?\nThe other thing that I feel is lacking in this paper is discussion around WHO the participants are. Recruiting is all well and good, but if RECOVERY recruited only white British people and no minorities, then that's clearly a problem - particularly given the disparities around the impact of COVID on different ethnic minority groups in our society. The fact that RECOVERY doesn't appear to have collected data on the ethnicity of participants is problematic, and deserves a comment here.\nThe implications of this work are potentially wide-ranging, and have got me thinking about what research area(s) should be focused on to ensure that the day-to-day usual trials that are run in the UK are as effective (and quick) as they can be. I'd like to see what the authors think about this. Should we be thinking about engagement with the public, involvement of public and patient partners, research into how the option to take part in a trial could potentially be a part of routine care? The fact that patients have died as a result of poor recruitment in RECOVERY is a fantastic starting point for discussion - the numbers are stark, but what to do with them now?\nOverall, this is an important and thought-provoking article, and one that should act as a wake-up call to those of us working to improve trials. There is much work to be done in this area.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": [
{
"c_id": "6203",
"date": "23 Dec 2020",
"name": "David Torgerson",
"role": "Author Response",
"response": "With respect to the second reviewer (Heidi Gardener) we very much agree with you that it would be interesting to find out what the differences were between the participating hospitals that caused the wide range in recruitment rate, as this would be incredibly valuable for current and future trials. Unfortunately, as we are not involved in the RECOVERY trial we do not have access to data broken down by site to determine whether there is a correlation between poor recruitment and the number of covid admissions. You have definitely touched upon an interesting area for consideration, around how different sites view trials. Although this was outside the scope of this study, it is certainly something we would be interested in researching in the future. You brought up another excellent point about who the participants are that participated in the trial. In two of the three published RECOVERY papers these data are not reported. However, one of the papers does report the proportion of participants from ethnic minorities and we have updated the paper to include a section on this in the discussion, which as far as we can see there does not seem to be a disparity. However, unfortunately we are unable to look at BAME admission rates across the participating sites to know for sure if there is any disparity as we do not have access to these data. Again, this is a potential area of future research if the RECOVERY team can provide this information. It would be fantastic if we could improve recruitment rates at those sites which do not deliver to target. We think that there is unlikely to be a single factor that is involved, and it will very much depend on individual sites and reasons may change over time too (e.g. site capacity, competing studies, Principal Investigator engagement). Indeed, we hope that the increased media attention on clinical trials recently will encourage more patients to seek participation in trials, however we think that the biggest effect would be at a site level as we believe that all patients should be given the opportunity to take part in a trial. We have addressed these issues in the updated version of our paper."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1017
|
https://f1000research.com/articles/9-1295/v1
|
03 Nov 20
|
{
"type": "Opinion Article",
"title": "Cancer T-cell therapy: building the foundation for a cure",
"authors": [
"Alexander Kamb",
"William Y. Go",
"William Y. Go"
],
"abstract": "T-cell cancer therapy is a clinical field flush with opportunity. It is part of the revolution in immuno-oncology, most apparent in the dramatic clinical success of PD-1/CTLA-4 antibodies and chimeric antigen receptor T-cells (CAR-Ts) to cure certain melanomas and lymphomas, respectively. Therapeutics based on T cells ultimately hold more promise because of their capacity to carry out complex behaviors and their ease of modification via genetic engineering. But to overcome the substantial obstacles of effective solid-tumor treatment, T-cell therapy must access novel molecular targets or exploit existing ones in new ways. As always, tumor selectivity is the key. T-cell therapy has the potential to address target opportunities afforded by its own unique capacity for signal integration and high sensitivity. With a history of breathtaking innovation, the scientific foundation for the cellular modality has often been bypassed in favor of rapid advance in the clinic. This situation is changing, as the mechanistic basis for activity of CAR-Ts and TCR-Ts is backfilled by painstaking, systematic experiments—harking back to last century’s evolution and maturation of the small-molecule drug discovery field.\n\nWe believe this trend must continue for T-cell therapy to reach its enormous potential. We support an approach that integrates sound reductionist scientific principles with well-informed, thorough preclinical and translational clinical experiments.",
"keywords": [
"CAR",
"TCR",
"cancer",
"mechanism of action",
"clinical translation",
"innovation"
],
"content": "\n\nIf you have built castles in the air, your work need not be lost; that is where they should be. Now put the foundations under them.\n\nHenry David Thoreau, Walden 1854\n\n\nT-cell therapies are the future of oncology\n\nIt is astounding how the contents of the typical pharmacy have changed over the last 100 years. A century ago, pharmacists stocked their shelves with aspirin, opiates, mercury, arsenic, magnesium sulfate, iodine and a few other substances of legitimate medical value (Pharmacopoeia of the US, 1907). Since then, hundreds of small-molecule drugs, dozens of recombinant antibodies, and even a few nucleic acid therapeutics have been proven by rigorous scientific and clinical studies to treat a wide variety of human ailments. It is likely, however, that for a large number of patients yet to enjoy effective remedies for their disease, including cancer, cell therapy will ultimately provide the solution.\n\nThis prediction follows from the inherent strengths of cells as therapeutic entities. T cells, for example, are honed by evolution to execute numerous complex biological functions, among them identification and elimination of infected or damaged tissue (Janeway et al., 1999). They have tremendous natural advantage over other therapeutic modalities that are often limited to a single activity: binding to other molecules. Simple binding behavior may be sufficient to trigger salutary physiological changes and, indeed, there are many examples. However, the limitations imposed by having only hundreds of atoms like small molecules, or even thousands like antibodies, is evident. T cells, on the other hand, are composed of thousands of different molecules, prewired by evolution to work in concert to accomplish tasks of extraordinary complexity (Janeway et al., 1999). Specific killing is one of the simpler cellular behaviors, and is therefore among the first successful achievements of T-cell therapy, exemplified by three CD19-targeting chimeric antigen receptor T-cells (CAR-T cells) registered or close to registration (Abramson, 2020; Neelapu et al., 2017; Neelapu et al., 2020a; Schuster et al., 2019). The next frontier for engineered T-cell therapy is solid tumors, which pose additional challenges. But cells have a second huge advantage as a therapeutic option: they can be readily manipulated with genetic alterations to augment or suppress their natural behaviors. The methods to do this are now routine and are improving with the advent of newer technologies such as CRISPR/Cas9 (Cong et al., 2013; Jinek et al., 2012). Combined with cellular reprogramming technologies, the possibilities to modulate natural cell properties or even create emergent ones are wide open (Takahashi & Yamanaka, 2006; Yu et al., 2007). T cells are naturally endowed with the attributes of (i) outstanding sensitivity, able to detect a handful of molecules on a cell surface; (ii) multivariate signal integration, permitting them to react to different environments and discriminate among a variety of cell types; and, (iii) the capacity to proliferate. These traits are exactly those needed to overcome obstacles posed by solid tumor therapy.\n\n\nWe need to build a robust mechanistic foundation\n\nTo overcome the obstacles to solid tumor therapy, we must first recognize certain facts. A hallmark of the T-cell therapy field is striking innovation, with towering figures such as S.A. Rosenberg who has spent 40 years spearheading the clinical use of T cells in cancer (Fisher et al., 1989; Yron et al., 1980). Others, including G. Gross and Z. Eshhar (CAR), M.R. Roberts and M.H. Finer (Gen2 CAR), and V.D. Fedorov and M. Sadelain (iCAR) have designed robust novel receptors that can substitute for, or extend, T-cell receptor (TCR) function (Fedorov et al., 2013; Gross et al., 1989; Roberts et al., 1994).\n\nNotwithstanding the innovation and clinical success, the field lacks a strong foundation of mechanistic understanding. For example, it is still unclear how the TCR transduces peptide major histocompatibility complex (pMHC) ligand-binding into a cellular response, let alone how CARs signal (see for review Courtney et al., 2018; Nerreter et al., 2019). These gaps impede progress in areas that need to be addressed so that solid tumors can reliably and predictably be treated. It is instructive to draw an analogy with small-molecule drug discovery, a field that developed over the 20th century from rudimentary industrial processes to a highly sophisticated discipline of quantitative structure-activity relationships based on structural chemistry, computational modeling, and pharmacodynamic analysis in vitro and in vivo (Figure 1).\n\nThe goal is to control variables and improve the predictability of substantive advances.\n\nAs an emerging field, engineered T-cell therapy is not on a similarly solid footing. The standard suite of in vitro assays is crude when compared to those used in modern small-molecule or antibody optimization laboratories. Assays that vary effector:target ratios are convenient, but have high background and poor dynamic range. They are typically insensitive and subject to conflation of important biological variables; for instance, T cell proliferation and cytotoxicity as well as target-cell proliferation over time (Rossi et al., 2018). Primary human T cells are heterogeneous and cumbersome to grow, and the relationship between them and model cell lines, such as Jurkat, is not well understood (Salter & Creswell, 1986). Murine cancer models must also trade off tractability with relevance, and have some obvious prima facie weaknesses. Assays of therapeutic efficacy and safety in murine models are notoriously unpredictive for clinical behavior (Kamb, 2005). In immuno-oncology specifically, even the best models use syngeneic grafts that do not originate in the host and, though matched at MHC, contain hundreds of nonsynonymous mutations and elicit immune response1. Many experiments employ chimeric murine models with a complicated mixture of murine and human immune components (e.g., humanized murine models, patient-derived xenografts). These models have utility and are chosen for practical reasons, but they are often regarded as decisive in selection of clinical candidates because of presumptive experimental supremacy. In our view this is specious. The ultimate destination of a clinical candidate is the complex milieu of the human body and specifically the tumor microenvironment. But understanding the steps that must occur, one by one, to achieve a successful outcome in the clinic should not be dismissed as irrelevant just because they are studied outside the system biology of a human body. In vivo experiments should be used and interpreted judiciously in the context of robust in vitro data.\n\nReferencing small-molecule discovery again, the most successful efforts have involved deliberate construction of a mechanistic picture; from biochemical assays, through cell-based assays, to cautiously interpreted in vivo testing of pharmacodynamics. A clear example is the history of imatinib’s discovery (Buchdunger et al., 1996). T-cell therapy would benefit from adoption of this approach to control as many of the variables as possible within a reasonable timeframe of drug discovery. Only then can the predictability of the discovery process improve to the point needed to address the challenges of solid tumor therapy. If we wish to continue to innovate and not settle for incremental advances to CD19-directed therapies where there are currently hundreds of ongoing clinical trials for an unmet need, now estimated at ~6,000 deaths/year in the US, we must improve the mechanistic understanding and economical testing of candidate therapeutics. Otherwise, the opportunity costs will be enormous.\n\n\nThe acute shortage of solid-tumor drug targets: targeting genetic gains and losses\n\nSelectivity is the supreme challenge of oncology. At the genetic level, a tumor differs on average at ~10,000 nucleotide positions from the normal tissues from which it arose—less than 0.01% of the human genome (Vogelstein et al., 2013). In contrast, siblings differ by about 10 million nucleotides. Perhaps even worse from a conventional therapeutic perspective, very few of these genetic changes are shared among a significant percentage of cancers. Only a handful of mutations, such as mutant KRAS and P53, occur at frequencies above 5% of cancers. The vast majority are private mutations unique to each tumor. For decades, drug discoverers have searched for “magic bullets” that can discriminate reliably among tumor and normal cells, with some success. Good examples include imatinib for chronic myeloid leukemia, which inhibits the Abl kinase, and rituximab, a CD20 antibody that mediates the destruction of B-cell lineage cells such as non-Hodgkin lymphoma (Anderson et al., 1997; Buchdunger et al., 1996). Both these medicines are extremely effective within the subset of cancers they are designed to treat. In solid tumors, there are a few proteins, known loosely as tumor-selective antigens, whose expression is sufficiently limited in adult normal tissues that they continue to attract attention as possible cancer targets. These include CEA, MSLN, PSMA, and the MAGE family members (Lu et al., 2017; Parkhurst et al., 2011).\n\nIn 2001 the complete human gene list of ~20,000 was defined, establishing a boundary for new discoveries. Cancer researchers have scoured this gene set for the last two decades with diminishing success, visible in the shrinking, overlapping group of cancer targets swarmed by academic research laboratories and pharma/biotech industry R&D organizations. We desperately need new options; and these will likely require utilization of known gene products in novel ways. Immuno-oncology offers prospects for doing so. The large majority of recurrent somatic mutations affect proteins expressed inside cells. Thus, it is necessary to overcome the barrier of the cell membrane that excludes antibodies and most other macromolecules to exploit somatic mutations as a source of selective cancer targets. The immune system has evolved the means to do so through the aegis of antigen presentation. Molecular complexes of major histocompatibility antigens bound to peptides derived from cellular proteins (pMHCs) give T cells a view of the internal contents of cells. Some of these pMHCs are likely the basis for PD-1 antibodies’ and tumor infiltrating lymphocytes’ (TILs) remarkable power to trigger tumor-specific killing by the immune system (Chamoto et al., 2020; Hinrichs & Rosenberg, 2014). pMHCs that contain mutant peptides are currently the intended targets for numerous investigational vaccines and T-cell therapy efforts to engineer or select neoantigen-reactive T cells (Castle et al., 2019; Ng et al., 2019). The small number of recurrent mutations constrain the target options on this front. Though there are dozens—even hundreds—more private neoantigens, therapeutic exploitation of these via T cell engineering presents other challenges (Ng et al., 2019).\n\nLoss of genetic material, rather than gain of somatic mutations, represents another opportunity to achieve absolute discrimination at the genetic level between tumor and normal cells. The most common form of genetic loss in cancer is loss of heterozygosity (LOH). An astonishing 20% of the genome in a typical cancer cell exhibits LOH. These LOH regions include loci that encode polymorphic surface antigens that can be recognized by T cells. Genetic loss is irrevocable and furnishes a basis for discrimination, provided a method can be devised to take advantage of LOH. The workings of a primordial branch of the immune system show the way. Natural killer (NK) cells, which evolved before the adaptive immune system, employ a system of signal integration that differentiates self from non-self by combining inputs from families of activating and inhibitory receptors (Bryceson & Long, 2008). The logic of the NK system has been reproduced in an artificial circuit involving CARs (Fedorov et al., 2013). Versions of this basic circuit are capable in principle of utilizing LOH as a black-and-white difference between tumor and normal cells (Hamburger et al., 2020). These attempts to widen the target source for selective cancer targets, both neoantigens and LOH, are in their early stages, but they hold promise to dramatically increase the therapeutic options available for solid tumor patients.\n\n\nAdditional challenges for T-cell therapy\n\nThe justifiable excitement around cancer T-cell therapy must be balanced with acknowledgement that many significant challenges remain beyond tumor-selective targeting. Difficulties in T-cell manufacturing and delivery to patients translate into high production costs and time-delays (Fiorenza et al., 2020; Locke et al., 2020). Despite the technical hurdles, we view these issues as solvable through the iterative improvement cycles that are part of the standard practice of engineers. Efforts to automate, miniaturize and accelerate the production of autologous cells are underway (Castella et al., 2020). The opportunity to improve efficiency seems extremely attractive because the current doses of T cells can exceed 100 billion cells—orders of magnitude beyond the number involved in a typical immune response in the body (Gudmundsdottir et al., 1999). Meanwhile, production methods for off-the-shelf allogeneic cell products have demonstrated early clinical success (Neelapu et al., 2020b).\n\nPerhaps more significant, efficacy to date in solid tumors is unimpressive and safety issues, either off- or on-target, continue to plague clinical programs (Lu et al., 2017; Norberg et al., 2020; Parkhurst et al., 2011). We believe these problems are also solvable. They will be addressed by biological solutions, as they are not generally the result of limits imposed by laws of physics and chemistry which constrain more mature modalities. Indeed, there are a myriad of levers to pull to improve T-cell therapy outcomes. In some respects, the opportunity set for improved design of T-cell therapeutics is so large, that the challenge is to prioritize and test the possibilities efficiently.\n\n\nAn approach to future T-cell therapeutic discovery\n\nWe do not subscribe to the common view that human testing always trumps preclinical data, not because it is false in concept, but because it is problematic in practice. Variation in the clinic is typically large, the number of observations small, the expense high and timelines long (Locke et al., 2020; Silbert et al., 2019). We believe that well-designed preclinical experiments, interpreted within a solid framework of pharmacology and biology, will greatly aid analysis of clinical results, and in the long run support translational innovation that saves lives.\n\nTo this end, we propose a roadmap that begins by reducing the problem of solid tumor cell therapy into its components (Figure 2). These components incorporate essential requirements for solid tumor cell therapy to achieve efficacy and safety, including that the engineered cells must: (i) survive in the body post infusion; (ii) migrate through the body’s tissues into the tumor microenvironment; (iii) overcome the potentially anti-inflammatory environment of the tumor; (iv) specifically recognize the tumor cells in a vast excess of normal cells; and, (iv) deliver a sustained cytotoxic blow sufficient to remove most, if not all, of the tumor bulk. These component activities can be parsed into scientific disciplines of biochemistry, pharmacology, cell biology, immunology, and tissue/organismal physiology.\n\nTSA, Tumor Specific Antigen; pMHC, peptide-major histocompatibility antigen; LOH, loss of heterozygosity.\n\nThere are many widely-held assumptions that have not been tested systematically, and the field would benefit from their thorough examination (Table 1). It would be useful to have sufficiently large datasets to delineate the connection between tractable models and the more complicated preclinical systems, and ultimately, the clinic. The collective time and expense on the one hand, and risk of irrelevant or non-robust results on the other, create significant overhangs for the field. Effort should be directed toward providing clear evidence to connect receptor properties to function, and T cell lines to primary cells. Given the potential importance of long-term survival and function of T cells for curative treatment of solid tumors, there is a pressing need for plausible in vitro models of chronic T cell activity. It is impractical to funnel large numbers of candidate receptors through in vivo models. This foundation-building work may not be glamorous, but is of great consequence and should be valued by scientific journals. If the field as a whole invests to build the infrastructure and expertise of better preclinical models and larger datasets, and allocates time to define key mechanistic details prior to clinical testing, we believe the risks required to develop inventive, differentiated therapies will be rewarded with success.\n\n\nConclusion\n\nThe head of Novartis’ drug discovery organization, J. Bradner, reportedly expressed the opinion last year that “money and scientific resources are being poured into attempts to make incremental progress at a time when there is an urgent need for disruptive change” (Usdin, 2019). We agree with this perspective, but would add that without proper investment in foundational understanding of the science and technology, efforts to innovate further engineered T-cell therapies are likely to bog down in frustrating unpredictability. Risk tolerance must be wedded to broad, deep preclinical datasets that enable better prediction of outcomes on the clinical frontier.\n\n\nData availability\n\nNo data is associated with this article.",
"appendix": "Footnotes\n\n1 https://www.criver.com/sites/default/files/resources/Whole-ExomeSomaticMutationAnalysisofMouseCancerModelsandImplicationsforPreclinicalImmunomodulatoryDrugDevelopment.pdf.\n\n\nReferences\n\nAbramson JS: Anti-CD19 CAR T-Cell Therapy for B-Cell Non-Hodgkin Lymphoma. Transfus Med Rev. 2020; 34(1): 29–33. PubMed Abstract | Publisher Full Text\n\nAnderson DR, Grillo-López A, Varns C, et al.: Targeted anti-cancer therapy using rituximab, a chimaeric anti-CD20 antibody (IDEC-C2B8) in the treatment of non-Hodgkin's B-cell lymphoma. Biochem Soc Trans. 1997; 25(2): 705–8. PubMed Abstract | Publisher Full Text\n\nBryceson YT, Long EO: Line of attack: NK cell specificity and integration of signals. Curr Opin Immunol. 2008; 20(3): 344–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBuchdunger E, Zimmermann J, Mett H, et al.: Inhibition of the Abl protein-tyrosine kinase in vitro and in vivo by a 2-phenylaminopyrimidine derivative. Cancer Res. 1996; 56(1): 100–4. PubMed Abstract\n\nCastella M, Caballero-Baños M, Ortiz-Maldonado V, et al.: Point-Of-Care CAR T-Cell Production (ARI-0001) Using a Closed Semi-automatic Bioreactor: Experience From an Academic Phase I Clinical Trial. Front Immunol. 2020; 11: 482. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCastle JC, Uduman M, Pabla S, et al.: Mutation-Derived Neoantigens for Cancer Immunotherapy. Front Immunol. 2019; 10: 1856. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChamoto K, Hatae R, Honjo T: Current issues and perspectives in PD-1 blockade cancer immunotherapy. Int J Clin Oncol. 2020; 25(5): 790–800. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCommittee of Revision and Publication of the Pharmacopoeia of the United States of America, United States Pharmacopœial Convention: The pharmacopœia of the United States of America. 8th decennial revision / [United States]: Board of Trustees. 1907. Reference Source\n\nCong L, Ann Ran F, Cox D, et al.: Multiplex genome engineering using CRISPR/Cas systems. Science. 2013; 339(6121): 819–23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCourtney AH, Lo WL, Weiss A: TCR Signaling: Mechanisms of Initiation and Propagation. Trends Biochem Sci. 2018; 43(2): 108–123. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFedorov VD, Themeli M, Sadelain M: PD-1- and CTLA-4-based inhibitory chimeric antigen receptors (iCARs) divert off-target immunotherapy responses. Sci Transl Med. 2013; 5(215): 215ra172. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFiorenza S, Ritchie DS, Ramsey SD, et al.: Value and affordability of CAR T-cell therapy in the United States. Bone Marrow Transplant. 2020; 55(9): 1706–1715. PubMed Abstract | Publisher Full Text\n\nFisher B, Packard BS, Read EJ, et al.: Tumor localization of adoptively transferred indium-111 labeled tumor infiltrating lymphocytes in patients with metastatic melanoma. J Clin Oncol. 1989; 7(2): 250–61. PubMed Abstract | Publisher Full Text\n\nGross G, Waks T, Eshhar Z: Expression of immunoglobulin-T-cell receptor chimeric molecules as functional receptors with antibody-type specificity. Proc Natl Acad Sci U S A. 1989; 86(24): 10024–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGudmundsdottir H, Wells AD, Turka LA: Dynamics and requirements of T cell clonal expansion in vivo at the single-cell level: effector function is linked to proliferative capacity. J Immunol. 1999; 162(9): 5212–23. PubMed Abstract\n\nHamburger AE, DiAndreth B, Cui J, et al.: Engineered T cells directed at tumors with defined allelic loss. Mol Immunol. 2020; S0161-5890(20)30488-0. PubMed Abstract | Publisher Full Text\n\nHinrichs CS, Rosenberg SA: Exploiting the curative potential of adoptive T-cell therapy for cancer. Immunol Rev. 2014; 257(1): 56–71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJaneway CA Jr, Travers P, Capra JD: Immunobiology: The Immune System in Health and Disease. Garland Publishers, New York. 1999. Reference Source\n\nJinek M, Chylinski K, Fonfara I, et al.: A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity. Science. 2012; 337(6096): 816–2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKamb A: What's wrong with our cancer models? Nat Rev Drug Discov. 2005; 4(2): 161–5. PubMed Abstract | Publisher Full Text\n\nLocke FL, Go WY, Neelapu SS: Development and Use of the Anti-CD19 Chimeric Antigen Receptor T-Cell Therapy Axicabtagene Ciloleucel in Large B-Cell Lymphoma: A Review. JAMA Oncol. 2020; 6(2): 281–290. PubMed Abstract | Publisher Full Text\n\nLu YC, Parker LL, Lu T, et al.: Treatment of Patients With Metastatic Cancer Using a Major Histocompatibility Complex Class II-Restricted T-Cell Receptor Targeting the Cancer Germline Antigen MAGE-A3. J Clin Oncol. 2017; 35(29): 3322–3329. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNeelapu SS, Locke FL, Bartlett NL, et al.: Axicabtagene Ciloleucel CAR T-Cell Therapy in Refractory Large B-Cell Lymphoma. N Engl J Med. 2017; 377(26): 2531–2544. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNeelapu SS, Jacobson CA, Oluwole OO, et al.: Outcomes of older patients in ZUMA-1, a pivotal study of axicabtagene ciloleucel in refractory large B-cell lymphoma. Blood. 2020a; 135(23): 2106–2109. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNeelapu, et al.: ASCO. 2020b.\n\nNerreter T, Letschert S, Götz R, et al.: Super-resolution microscopy reveals ultra-low CD19 expression on myeloma cells that triggers elimination by CD19 CAR-T. Nat Commun. 2019; 10(1): 3137. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNg AHC, Peng S, Xu AM, et al.: MATE-Seq: microfluidic antigen-TCR engagement sequencing. Lab Chip. 2019; 19(18): 3011–3021. PubMed Abstract | Publisher Full Text\n\nNorberg S, Nagarsheth N, Sinkoe A, et al.: Safety and clinical activity of gene-engineered T-cell therapy targeting HPV-16 E7 for epithelial cancers. J Clin Oncol. 2020; 38(15_suppl): 101. Publisher Full Text\n\nParkhurst MR, Yang JC, Langan RC, et al.: T cells targeting carcinoembryonic antigen can mediate regression of metastatic colorectal cancer but induce severe transient colitis. Mol Ther. 2011; 19(3): 620–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoberts MR, Qin L, Zhang D, et al.: Targeting of human immunodeficiency virus-infected cells by CD8+ T lymphocytes armed with universal T-cell receptors. Blood. 1994; 84(9): 2878–89. PubMed Abstract | Publisher Full Text\n\nRossi J, Paczkowski P, Shen YW, et al.: Preinfusion polyfunctional anti-CD19 chimeric antigen receptor T cells are associated with clinical outcomes in NHL. Blood. 2018; 132(8): 804–814. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSalter RD, Creswell P: Impaired assembly and transport of HLA-A and -B antigens in a mutant TxB cell hybrid. EMBO J. 1986; 5(5): 943–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchuster SJ, Bishop MR, Tam CS, et al.: Tisagenlecleucel in Adult Relapsed or Refractory Diffuse Large B-Cell Lymphoma. N Engl J Med. 2019; 380(1): 45–56. PubMed Abstract | Publisher Full Text\n\nSilbert S, Yanik GA, Shuman AG: How Should We Determine the Value of CAR T-Cell Therapy? AMA J Ethics. 2019; 21(10): E844–851. PubMed Abstract | Publisher Full Text\n\nTakahashi K, Yamanaki S: Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell. 2006; 126(4): 663–76. PubMed Abstract | Publisher Full Text\n\nUsdin: Biocentury. 2019.\n\nVogelstein B, Papadopoulos N, Velculescu VE, et al.: Cancer genome landscapes. Science. 2013; 339(6127): 1546–58. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYron I, Wood Jr TA, Spiess J, et al.: In vitro growth of murine T cells. V. The isolation and growth of lymphoid cells infiltrating syngeneic solid tumors. J Immunol. 1980; 125(1): 238–45. PubMed Abstract\n\nYu J, Vodyanik MA, Smuga-Otto K, et al.: Induced pluripotent stem cell lines derived from human somatic cells. Science. 2007; 318(5858): 1917–20. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "74217",
"date": "17 Nov 2020",
"name": "John R. James",
"expertise": [
"Reviewer Expertise T cell signalling",
"signal transduction",
"reductionist approaches",
"Synthetic biology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe review article from Kamb and Go provides an overview of how T-cell based therapies are being used in cancer treatment. The challenges arising from this approach and how the foundation of T-cell therapy research can be built for development of more potent treatment of solid tumours are discussed. Their main point is that in the rush to get CAR-T therapy to patients, some of the underlying foundational research has been bypassed, which needs to be ‘filled in’ so the potential benefits of T-cell therapies can be fully realised, which is of course an important concern to raise.\nThere are a few points that should be addressed to improve this version of the manuscript:\nThe paper would be clearer if the authors could inform about the distinct challenges of CAR-T therapy used for treating blood cancers compared to those of solid tumours. At times, the information is overlapped and slightly unclear.\n\nIn the early discussion of significant players in the CAR-T field, it is remiss not to state the contribution of Carl June’s lab. While much of his group’s work has primarily been in leukaemia rather than solid tumours, it has nonetheless provided real impetus that this approach could be transformational in cancer therapies.\n\nThere is a slight pessimism to the state of knowledge on the mechanism of TCR triggering; while no true consensus will ever be reached on this question, there is little doubt that the fundamental aspects of this signal transduction have been elucidated.\n\nThe authors compare the state-of-the-art development of small-molecule drugs to the equivalent process for T-cell based therapies. They argue that mouse models are not appropriate tools to study (human) immuno-oncology, which is of course strictly true but a charge that can be just as easily levelled at small-molecule drug approaches too and so perhaps unfair for T-cell therapies to be singled out.\n\nThere is no mention of BiTE or ImmTAC therapeutics as alternative T-cell based therapies, which do have potentially greater likelihood of being effective in treating solid tumour masses.\n\nThe authors have listed four additional requirements for effective and safe solid tumour therapy (page 5 under heading Additional challenges for T-cell therapy) along with identifying drug target. The review could do well with more information on these listed points such as current research being carried out to address these limitations.\n\nThere are many labs around the world trying to combine engineering approaches to provide ‘logic-gating’ to CAR-T cell targeting. As the authors state, targets are hard to come by, but the potential for combinatorial CAR-T inputs (AND/NAND/NOT gating) significantly extends the usefulness of some likely targets to more accurately define solid tumour targets.\n\nTable 1 describes some commonly-held assumptions about T-cell therapies “not yet rigorously tested by mechanistic data”. The authors do provide a basis for these assumptions but no references to back these up. Whose “commonly held assumptions” are they?\n\nThe authors state: “This foundation-building work may not be glamorous but is of great consequence and should be valued by scientific journals. If the field as a whole invests to build the infrastructure and expertise of better preclinical models and larger datasets and allocates time to define key mechanistic details prior to clinical testing, we believe the risks required to develop inventive, differentiated therapies will be rewarded with success.” This point should be elaborated on to explain the roles of pharmaceutical companies, scientists and research institutes. Who takes the “unglamourous” job of foundation building? We would argue that academia is taking these risks and doing the foundational work; perhaps the point is aimed more at Pharma that they should also invest more heavily in this work too?\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly",
"responses": [
{
"c_id": "6197",
"date": "22 Dec 2020",
"name": "Alexander Kamb",
"role": "Author Response",
"response": "We thank the reviewers for their constructive comments and believe they have understood our key point. We have attempted to address most of the reviewers’ suggestions in the planned revised publication. We point out that we intend our paper to be an opinion or perspective, and not a review. Consequently, we have limited some of the references and discussion. The paper would be clearer if the authors could inform about the distinct challenges of CAR-T therapy used for treating blood cancers compared to those of solid tumours. At times, the information is overlapped and slightly unclear. We agree and have added text to clarify the specific challenges of solid tumors. Perhaps most dramatically, infused T-cell therapeutics directed against solid tumors must extravasate to reach their targets, targets that may be present on a subset of vital normal tissues as well. In the early discussion of significant players in the CAR-T field, it is remiss not to state the contribution of Carl June’s lab. While much of his group’s work has primarily been in leukaemia rather than solid tumours, it has nonetheless provided real impetus that this approach could be transformational in cancer therapies. We agree and have included Dr. June’s name and referenced his contributions. Still others have made substantive contributions to understanding, design and development of next-generation CAR-Ts; for example, C. June and P. Greenberg (see for review Guedan et al., 2019). There is a slight pessimism to the state of knowledge on the mechanism of TCR triggering; while no true consensus will ever be reached on this question, there is little doubt that the fundamental aspects of this signal transduction have been elucidated. We do not intend pessimism, and have clarified our view that, though important basic mechanistic questions remain (e.g., altered-peptide ligands, APLs), the TCR and CAR signaling mechanisms are understood in outline at least (a good example is the work of Dr. James): For example, there is not a broadly accepted model that explains key behavior of TCRs with respect to sensitivity and selectivity toward their ligands, peptide major histocompatibility complexes (pMHCs). CAR signaling, though understood in outline, also lacks important details (see for review Courtney et al., 2018; Nerreter et al., 2019). The authors compare the state-of-the-art development of small-molecule drugs to the equivalent process for T-cell based therapies. They argue that mouse models are not appropriate tools to study (human) immuno-oncology, which is of course strictly true but a charge that can be just as easily levelled at small-molecule drug approaches too and so perhaps unfair for T-cell therapies to be singled out. We agree wholeheartedly and have clarified this point: These deficits apply to small- and large-molecule therapeutic discovery. There is no mention of BiTE or ImmTAC therapeutics as alternative T-cell based therapies, which do have potentially greater likelihood of being effective in treating solid tumour masses. We know these modalities well, but believe that cell therapy holds more promise for solid tumor therapies. Cells can be engineered, if they do not do so already, to distribute into tissues. Large molecules (soluble proteins) are much more limited in what they can be engineered to do, beyond binding things, and are constrained by their physico-chemical properties. The authors have listed four additional requirements for effective and safe solid tumour therapy (page 5 under heading Additional challenges for T-cell therapy) along with identifying drug target. The review could do well with more information on these listed points such as current research being carried out to address these limitations. We should be clear that we do not intend to review the topic; our publication is more properly classified as an opinion piece. These topics are beyond the scope of our paper, and there are numerous reviews in the literature. There are many labs around the world trying to combine engineering approaches to provide ‘logic-gating’ to CAR-T cell targeting. As the authors state, targets are hard to come by, but the potential for combinatorial CAR-T inputs (AND/NAND/NOT gating) significantly extends the usefulness of some likely targets to more accurately define solid tumour targets. We have added a sentence to emphasize this point; i.e., that there are other logic systems beyond the AND NOT logic we describe in brief: Other approaches are under study, including transcriptional logic circuits and receptor masking (Roybal et al., 1995; Desnoyers et al., 2013). These attempts to widen the target source for selective cancer targets to other targets, including neoantigens and LOH… Table 1 describes some commonly-held assumptions about T-cell therapies “not yet rigorously tested by mechanistic data”. The authors do provide a basis for these assumptions but no references to back these up. Whose “commonly held assumptions” are they? We encounter people with different subsets of these assumptions frequently, but it is difficult to provide a suitable reference. We have changed the wording of the Table 1 title and in the text: There are many potential differences between, for example, TCRs and CARs which have not been tested systematically, and the field would benefit from their thorough examination (Table 1). We are certainly open to alternative phrasing. The authors state: “This foundation-building work may not be glamorous but is of great consequence and should be valued by scientific journals. If the field as a whole invests to build the infrastructure and expertise of better preclinical models and larger datasets and allocates time to define key mechanistic details prior to clinical testing, we believe the risks required to develop inventive, differentiated therapies will be rewarded with success.” This point should be elaborated on to explain the roles of pharmaceutical companies, scientists and research institutes. Who takes the “unglamourous” job of foundation building? We would argue that academia is taking these risks and doing the foundational work; perhaps the point is aimed more at Pharma that they should also invest more heavily in this work too? We strongly agree that there should be investment in foundation-building academic research by granting agencies, and have added this opinion explicitly: Our strong view is that granting agencies should invest in foundation-building academic research, in part because shorter-term translational work is often attractive to the private sector."
}
]
},
{
"id": "74214",
"date": "30 Nov 2020",
"name": "Barbra J. Sasu",
"expertise": [
"Reviewer Expertise T cell biology",
"CAR T"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe review deals with an important topic in T cell therapy - the use of appropriate methodology to increase mechanistic understanding and hopefully, eventually, translatability to the clinic.\nAt the outset, the authors quote some pioneers in the field of T cell and CAR therapy. Although it’s not possible to add everyone, I would suggest that the addition of Phil Greenberg for his long history of pioneering work on TCR T cells and understanding how to apply engineering to these cells. Perhaps also Mike Jensen and Carl June for understanding the nature of first and second gen CAR T and scientists such as Malcolm Brenner for insights into competitive expansion of CARs in vivo are some suggestions for additions. Analogy between development of SM and T cell therapeutics is an interesting comparison. Text in diagram should perhaps be bigger and QSAR needs to be defined in the legend.\nThe paragraph about building PD or mechanistic assays in the same spirit as the SM field is valuable but could be fleshed out more. E.g. The authors comment correctly that most work has to be done with primary cells since Jurkat unable to kill or behave in many other ways like normal T cells. There is reference to heterogeneousness, but perhaps calling out specifically that there is large donor to donor variability would be valuable. Adding to in vivo model difficulties I might talk about mouse cytokine environment not supporting human cells without model modification and in syngeneic models, inherent difficulties between human and mouse T cells such as the need for different signaling strengths.\n\nWhen pointing out shortcomings of screening approaches, say it needs to be more mechanistic. Some for instances might be useful, perhaps discussing possibilities for assays that might apply to parts of diagram 2, rather than the traditional endpoints in the T cell field of cytokine secretion, exhaustion markers or killing. Comments that in vivo assay can’t deal with high throughput of candidates is true, but there is the potential for rapid in vivo assay to potentially look at certain aspects covered in diagram e.g. migration, activation, specificity that are hard to cover in vitro. Perhaps a compare on contrast on this would be useful.\nHighlight other potential approaches to increasing tuor specificity e.g. synthetic biology from Wendel Lim or masking would be useful rather than just outlining one approach.\nQuotes that cell therapy doses can exceed 100 billion cells - seems like an outlier and more normal ranges should be quoted both for CARs and TCRs, especially in light of the table comparing CARs and TCRs and in fact cell dose could be added to this table.\n\nIt was unclear to me if the table is meant to state dogma or the belief of authors. Some of the assumptions already have data challenging them and discussing some of this as a ‘start of the journey’ may be valuable, for example that CARs show good combination with PD-1 Abs in preclinical models.\nAt the end, the authors make strong statements that better assays are needed, which I can’t argue with. Perhaps compare and contrast some assays and say what areas merit more development would be good. Potentials for solutions would make this review more valuable and might stimulate some of the general advances in the field that the review calls for. The review deals with an important topic in T cell therapy - the use of appropriate methodology to increase mechanistic understanding and hopefully, eventually, translatability to the clinic.\nAt the outset, the authors quote some pioneers in the field of T cell and CAR therapy. Although it’s not possible to add everyone, I would suggest that the addition of Phil Greenberg for his long history of pioneering work on TCR T cells and understanding how to apply engineering to these cells. Perhaps also Mike Jensen and Carl June for understanding the nature of first- and second-gen CAR T and scientists such as Malcolm Brenner for insights into competitive expansion of CARs in vivo are some suggestions for additions.\n\nAnalogy between development of SM and T cell therapeutics is an interesting comparison. Text in diagram should perhaps be bigger and QSAR needs to be defined in the legend.\nHighlight other potential approaches to increasing tuor specificity e.g. synthetic biology from Wendel Lim or masking would be useful rather than just outlining one approach.\nQuotes that cell therapy doses can exceed 100 billion cells - seems like an outlier and more normal ranges should be quoted both for CARs and TCRs, especially in light of the table comparing CARs and TCRs and in fact cell dose could be added to this table.\n\nIt was unclear to me if the table is meant to state dogma or the belief of authors. Some of the assumptions already have data challenging them and discussing some of this as a ‘start of the journey’ may be valuable, for example that CARs show good combination with PD-1 Abs in preclinical models.\nAt the end, the authors make strong statements that better assays are needed, which I can’t argue with. Perhaps compare and contrast some assays and say what areas merit more development would be good. Potentials for solutions would make this review more valuable and might stimulate some of the general advances in the field that the review calls for.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly",
"responses": [
{
"c_id": "6198",
"date": "22 Dec 2020",
"name": "Alexander Kamb",
"role": "Author Response",
"response": "At the outset, the authors quote some pioneers in the field of T cell and CAR therapy. Although it’s not possible to add everyone, I would suggest that the addition of Phil Greenberg for his long history of pioneering work on TCR T cells and understanding how to apply engineering to these cells. Perhaps also Mike Jensen and Carl June for understanding the nature of first and second gen CAR T and scientists such as Malcolm Brenner for insights into competitive expansion of CARs in vivo are some suggestions for additions. We have mentioned Drs. Greenberg and June in the revision, but ask the reviewer to bear in mind that this is an opinion or perspective, not a review. We have added “An opinion” to the title to clarify this. Analogy between development of SM and T cell therapeutics is an interesting comparison. Text in diagram should perhaps be bigger and QSAR needs to be defined in the legend. We have defined QSAR and requested the additional change in size. The paragraph about building PD or mechanistic assays in the same spirit as the SM field is valuable but could be fleshed out more. E.g. The authors comment correctly that most work has to be done with primary cells since Jurkat unable to kill or behave in many other ways like normal T cells. There is reference to heterogeneousness, but perhaps calling out specifically that there is large donor to donor variability would be valuable. We have included text to call this variability out specifically: …with considerable donor-to-donor variability; Adding to in vivo model difficulties I might talk about mouse cytokine environment not supporting human cells without model modification and in syngeneic models, inherent difficulties between human and mouse T cells such as the need for different signaling strengths. We have added text to highlight this specific problem (i.e., the mismatch between mouse/human cytokine signaling that can be understood partly as divergence between ligands and receptor pairs over 90 million years of evolutions (e.g., IL-2 ligand and receptor): The human and mouse components of these chimeras, e.g., IL-2 and IL-2R, do not mesh perfectly (Nemoto et al., 1995). When pointing out shortcomings of screening approaches, say it needs to be more mechanistic. Some for instances might be useful, perhaps discussing possibilities for assays that might apply to parts of diagram 2, rather than the traditional endpoints in the T cell field of cytokine secretion, exhaustion markers or killing. Comments that in vivo assay can’t deal with high throughput of candidates is true, but there is the potential for rapid in vivo assay to potentially look at certain aspects covered in diagram e.g. migration, activation, specificity that are hard to cover in vitro. Perhaps a compare on contrast on this would be useful. We agree with the reviewer and have pointed out the need to study certain aspects of T-cell biology in vivo, comparing mechanisms that can be studied in vitro with those that require in vivo experimentation. We have added to the legend of Fig. 2 a comment about the need for more mechanistic information. : This diagram illustrates the number and complexity of the steps required to achieve efficacy. Many of these steps can be studied in vitro; for others (e.g., extravasation), in vitro models are inherently problematic. Highlight other potential approaches to increasing tuor specificity e.g. synthetic biology from Wendel Lim or masking would be useful rather than just outlining one approach. We have mentioned the SynNotch approach of Dr. Lim and colleagues, (Williams et al., 2020). We have also referenced ligand-binding domain masking approaches and added one reference (Desnoyers et al., 2013): Other approaches are under study, including transcriptional logic circuits and receptor masking (Roybal et al., 2016; Desnoyers et al., 2013). These attempts to widen the target source for selective cancer targets to other targets, including neoantigens and LOH,… Quotes that cell therapy doses can exceed 100 billion cells - seems like an outlier and more normal ranges should be quoted both for CARs and TCRs, especially in light of the table comparing CARs and TCRs and in fact cell dose could be added to this table. We have added a range of T-cell therapeutic doses and altered the sentence: …range from 100 million to 100 billion cells—well beyond the number… It was unclear to me if the table is meant to state dogma or the belief of authors. Some of the assumptions already have data challenging them and discussing some of this as a ‘start of the journey’ may be valuable, for example that CARs show good combination with PD-1 Abs in preclinical models. We have changed the title of the legend and added a clause that states: Potential differences among cell therapy targets, receptors, and regulation not yet rigorously tested by mechanistic data. Experiments to test many of these assumptions are underway. At the end, the authors make strong statements that better assays are needed, which I can’t argue with. Perhaps compare and contrast some assays and say what areas merit more development would be good. Potentials for solutions would make this review more valuable and might stimulate some of the general advances in the field that the review calls for. The review deals with an important topic in T cell therapy - the use of appropriate methodology to increase mechanistic understanding and hopefully, eventually, translatability to the clinic. We make general statements about the kind of assays, but have added text indicating that sensitivity in particular is a useful parameter to measure because it provides a connection among different targets and receptors: In particular, we believe that quantitative assays that measure absolute sensitivity of receptors should be more widely employed, allowing direct comparisons among different targets and receptors."
}
]
},
{
"id": "74926",
"date": "02 Dec 2020",
"name": "C. Glenn Begley",
"expertise": [
"Reviewer Expertise Translational research - oncology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis review highlights the success of the recent advances in immune-oncology focusing particularly on cellular therapies, and outlines some of the fundamental scientific criteria that have been ‘by-passed’ in moving into the clinic, but that will likely need to be understood to make the cell-therapy approach applicable to solid tumors.\nI note the valuable comments of James et al., and in addition suggest:\nAs part of the “revolution in immune-oncology” seen with checkpoint inhibitors, CAR-T cells, the authors should acknowledge another ‘recombinant cellular therapy’ – oncolytic viruses.\n\nThe authors are appropriately critical of mouse models that “trade off tractability with relevance“ and are then “often regarded as decisive in selection of clinical candidates because of presumptive experimental supremacy. In our view this is specious.” I agree completely! However given the ubiquity of these models regardless of therapeutic modality, it could be helpful to provide additional commentary as to how these models should be appropriately exploited.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": [
{
"c_id": "6199",
"date": "22 Dec 2020",
"name": "Alexander Kamb",
"role": "Author Response",
"response": "As part of the “revolution in immune-oncology” seen with checkpoint inhibitors, CAR-T cells, the authors should acknowledge another ‘recombinant cellular therapy’ – oncolytic viruses. Our focus is on T-cell therapies, and not intended as an inclusive review. We have added “An opinion” to the title to clarify this. We acknowledge the interest of oncolytic viruses but do not see an unobtrusive way to feather them into our opinion piece. The authors are appropriately critical of mouse models that “trade off tractability with relevance“ and are then “often regarded as decisive in selection of clinical candidates because of presumptive experimental supremacy. In our view this is specious.” I agree completely! However given the ubiquity of these models regardless of therapeutic modality, it could be helpful to provide additional commentary as to how these models should be appropriately exploited. We appreciate this comment, and now include a statement about xenograft models as an example: As a T-cell therapy example, simple xenograft models demonstrate that therapeutic function is compatible with the environment of a mammalian body; nothing more, but nothing less."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1295
|
https://f1000research.com/articles/9-1497/v1
|
22 Dec 20
|
{
"type": "Research Article",
"title": "Glibenclamide, ATP and metformin increases the expression of human bile salt export pump ABCB11",
"authors": [
"Nisha Vats",
"Ravi Chandra Dubey",
"Madhusudana Girija Sanal",
"Pankaj Taneja",
"Senthil Kumar Venugopal",
"Nisha Vats",
"Ravi Chandra Dubey",
"Pankaj Taneja",
"Senthil Kumar Venugopal"
],
"abstract": "Background: Bile salt export pump (BSEP/ABCB11) is important in the maintenance of the enterohepatic circulation of bile acids and drugs. Drugs such as rifampicin and glibenclamide inhibit BSEP. Progressive familial intrahepatic cholestasis type-2, a lethal pediatric disease, some forms of intrahepatic cholestasis of pregnancy, and drug-induced cholestasis are associated with BSEP dysfunction. Methods: We started with a bioinformatic approach to identify the relationship between ABCB11 and other proteins, microRNAs, and drugs. A microarray data set of the liver samples from ABCB11 knockout mice was analyzed using GEO2R. Differentially expressed gene pathway enrichment analysis was conducted using ClueGo. A protein-protein interaction network was constructed using STRING application in Cytoscape. Networks were analyzed using Cytoscape. CyTargetLinker was used to screen the transcription factors, microRNAs and drugs. Predicted drugs were validated on human liver cell line, HepG2. BSEP expression was quantified by real-time PCR and western blotting. Results: ABCB11 knockout in mice was associated with a predominant upregulation and downregulation of genes associated with cellular component movement and sterol metabolism, respectively. We further identified the hub genes in the network. Genes related to immune activity, cell signaling, and fatty acid metabolism were dysregulated. We further identified drugs (glibenclamide and ATP) and a total of 14 microRNAs targeting the gene. Western blot and real-time PCR analysis confirmed the upregulation of BSEP on the treatment of HepG2 cells with glibenclamide, ATP, and metformin. Conclusions: The differential expression of cell signaling genes and those related to immune activity in ABCB11 KO animals may be secondary to cell injury. We have found glibenclamide, ATP, and metformin upregulates BSEP. The mechanisms involved and the clinical relevance of these findings need to be investigated.",
"keywords": [
"BSEP/ABCB11",
"ABCB11-KO",
"Insilico",
"upregulation",
"HepG2",
"glibenclamide",
"ATP",
"and metformin",
"Nuclear Receptors"
],
"content": "Introduction\n\nThe bile salt export pump (BSEP), the major bile salt transporter in the liver canalicular membrane, is coded by ABCB11 gene, and mutations in this gene cause progressive familial intrahepatic cholestasis type- 2 (PFIC-2)1,2. Besides PFIC-2, mutations or insufficiency of BSEP is associated with a variety of diseases such as drug-induced cholestasis, pregnancy induced cholestasis, cryptogenic cholestasis, cholangiocarcinoma and hepatocellular carcinoma, which are cancers of the liver3–7. Naturally, ABCB11 expression is induced by bile salts and is mediated by FXR- RXR heterodimer8. Here in this pilot study we explored in silico the interactions/networks around ABCB11. We wanted to identify the genes, drugs, microRNAs which might influence the expression of ABCB11. Drugs which could upregulate ABCB11 expression may be useful in ABCB11 haploinsufficiency and inhibition of the pump could result in the accumulation of toxic bile salts inside hepatocytes. Modulation of ABCB11 expression could be clinically beneficial in a variety of medical conditions.\n\n\nMethods\n\nWe analyzed the microarray data set of the liver samples from ABCB11 knockout mice (GEO accession GSE70179) using GEO2R online tool from NCBI9. All differentially expressed genes (DEGs) were filtered with two criteria: -1> log2FC >+1 and adj. p-value <0.05.\n\nTo identify DEGs which are significant, pathway enrichment analysis was conducted using the ClueGo v2.5.5 app from Cytoscape10. ClueGo constructed and compared networks of functionally related GO terms with kappa statistics, which was adjusted at >0.4 in this study.\n\nThe protein-protein interaction (PPI) networks were built by the Search Tool for the Retrieval of Interacting Genes (STRING v11.0)11 and Cytoscape v3.7.1 software. The Molecular Complex Detection (MCODE v1.6), app from Cytoscape was used to screen modules of the PPI network with degree cut-off = 2, node score cut-off = 0.2, k-core = 2, and maximum depth = 100. The hub genes were identified by the CytoHubba v0.1 app. The top 10 nodes were considered as notable hub genes and displayed.\n\nCyTargetLinker v4.1.0 from Cytoscape was used to identify the transcription factors (TFs) and microRNAs using ENCODE and Target-scan databases, respectively. We drew Homo sapiens TF-target interactions linkset from database (ENCODE)12 and drug-target interactions linkset from the database (DrugBank)13. The networks were visualized and analyzed using Cytoscape v3.7.1 Cytoscape app CyTargetlinker version 4.1.0[6] was used to screen the transcription factors and microRNAs\n\nHepG2 cells were grown in high-glucose DMEM (Hi-Media Lab, Mumbai, Cat. # AL111-500ML) supplemented with 10% fetal bovine serum (CellClone, Genetix Biotech Asia, New Delhi, Cat.# CCS-500-SA-U), penicillin and streptomycin (Hi-Media, Mumbai Cat. # A018-5X100ML). When cells became 80% confluent, they were individually treated with glibenclamide (500 ng/mL)14, metformin (25 mg/L)15 or ATP (1 mM) for 48 h. After 48 h cells were scraped out for total protein and RNA.\n\nTotal proteins from HepG2 cells were prepared and run on 10% SDS-PAGE and transferred to a PVDF membrane using a transfer apparatus following the standard protocols (Bio-Rad). After incubation with 5% nonfat milk in TBST (10 mM Tris, pH 8.0, 150 mM NaCl, 0.5% Tween 20) for 1 h the membrane was washed once with TBST and incubated overnight at 4°C with rabbit antibodies against human ABCB11 (Affinity, Catalog #DF 9278) 1: 2000 dilution; mouse anti-human β-actin (Santa Cruz Cat.# SC4778), dilution 1:1000. The membrane was washed three times (TBST) and incubated with a 1:5000 dilution of horseradish peroxidase-conjugated anti-rabbit (Santa Cruz Cat# SC-2004)/anti-mouse antibodies (Cat.#SC-2005) for 2 h. Blots were washed with TBST four times and developed with the ECL system (Bio-Rad, US Cat.#170-5060) according to the manufacturer's protocol. The western blot images were acquired using iBright CL1000 (Invitrogen, Thermo Fisher Scientific).\n\nTotal RNA was isolated using NucleoZOL (Takara Cat. No. 740404.200) following manufacturer's instruction. cDNA was prepared from (deoxyribonuclease treated) total RNA using RevertAid Reverse Transcriptase (Thermo Cat. No. EP0441) following the manufacturer's instructions. Real Time PCR was done with unique oligonucleotide primers targeting ABCB11 and GAPDH, Ta=60°C, in triplicates and two repeats, using GoTaq® qPCR Master Mix (Promega Cat. No. A6001) following 'manufacturer's instructions on a Veriti Thermo Cycler from Applied Biosystems Waltham, Massachusetts, USA and data was acquired using the software associated with the same machine (ViiA7 V1.2) and relative quantification was calculated using the by 2(–ΔΔCt) method. Oligonucleotide primer sequences are listed in Table 3.\n\nAn earlier version of this article can be found on biorxiv.org (DOI: https://doi.org/10.1101/2020.09.01.277434).\n\n\nResults\n\nGene expression profile ABCB11 knockdown dataset GSE 70179 from GEO datasets were analysed with GEO2R tool. Genes with >2-fold change in expression value and <0.05 adjusted p-value was filtered. Identified differentially expressed genes (DEG) from the GSE dataset were classified in two groups - upregulated (375 genes) and downregulated (185 genes) (Extended data, Supp.Table-1)36. Gene ontology analysis was performed for functional analysis of DEGs by using ClueGo app from Cytoscape. PPIs of DEGs were constructed using STRING database showed an upregulation of genes related to cellular transport (pink colored nodes), and these nodes were also shared by Toll-like receptor (TLR) signalling (Figure 1). Downregulated genes were involved in metabolic pathways (sterol, carbohydrate, alcohol, etc.) (Extended data, Supp. Table-2)36. We next identified top hub genes in PPI network using CytoHubba app from Cytoscape (Table 1). Immunologically important genes were among the top ranked upregulated hub genes (Figure 2a) downregulated group majorly represents cell signaling and fatty acid metabolism (Figure 2b). Epidermal growth factor receptor (EGFR) ranked first among the genes involved in signaling pathways. Kinases play a role in the transcription, activity, or intracellular localization of ABC transporters as do protein interactions16. Proteins interacting with ABCB11 are represented in Figure 3 which includes nuclear receptors NR1H4 and NR0B2. Most proteins were associated with bile acid metabolism and transport.\n\nGene ontology analysis was performed for functional analysis of DEGs by using ClueGo app from Cytoscape. This app allows simultaneous analysis of multiple annotation and ontology sources. Functionally grouped network is represented Figure 1a (upregulated genes) 1b (downregulated genes) . The node size represents enrichment significance and connections are based on kappa score (> 0.4). In upregulated gene group maximum number of nodes which are in pink color represent the cellular component movement. These nodes are shared by toll like receptor signaling pathway.\n\nWe observed that the top ranked hub genes in PPI network which were upregulated were associated with immune activity while those downregulated are associated with cell signaling and fatty acid metabolism. EGFR came first in the ranking which is a critical receptor in several cell signaling pathways.\n\nWe observed that the top ranked hub genes in the upregulated group were mainly related to immune activity (a). The top hub genes in downregulated group were associated with cell signaling and fatty acid metabolism. EGFR emerged as the top hub gene, a growth factor receptor which is crucial factor several cell signaling pathways (b).\n\nThis network was constructed to analyze the relationship between ABCB11 and other proteins. Cytohubba app was used to calculate centrality of each node by MCC method. Node colour (red to yellow) represents the significance of the centrality in the group. In this analysis, we counted 11 nodes and 42 edges. These proteins majorly involved in bile acid metabolism and transport. Most of these genes are participant of more than one pathway which was expected because these pathways intersect and coregulated. We also mapped the NR0B2 protein, which is participate in sterol metabolism.\n\nAs described, sub-network analysis was performed using MCODE (Figure 4), and CMPK2, ACTG1, and SSTR2 emerged as seed nodes among upregulated genes (Table 2). Among downregulated gene groups, only one subnetwork was found to be significant which had three genes: MIA3 (which codes a protein which is important in the transport of cargos that are too large to fit into COPII-coated vesicles such as collagen VII), IGFBP4 (encoding a protein that binds to both insulin-like growth factors and modifies their functions) and NOTUM (encoding a carboxylesterase that acts as a key negative regulator of the Wnt signaling pathway by specifically mediating depalmitoleoylation of WNT proteins).\n\nTop sub-networks on the basis of MCODE score (Degree cut-off= 2, node score cut-off = 0.2, k-core = 2 and max. depth = 100). Upregulated gene group clusters, we identified seed nodes (CMPK2, ACTG1 and SSTR2) in the network (green and blue). In downregulated gene group, we identified only one subnetwork which qualified cut off criteria. Three genes in this sub-network was identified: MIA3, IGFBP4 and NOTUM (red).\n\nSub-network analysis was performed using the Molecular Complex Detection (MCODE) app from Cytoscape and CMPK2, ACTG1 and SSTR2 emerged as seed nodes among upregulated.\n\nUsing CyTargetLinker identified two drugs, glibenclamide, and ATP, directly targeting ABCB11. We subsequently looked for microRNAs [Target-scan database]17 that were associated with ABCB11, and a total of 14 microRNAs were identified targeting the gene (Figure 5). Transcription factors and microRNAs targeting ABCB11 and interacting partners are represented in Figure 6.\n\nWe identified microRNAs that were associated with ABCB11. In total 14 microRNA identified targeting the gene.\n\nIn the screen of transcription factors of ABCB11 interaction network we observed 21 nodes and 52 edges. Among these transcription factors, FOXA have been suggested an important factor in bile duct development and lipid accumulation. HNF4A in the regulation dyslipidaemia and terminal liver failure and JUND in fibrosis development. Others can be investigated in future studies. We counted 55 nodes and 89 edges in the search of microRNA targeting the ABCB11 network. Four genes (ABCB11, ATP8B1, SLC10A2 and NR1H4) targeted by multiple microRNAs also some microRNA such as has-miR-203a-3p.2 and has-miR-203a-3p.2 target more than one gene. By nature, a microRNA can regulate several pathways therefore it would be interesting to study in future the dysregulation of these microRNAs and interaction with Identified transcription factors.\n\nWe evaluated in vitro, the effect of three drugs, two of which were bioinformatically predicted (Glibenclamide, ATP) and one based on literature18. We found all the three compounds upregulating ABCB11 expression based on qPCR, and this was confirmed by western blot (Figure 7). Unannotated western blot images and raw qPCR Ct values are available as Underlying data36.\n\nAll the three compounds upregulated ABCB11 expression based on Real Time PCR data. This was further confirmed by Western Blot against anti-human ABCB11 antibody as the primary antibody. The upper image in 7b shows the PVDF membrane probed with antibody against ABCB11 and the lower image shows the same membrane stripped and probed with antibody against beta-actin (loading control).\n\n\nDiscussion\n\nWe identified several immunologically important genes being upregulated during ABCB11 deficiency. The reason could be liver cell injury secondary to bile salt accumulation, which triggers the sterile immune response19,20 and the downregulation of transport proteins and metabolically important genes could be because of decreased liver function following damage. A regenerative response follows cell injury, and a host of genes involved in regeneration are upregulated21–23; however, it appears that bile salts in the absence of BSEP hamper the regenerative response reflected by dysregulated collagen transporting protein MIA3 and NOTUM a protein involved in Wnt signaling. It's also possible that EGFR is dysregulated via accumulating bile salts mediated by STAT324. We have observed an upregulation of ABCB11 in a liver cell line (HepG2) on treatment with glibenclamide, metformin, and ATP. This expression is upregulation may be a compensatory mechanism in the case of glibenclamide and metformin because these drugs are known to inhibit ABCB1125. Metformin is known to interfere with ABCB11 function, mediated through AMPK-FXR crosstalk18 involving metformin induced FXR phosphorylation. ATP acts through ATP receptors on hepatocytes26,27. ATP is known to cross the plasma membrane28 and this can act via AMPK. However, ATP has a very short half-life29, and it may be converted to ADP, which can activate AMPK30. In a recent report, metformin was shown to suppress ABCB11 expression, which is not in agreement with our observation, however, they performed their experiment on primary human hepatocytes, and they have also treated their cells with dimethylsulfoxide (DMSO)31.\n\nThere are many reports stating the influence of DMSO on human gene expression. For example, Verheijen et al. “exposed 3D cardiac and hepatic microtissues to medium with or without 0.1% DMSO and analyzed the transcriptome, proteome and DNA methylation profiles”. They found that “in both tissue types, transcriptome analysis detected >2000 differentially expressed genes affecting similar biological processes, thereby indicating consistent cross-organ actions of DMSO”. In both tissue types, the transcriptome analysis detected over 2000 differentially expressed genes affecting similar biological processes32. Moskot et al. reported alterations of lysosomal ultrastructure upon DMSO treatment33. Alizadeh et al. reported that DMSO catalyzes hepatic differentiation of adipose tissue-derived mesenchymal stem cells34. It has been observed that “culturing pluripotent stem cells in DMSO activates the retinoblastoma protein, increases the proportion of cells in the early G1 phase of the cell cycle, and subsequently improves their competency for directed differentiation into multiple lineages in more than 25 stem cell lines”35. However, we are not sure whether the observed difference is attributed to DMSO.\n\nIn conclusion, we need more experiments to determine the mechanisms of action of these drugs on the upregulation of ABCB11. Many changes in gene expression following ABCB11 knockout could be secondary to stress, immune and regenerative responses following hepatocyte injury in mice liver.\n\n\nData availability\n\nHarvard Dataverse: Real Time PCR for ABCB11 and few NRs. https://doi.org/10.7910/DVN/AOYKY736.\n\nThis project contains the following underlying data:\n\n2020-09-12 092712-ViiA7-export.xls. (qPCR data following addition of ATP, metformin or gilbenclamide.)\n\nABCB11 WESTERN BLOT DRUG.tif. (Unannotated western blot image for ABC11.)\n\nABCB11_WB_Repeat_Drugs.tif. (Unannotated repeat western blot image for ABC11.)\n\nactin drug.tif. (Unannotated western blot image for β-actin.)\n\nActin_2020_07_11_182456.jpg. (Unannotated western blot image, including β-actin loading control.)\n\nActin_2020_07_11_182456.tif. (As above, but in tif format.)\n\nnisha_qPCR DATA_ 7142020.xls. (qPCR data for ABC11 and other indicated genes.)\n\nrealtime and western blottt (1).pptx. (Western blot and qPCR data pooled into a single file.)\n\nRepeat_Actin_drug_WB.tif. (Unannotated repeat western blot image for β-actin.)\n\nHarvard Dataverse: Real Time PCR for ABCB11 and few NRs. https://doi.org/10.7910/DVN/AOYKY736.\n\nThis project contains the following extended data:\n\nSupp-Table-1-Dysregulated genes GSE70179, GEO2R, NCBI. (Differentially expressed genes\n\nSupp-Table-2-Gene ontology analysis, DAVID. (Gene ontology analysis of\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).",
"appendix": "References\n\nStrautnieks SS, Byrne JA, Pawlikowska L, et al.: Severe bile salt export pump deficiency: 82 different ABCB11 mutations in 109 families. Gastroenterology. 2008; 134(4): 1203–1214. PubMed Abstract | Publisher Full Text\n\nChilds S, Yeh RL, Georges E, et al.: Identification of a sister gene to P-glycoprotein. Cancer Res. 1995; 55(10): 2029–2034. PubMed Abstract\n\nDavit-Spraul A, Gonzales E, Jacquemin E: NR1H4 analysis in patients with progressive familial intrahepatic cholestasis, drug-induced cholestasis or intrahepatic cholestasis of pregnancy unrelated to ATP8B1, ABCB11 and ABCB4 mutations. Clin Res Hepatol Gastroenterol. 2012; 36(6): 569–573. PubMed Abstract | Publisher Full Text\n\nDixon PH, van Mil SWC, Chambers J, et al.: Contribution of variant alleles of ABCB11 to susceptibility to intrahepatic cholestasis of pregnancy. Gut. 2009; 58(4): 537–544. PubMed Abstract | Publisher Full Text\n\nScheimann AO, Strautnieks SS, Knisely AS, et al.: Mutations in bile salt export pump (ABCB11) in two children with progressive familial intrahepatic cholestasis and cholangiocarcinoma. J Pediatr. 2007; 150(5): 556–559. PubMed Abstract | Publisher Full Text\n\nKnisely AS, Strautnieks SS, Meier Y, et al.: Hepatocellular carcinoma in ten children under five years of age with bile salt export pump deficiency. Hepatology. 2006; 44(2): 478–486. PubMed Abstract | Publisher Full Text\n\nVitale G, Gitto S, Raimondi F, et al.: Cryptogenic cholestasis in young and adults: ATP8B1, ABCB11, ABCB4, and TJP2 gene variants analysis by high-throughput sequencing. J Gastroenterol. 2018; 53(8): 945–958. PubMed Abstract | Publisher Full Text\n\nPlass JRM, Mol O, Heegsma J, et al.: Farnesoid X receptor and bile salts are involved in transcriptional regulation of the gene encoding the human bile salt export pump. Hepatology. 2002; 35(3): 589–596. PubMed Abstract | Publisher Full Text\n\nGEO2R - GEO - NCBI [Internet]. Reference Source\n\nShannon P, Markiel A, Ozier O, et al.: Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003; 13(11): 2498–2504. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSzklarczyk D, Gable AL, Lyon D, et al.: STRING v11: protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res. 2019; 47(D1): D607–D613. PubMed Abstract | Publisher Full Text | Free Full Text\n\nENCODE Project Consortium: The ENCODE (encyclopedia of DNA elements) project. Science. 2004; 306(5696): 636–640. PubMed Abstract | Publisher Full Text\n\nWishart DS, Feunang YD, Guo AC, et al.: DrugBank 5.0: a major update to the DrugBank database for 2018. Nucleic Acids Res. 2018; 46(D1): D1074–D1082. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMatsuda A, Kuzuya T, Sugita Y: Plasma levels of glibenclamide in diabetic patients during its routine clinical administration determined by a specific radioimmunoassay. … and metabolic research. 1983.\n\nKajbaf F, De Broe ME, Lalau JD: Therapeutic concentrations of metformin: A systematic review. Clin Pharmacokinet. 2016; 55(4): 439–459. PubMed Abstract | Publisher Full Text\n\nCrawford RR, Potukuchi PK, Schuetz EG, et al.: Beyond Competitive Inhibition: Regulation of ABC Transporters by Kinases and Protein-Protein Interactions as Potential Mechanisms of Drug-Drug Interactions. Drug Metab Dispos. 2018; 46(5): 567–580. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAgarwal V, Bell GW, Nam JW, et al.: Predicting effective microRNA target sites in mammalian mRNAs. Elife. 2015; 4: e05005. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLien F, Berthier A, Bouchaert E, et al.: Metformin interferes with bile acid homeostasis through AMPK-FXR crosstalk. J Clin Invest. 2014; 124(3): 1037–1051. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAllen K, Jaeschke H, Copple BL: Bile acids induce inflammatory genes in hepatocytes: a novel mechanism of inflammation during obstructive cholestasis. Am J Pathol. 2011; 178(1): 175–186. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWoolbright BL, Jaeschke H: The impact of sterile inflammation in acute liver injury. J Clin Transl Res. 2017; 3(Suppl 1): 170–188. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDransfeld O, Gehrmann T, Köhrer K, et al.: Oligonucleotide microarray analysis of differential transporter regulation in the regenerating rat liver. Liver Int. 2005; 25(6): 1243–1258. PubMed Abstract | Publisher Full Text\n\nWhite P, Brestelli JE, Kaestner KH, et al.: Identification of transcriptional networks during liver regeneration. J Biol Chem. 2005; 280(5): 3715–3722. PubMed Abstract | Publisher Full Text\n\nLi SC, Wang FS, Yang Y, et al.: Microarray Study of Pathway Analysis Expression Profile Associated with MicroRNA-29a with Regard to Murine Cholestatic Liver Injuries. Int J Mol Sci. 2016; 17(3): 324. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMair M, Zollner G, Schneller D, et al.: Signal transducer and activator of transcription 3 protects from liver injury and fibrosis in a mouse model of sclerosing cholangitis. Gastroenterology. 2010: 138(7): 2499–508. PubMed Abstract | Publisher Full Text\n\nKrivoy N, Zaher A, Yaacov B, et al.: Fatal toxic intrahepatic cholestasis secondary to glibenclamide. Diabetes care. 1996; 19(4): 385–6. PubMed Abstract | Publisher Full Text\n\nSchlosser SF, Burgstahler AD, Nathanson MH: Isolated rat hepatocytes can signal to other hepatocytes and bile duct cells by release of nucleotides. Proc Natl Acad Sci U S A. 1996; 93(18): 9948–9953. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCharest R, Blackmore PF, Exton JH: Characterization of responses of isolated rat hepatocytes to ATP and ADP. J Biol Chem. 1985; 260(29): 15789–15794. PubMed Abstract\n\nChaudry IH: Does ATP cross the cell plasma membrane. Yale J Biol Med. 1982; 55(1): 1–10. PubMed Abstract | Free Full Text\n\nSkrabanja ATP, Bouman EAC, Dagnelie PC: Potential value of adenosine 5’ -triphosphate (ATP) and adenosine in anaesthesia and intensive care medicine. Br J Anaesth. 2005; 94(5): 556–562. PubMed Abstract | Publisher Full Text\n\nHardie DG: AMP-activated protein kinase: an energy sensor that regulates all aspects of cell function. Genes Dev. 2011; 25(18): 1895–1908. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGarzel B, Hu T, Li L, et al.: Metformin disrupts bile acid efflux by repressing bile salt export pump expression. Pharm Res. 2020; 37(2): 26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVerheijen M, Lienhard M, Schrooders Y, et al.: DMSO induces drastic changes in human cellular processes and epigenetic landscape in vitro. Sci Rep. 2019; 9(1): 4641. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoskot M, Jakóbkiewicz-Banecka J, Kloska A, et al.: The role of dimethyl sulfoxide (DMSO) in gene expression modulation and glycosaminoglycan metabolism in lysosomal storage disorders on an example of mucopolysaccharidosis. Int J Mol Sci. 2019; 20(2): 304. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlizadeh E, Zarghami N, Eslaminejad MB, et al.: The effect of dimethyl sulfoxide on hepatic differentiation of mesenchymal stem cells. Artif Cells Nanomed Biotechnol. 2016; 44(1): 157–164. PubMed Abstract | Publisher Full Text\n\nChetty S, Pagliuca FW, Honore C, et al.: A simple tool to improve pluripotent stem cell differentiation. Nat Methods. 2013; 10(6): 553–556. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMadhusudana Girija S: Real Time PCR for ABCB11 and few NRs. Harvard Dataverse, V6, 2020. http://www.doi.org/10.7910/DVN/AOYKY7"
}
|
[
{
"id": "76410",
"date": "25 Jan 2021",
"name": "Premkumar Kumpati",
"expertise": [
"Reviewer Expertise Genomics",
"Genetics",
"Nanotheraputics",
"Cancer biology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors have provided an interesting manuscript that described the role of Bile salt export pump (BSEP/ABCB11) and its relationship with other proteins, microRNAs and Drugs. Various bioinformatics tools were used to explore and carried out experiments to validate and correlate their findings.\nOverall, I believe that this manuscript is well written and will provide some useful information for the scientific community. Having said that there are some concerns I have that the authors should address before it is ready/accept for indexing.\nMinor Comments to the Author\nThe author can provide more information on biological mechanism of ABCB11.\n\nWhat is the rational for selecting concentration of drugs used glibenclamide (500 ng/mL), metformin (25 mg/L) or ATP (1 mM).\n\nWhat is the basis for selecting Glibenclamide & ATP out of the total hits identified using bioinformatic analysis.\n\nProviding Uniport accession number of each gene in table 1 could be more informative, authors may consider doing it.\n\nImage quality for Fig.1 & 4 may be increased.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "76680",
"date": "03 Feb 2021",
"name": "Jayanta Roy-Chowdhury",
"expertise": [
"Reviewer Expertise Hepatology",
"Genetic disorders",
"Cell transplantation",
"Gene therapy"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this manuscript the authors have used the data on hepatic gene expression submitted to NCBI by Zhang Y, Neale G and Schuetz to determine the effect on the haplodeficiency of ABCB11 (BSEP) in mouse liver. The 375 upregulated genes included those involved in cellular transport and innate immunity (e.g. IFN signaling), whereas the 185 downregulated genes included signal transduction proteins, e.g. epidermal growth factor receptor and those involved with metabolic pathways. In addition, they have examined the effect of Glibenclamide, metformin and ATP on the expression of human ABCB11 in the human cell line HepG2. All three chemicals were found to increase ABCB11 expression.\nComments:\nThe information gene pathway analysis provides useful information as it has revealed \"nodes\" and \"hubs\" that mediate the up regulation and down regulation of the genes the expression of which is dysregulated in deficiency of ABCB11 function. However, as the original data set had been derived from whole mouse liver, it comprises gene expression by hepatocytes as well as non-parenchymal cells. It is known that bile acid accumulation can affect gene expression in various cell types in addition to hepatocytes (such as lung cells). Perhaps, the authors can mention this in the Discussion section.\nA second concern is that HepG2 cells are not the most appropriate cell line for modeling the induction of ABCB11 in human hepatocytes, because, unlike some other human hepatoma cell lines, these cells lack the most abundant hepatocyte microRNA, miR-122. miR-122 is known to target various genes, thereby enhancing IFN signaling. On the other hand IFN downregulates the expression of miR-122. Thus, transcriptional induction of ABCB11 in HepG2 cell may not parallel that in primary human hepatocytes. Therefore, the drug induction study should be validated in a different human hepatoma cell line, or at least, the authors should discuss this complexity in interpreting the results.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "79055",
"date": "09 Mar 2021",
"name": "Amit Dash",
"expertise": [
"Reviewer Expertise Cancer biology",
"transcription",
"nuclear receptor",
"cell biology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn the article by Vats et al., the authors used the preexisting microarray data set of the liver samples from ABCB11 knockout mice and used different bioinformatics methods to draw a relationship between ABCB11 and other proteins, microRNAs, and drugs. They validated the predicted drugs by real-time PCR and Western blot in HepG2 cells. The authors did an outstanding job in this pilot study which may be helpful for future researchers. Though their approach and methodologies are right but for better reproducibility and understanding, authors need to explain the following points\nMajor comments:\nWhy authors choose HepG2 cells and not other hepatoma cell lines? They need to describe it in the discussion.\n\nAuthors need to give justification about their chosen drug concentration. Is it the optimum concentration?\n\nFor more reproducibility of the data, it will be nice if authors mention the detailed protein extraction method and whether they used whole-cell extract or any fractions like nuclear and cytoplasmic extract.\n\nIn figure 7a, the authors need to mention what is on the Y-axis of the graph and need to put error bars and statistical significance. The method of statistical analysis needs to discuss in materials and methods. Also, it will be nice if authors make a graph for the western blot taking the averages and errors of three independent experiments and normalized with actin.\n\nFrom their raw data, it is surprising that the endogenous control (GAPDH or 18S) is highly variable between controls and treatments. Did the authors start with an equal amount of RNA? How much RNA they used for cDNA. Also, it is confusing out of 18S and GAPDH, which they used for endogenous control. If they used 18S, they need to mention the primer sequences.\n\nAuthors need to recheck the GAPDH primers as the melting point is more than one.\nMinor comments:\nUnder introduction, on line “we explored in silico the interactions/networks around,” \"the\" can be omitted.\n\nFig1 resolution needs to be increased.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1497
|
https://f1000research.com/articles/9-1088/v1
|
03 Sep 20
|
{
"type": "Brief Report",
"title": "Cyclic stroke mortality variations follow sunspot patterns",
"authors": [
"Stella Geronikolou",
"Alexandros Leontitsis",
"Vasilis Petropoulos",
"Constantinos Davos",
"Dennis Cokkinos",
"George Chrousos",
"Alexandros Leontitsis",
"Vasilis Petropoulos",
"Constantinos Davos",
"Dennis Cokkinos",
"George Chrousos"
],
"abstract": "Mapping time-structures is a burgeoning scientific field enriching the (P4) medicine models. Local evidence in Mediterranean populations is underinvestigated. The Censed stroke-related death events (D) in the largest East-Mediterranean port (Piraeus), during (1985-1989), when local population had diet and genetic homogeneity-been interrupted by the immigration into Greece in 1990s, and Sunspot numbers indexed by Wolf numbers (Rz) (1944-2004), were evaluated using Fast Fourier Analysis and Singular Spectrum Analysis in MATLAB. D were turned with fluctuations >35% in Rz. A non-anthropogenic 6.8 days cycle was recognized. This study may be taken into consideration in future public health planning and chronotherapy evaluations.",
"keywords": [
"Sunspot numbers",
"Chronome",
"Stroke mortality",
"Singular Spectrum Approach",
"NCOR1",
"R1 interactome"
],
"content": "Introduction\n\nStrokes are the second leading cause of death and disability worldwide (Mozaffarian et al., 2015). Strokes share all of the recommended interventions for chronic noncommunicable disease (NCD): life-style modifications, including a low fat/salt/sugar diet, moderate physical activity, discontinuation of smoking, sufficient sleep, and control of arterial blood pressure control, and, if necessary, pharmacologic therapy (Mozaffarian et al., 2015). In addition, risk factors such as chronic stress, underlying diseases, such as obesity, diabetes mellitus, chronic obstructive pulmonary disease, and renal insufficiency, as well as predisposing genetic factors, have been implicated in stroke morbidity and mortality (Malik & Dichgans, 2018). The incidence and prevalence of stroke subtypes vary greatly, depending on ethnicity and country income. As stroke statistics fail to cover all etiologies, some remain unknown.\n\nStroke has been previously associated to solar activity (Halgberg et al., 2001; Otsuka et al., 2001; Stoupel et al., 1995; Stoupel et al., 1996). Such activity, as indexed by sunspot numbers, has been generally associated with health (Feigin et al., 2014; Halberg et al., 1998; Petropoulos & Geronikolou, 2005). However, relevant chaos and trend analyses in local populations have been limited but strongly suggested (Reinberg et al., 2017).\n\nOur aim was to investigate the dynamics and trends in the time series, to determine sunspot numbers vs. daily and monthly stroke deaths in synchronized periodicities with gradual time delays (chronomes), and to define a sunspot number threshold for presence of stroke mortality.\n\n\nMethods\n\nIn this study we focused on monthly stroke mortality events between 1985 and 1989, based on the underlined cause of death data from the archives of Piraeus Civil Registry (Geronikolou & Zikos, 1991).\n\nThe sunspot numbers were derived from the archives of measurements published by Solar Geographical Data. Sunspot Number, denoted Rz (Zurich number), is defined as: Rz = K(10g + f), where g is the number of sunspot groups visible on the Sun, f represents the total number of individual spots visible; and K is an instrumental factor to take into account differences between observers and observatories.\n\nThe stroke death rate in Piraeus, was calculated over the formula (number of all deaths per year per 1000 people in June 30th, year x). The overall death rate was calculated with the denominators provided by the 1981 census.\n\nIn the analysis of our short time series, we employed fast Fourier transform (FFT) analysis and the singular spectrum approach (SSA). Thus, we first performed a square root transformation of the sunspot time series. We subsequently analyzed the second time series of the stroke deaths using the SSA, to find the principal components that formulated it using principal component analysis (PCA). We applied Pearson correlation analysis to detect the coefficients of variation between the principal components of the sunspots and the strokes time series. All calculations were performed with MATLAB 7 software.\n\n\nResults\n\nWe focused on monthly stroke deaths, based on all death events archived in the local Civil Registry (Geronikolou & Zikos, 1991). There were 792 stroke deaths out of 4324 total deaths events distributed in the 60 months of the quinquennium (1985–1989) examined. Over 54% were women and over 61% occurred at ages over 69 years. The stroke death rate (stroke deaths in year x/overall deaths in year x × 100) was calculated as 17.668 in 1985, 20.089 in 1986, 19.372 in 1987, 17.647 in 1988, and 15.531 in 1989. The overall death rate (all deaths in year x/local population in year x × 100) was 5.5 in 1985, 4.4 in 1986, 4.9 in 1987, 3.5 in 1988, and 3.7 in 1989.\n\nThe observed time series of both monthly and daily sunspot numbers and monthly and daily stroke death events in Piraeus between 1985 and 1989 are illustrated in Figure 1a. The PCA distinguished two principal components, as shown in Figure 1b: 6.8 and 20 days. The sunspot numbers observed (1944–2004) transformed to squared roots are described in Figure 1c. The singular values of the transformed sunspots time series showed that the noise plateau began at the 3rd ordered singular value (Figure 1d). Thus, monthly sunspot numbers by squared root variation and their violent fluctuation of over 35% was correlated to monthly stroke mortality, establishing a negative correlation between the two time-series (sunspot numbers and deaths of strokes) (Figure 1b, d). FFT showed frequencies of 3.5 and 6.85 days.\n\nData on stroke deaths and sunspots by month are available as Underlying data (Geronikolou & Leontitsis, 2020).\n\n(a) Time series of strokes and sunspots from January 1985 to December 1989. A square root transformation is applied to the time series of the sunspots. (b) Time series of sunspots from December 1984 to November 2004. (c) Pearson correlation coefficient between the principal components of the sunspots and strokes time series. (d) Singular values of the transformed sunspots time series. The noise plateau clearly begins at the third-ordered singular value.\n\n\nDiscussion\n\nMapping time-structures is a rapidly growing scientific field, enriching P4 medicine (predictive, preventative, personalized, participatory medicine) models with chronotherapy aspects (Yan, 2015). Human biological clocks are intensively studied. They represent adaptive body mechanisms necessary to assist with homeostatic changes caused by solar activity disturbances. These mechanisms have not been extensively investigated in Mediterranean populations.\n\nChronic NCDs account for over 70% of early deaths worldwide, while stroke is the second leading cause of death and disability; the latter is associated with high expenses in health services, and constitutes a public health challenge (WHO, 2018). This challenge is progressively increasing, considering the large population migrations that take place on the planet because of ethnic conflicts, economic crises and climate changes. Stroke has been associated with various risk factors, such as lifestyle-related eating habits, tobacco and/or alcohol use, and decreased physical activity, underlying comorbidities, such as obesity, hypertension, dyslipidemia, diabetes mellitus type 2, lung and kidney failures, etc., as well as exposure to environmental pollutants. Genetic propensities also contribute to various manifestations of the chronic noncommunicable diseases. Socioeconomic and geographic disparities have been suspected, while heliomagnetic influences have been proposed as possible etiologic contributors to human pathology.\n\nMortality data meeting validity and credibility criteria are a sine qua non in the study of stroke incidence (Feigin & Hoorn, 2004). Our study focused on stroke mortality in the largest Mediterranean port (Piraeus), ranked as the third most populated city in Greece. Moreover, its population is representative of the urban populations in Greece (Geronikolou, 1991). The quinquennium 1985–1989 was chosen, because, until then, Greece had a rather robust diet (low fat/sugar, proteins and vegetables/fruits daily, pure olive oil almost exclusively) and genetic homogeneity, while environmental pollution was limited. In this period, these major confounding factors were not present: major pollution, nonstandard diet, foreign gene inflow. The data used in this study were original and based on the underlined cause of death (Geronikolou, 1991). Importantly, the quinquennium selected was a period when the local population was of the same origin, while only small differences in the socioeconomically stratified levels were present. The covariates related to diet, hygiene and culture were stable in this period and, thus, they could be safely assumed.\n\nThe selected time period 1985–1989 emerged as an appropriate time to provide good reference observations, credible correlations, and future comparisons, and, hence, high inferential precision. Importantly, this period, although relatively short, included the maximum of the 22nd cycle: July 1989 (maximum 157.6 or smoothed sunspot numbers 158.9), as well as the minimum of the 21st solar cycle (minimum 13.4 or smoothed sunspot numbers 12.3). The 21st solar cycle lasted 10.3 years, beginning in June 1976 and ending in September 1986. The 22nd solar cycle lasted 9.7 years, beginning in September 1986 and ending in May 1996. The sunspot numbers do not affect Earth directly; however, the solar wind emanating from solar activity affects stratospheric ozone layer density, whose ionization promotes health morbidity on inhabitants, including the prevalence of strokes (Feigin et al., 2014; Petropoulos & Geronikolou, 2005).\n\nIt has been suggested that chaos and trends in local evidence are lacking (Halberg et al., 1998), and this study addresses this need. Chaotic dynamics analyses could unravel unknown patterns of stroke epidemiology -whose causes are not fully understood. Our work postulates that there is an inverse relation in two time series, between the timing of sunspot numbers and stroke deaths, a hypothesis posed by previous investigations (Geronikolou & Leontisis, 2005; Geronikolou & Petropoulos, 1996; Stoupel et al., 1999). We showed that an over 35% change in the sunspot numbers, shifted the upwards trend of stroke deaths with a delay of two months.\n\nThe interaction of living organisms and their functions with solar radiation has been previously described (Halberg et al., 1998; Reinberg et al., 2017). This consisted mainly of protein secretion studies, and, less extensively of local population dynamics. The molecular interactions network approach, where the inter-species functional interactome of nuclear steroid receptors (R1) was constructed on orthologues was employed (Geronikolou et al., 2018). R1 has interspecies dimensions and thus has evolutionary and historical value extending from insects to humans, that is, from early life eras till now. Solar activity exposure was certainly omnipresent before life appeared on the planet. Similar cycles existed, such as those detected in our study, although the rotation of our planet around its axis and around the sun were faster than they are today (Reinberg et al., 2017). R1 includes genes and their products involved in circadian rhythms, while its major hub NCOR1 in macrophages blocks the pro-atherogenic functions of peroxisome proliferator-activated receptor gamma (PPARγ) in atherosclerosis (Oppi et al., 2020), greatly implicating stroke pathophysiology. PPARγ has a pleiotropic role in intracerebral hemorrhage (Zhao et al., 2015) and ischemic brain injury (Culman et al., 2007). It is likely that as soon as R1 is disrupted, the atherogenic and/or other pathological processes progress dramatically, with lethal consequences. Here, both FFT and SSA revealed a novel common cycle of 6.8 days in Zurich numbers and stroke deaths. The cycle is smaller than the known anthropogenic circaseptan rhythm (Reinberg et al., 2017). The 17-ketosteroids were also found to have a <7 day cycle (Hamburger et al., 1985), confirming the steroid contribution to the phenomenon seen in R1 and herein.\n\nIt has been previously reported that the geomagnetic disturbance indices Kp and aa take the value of 6.75 (Halberg et al., 1998), but without a clear association to the biota (living organisms -flora/fauna/humans). Our finding, apart from its novelty, provides a new insight in stroke epidemiology: the observed patterns suggest an endogenous natural rhythm of renewing populations.\n\nOur work demonstrates a phase shift resulting from violent fluctuation in sunspots variation (>35%) with a clear correlation to monthly stroke deaths. A phase delay of two months was observed between the physical triggering and the death incidence shift. This should be investigated in the future over different and/or even longer periods of time and in different and/or larger populations. Still, the violent fluctuation of 35% of sunspots appears to be a hazard for mortality and, we assume, morbidity. Thus, future medical practice should probably take account of chronopathology so as to prevent stroke mortality shifts (chronotherapy and chronoprevention plans).\n\n\nConclusions\n\nOur work established clearly that of sunspot numbers and stroke mortality were inversely correlated, and that a violent fluctuation of sunspot numbers over 35% shifted monthly mortality to a phase delay of two months. In addition, a common, novel, non-anthropogenic chronome of 6.8 days in solar activity (sunspot numbers) and stroke mortality was revealed. Time structure patterns evaluated with non-linear methods revealed new information on the stroke epidemic, and, thus, contributed to precision inference and the need of sophisticated public health policy planning.\n\n\nData availability\n\nFigshare: Singular Spectrum Analysis of monthly stroke deaths and mean monthly sunspot numbers. https://doi.org/10.6084/m9.figshare.12644981.v2 (Geronikolou & Leontitsis, 2020).\n\nThis file contains the incidence of stroke deaths by month for 1985–1989 and the incidence of sunspots by month for 1944–2004.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Acknowledgements\n\nWe thank Prof Konstantinos Poulakos, Dr Mireille Harsoula and Prof Nikolaos Vogglis for their critical remarks.\n\n\nReferences\n\nCulman J, Zhao Y, Gohlke P: PPAR-gamma: therapeutic target for ischemic stroke. Trends Pharmacol Sci. 2007; 28(5): 244–249. PubMed Abstract | Publisher Full Text\n\nFeigin V, Hoorn SV: How to study stroke incidence. Lancet. 2004; 363(9425): 1920. PubMed Abstract | Publisher Full Text\n\nFeigin V, Parmar PG, Barker-Collo S, et al.: Geomagnetic Storms Can Trigger StrokeEvidence From 6 Large Population-Based Studies in Europe and Australasia. Stroke. 2014; 45(6): 1639–45. PubMed Abstract | Publisher Full Text\n\nGeronikolou S: Public and Environmental Health in Piraeus 1985-1989. Public Hygene Bsc. University of West Attica, Athens. 1991; 300. Reference Source\n\nGeronikolou S, Leontisis A: Investigating internal relations of sunspot numbers and health events with SSA. 18th Summer School/Pan-Hellenic Conference: Non Linear Science and Complexity, Volos, Greece. 2005. Reference Source\n\nGeronikolou S, Leontitsis A: Singular Spectrum Analysis of monthly stroke deaths and mean monthly sunspot numbers. figshare. Dataset. 2020. http://www.doi.org/10.6084/m9.figshare.12644981.v2\n\nGeronikolou S, Petropoulos V: Solar activity related to geomagnetic climatic parameters and biological effects in Greece. 3rd Soltip Symposium, Beijing, China. 1996. Reference Source\n\nGeronikolou S, Zikos P: Public and Environmental Health in Piraeus 1985-1989. Public Hygene. Bachelor. University of West Attica, Athens, Greece. 1991; 300. Reference Source\n\nGeronikolou SA, Pavlopoulou A, Kanaka-Gantenbein C, et al.: Inter-species functional interactome of nuclear steroid receptors (R1). Front Biosci (Elite Ed). 2018; 10: 208–228. PubMed Abstract | Publisher Full Text\n\nHalberg F, Siutkina EV, Cornelissen G, et al.: Chronomes render predictable the otherwise-neglected human \"physiological range\": position paper of the BIOCOS project. BIOsphere and the COSmos. Fiziol Cheloveka. 1998; 24(4): 14–21. PubMed Abstract\n\nHalgberg F, Cornelissen G, Watanabe Y, et al.: Near 10-year and longer periods modulate circadians: intersecting anti-aging and chronoastrobiological research. J Gerontol A Biol Sci Med Sci. 2001; 56(5): M304–24. PubMed Abstract | Publisher Full Text\n\nHamburger C, Sothern RB, Halberg F: Circaseptans and semicircaseptan aspects of human male sexual activity. Chronobiologia. 12:250. Chronobiologia. 1985; 12: 250.\n\nMalik R, Dichgans M: Challenges and opportunities in stroke genetics. Cardiovasc Res. 2018; 114(9): 1226–1240. PubMed Abstract | Publisher Full Text\n\nMozaffarian D, Benjamin EJ, Go AS, et al. American Heart Association Statistics Committee, Stroke Statistics Subcommittee.: Heart disease and stroke statistics–2015 update: a report from the American Heart Association. Circulation. 2015; 131(4): e29–322.PubMed Abstract | Publisher Full Text\n\nOppi S, Nusser-Stein S, Blyszczuk P, et al.: Macrophage NCOR1 protects from atherosclerosis by repressing a pro-atherogenic PPARγ signature. Eur Heart J. 2020; 41(9): 995–1005. PubMed Abstract | Publisher Full Text\n\nOtsuka K, Oinuma S, Cornelissen G, et al.: Alternating light-darkness-influenced human electrocardiographic magnetoreception in association with geomagnetic pulsations. Biomed Pharmacother. 2001; 55(Suppl 1): 63s–75s. PubMed Abstract | Publisher Full Text\n\nPetropoulos V, Geronikolou S: Stratospheric ozon, density variation solar activity and biological phenomena. 7th HELLASET Conference Hellenic Astronomical Society, Cefalonia Greece, 2005. Reference Source\n\nReinberg AE, Dejardin L, Smolensky MH, et al.: Seven-day human biological rhythms: An expedition in search of their origin, synchronization, functional advantage, adaptive value and clinical relevance. Chronobiol Int. 2017; 34(2): 162–191. PubMed Abstract | Publisher Full Text\n\nStoupel E, Petrauskiene J, Abramson E, et al.: Relationship between deaths from stroke and ischemic heart disease--environmental implications. J Basic Clin Physiol Pharmacol. 1999; 10(2): 135–45. PubMed Abstract | Publisher Full Text\n\nStoupel E, Petrauskiene J, Kalediene R, et al.: Clinical cosmobiology: the Lithuanian study 1990-1992. Int J Biometeorol. 1995; 38(4): 204–8. PubMed Abstract | Publisher Full Text\n\nStoupel E, Petrauskiene J, Kalediene R, et al.: Distribution of deaths from ischemic heart disease and stroke. Environmental and aging influences in men and women. J Basic Clin Physiol Pharmacol. 1996; 7(4): 303–19. PubMed Abstract | Publisher Full Text\n\nWHO: Noncommunicable diseases. 2018. Reference Source\n\nYan Q: Circadian Biomarkers and Chronotherapy: Implications for Personalized and Systems Medicine. Cellular Rhythms and Networks. SpringerBriefs in Cell Biology. Springer, Cham, 2015; 71–81. Publisher Full Text\n\nZhao XR, Gonzales N, Aronowski J: Pleiotropic role of PPARγ in intracerebral hemorrhage: an intricate system involving Nrf2, RXR, and NF-κB. CNS Neurosci Ther. 2015; 21(4): 357–366. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "71839",
"date": "03 Nov 2020",
"name": "Emmanuel Poulidakis",
"expertise": [
"Reviewer Expertise Heart failure",
"echocardiography",
"interventional cardiology."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nOverall it is a good work, in a rather unusual subject, but seems to well support its arguments. However, I believe that while its point is interesting, it does not sufficiently explain many of the terms used, as if addressing only those interested in the same field of research, somehow barring less expert audience. I suggest that in many cases, a few more phrases could make a difference.\n\nFor example, explaining certain terms (e.g. Sunspot numbers and time series) could make the article readable by a less specialized audience. I had to conduct research on my own to better understand these concepts. I would also suggest expanding the paragraph at the end of the introduction, detailing the aims of the study. I have attached a pdf copy with highlighted sections about possible corrections.\nAnother observation is the structure of the abstract, which in my opinion should be divided into sections (background/methods/results/conclusion).\n\nI noted a few spelling errors, highlighted in the pdf file.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? I cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "70999",
"date": "13 Nov 2020",
"name": "Kateřina Podolská",
"expertise": [
"Reviewer Expertise Stochastic Modeling",
"Demographic Analysis",
"Solar Terrestrial Interactions",
"Solar Flares"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this manuscript, the authors propose non-anthropogenic 6.8 days cycle of stroke-related death events in Pireus port during 1985-1989 period.\nReliable data sources of solar and magnetic activity are used.\n\nI suggest to answer the following questions and comments before the manuscript will be accepted for indexing:\nPlease provide codes of underlined cause of death from the of the International Classification of Diseases (ICD-9) under which was analyzed stroke-related death events registered.\nHave you tried to use as another proxy of the solar activity in your analysis also Solar radio flux F10.7? The Solar radio flux reflects the nature of the process in a better way.\n\nThe discussion part should be enlarged with consideration different dynamic of geomagnetic and solar activity during the ascending and descending phase of the solar cycle. Influence of this effect in changes of mortality from cardiovascular diseases was publicized.\n\nThe effect of cosmic rays during the the solar cycle minima, to which human physiology is sensitive, should be also mentioned.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1088
|
https://f1000research.com/articles/8-1585/v1
|
04 Sep 19
|
{
"type": "Research Article",
"title": "Predicting the chances of live birth for couples undergoing IVF-ICSI: a novel instrument to advise patients and physicians before treatment",
"authors": [
"Bruna Estácio da Veiga",
"Duarte Pedro Tavares",
"José Luis Metello",
"Fernando Ferreira",
"Pedro Ferreira",
"José Manuel Fonseca",
"Duarte Pedro Tavares",
"José Luis Metello",
"Fernando Ferreira",
"Pedro Ferreira",
"José Manuel Fonseca"
],
"abstract": "Background: In developed countries, the prevalence of infertility ranges from 3.5% to 16.7%. Therefore, the number of in vitro fertilization technique (IVF) and its subtype intracytoplasmic sperm injection (ICSI) treatments has been significantly increasing across Europe. Several factors affect the success rate of in vitro treatments, which can be used to calculate the probability of success for each couple. As these treatments are complicated and expensive with a variable probability of success, the most common question asked by IVF patients is ‘‘What are my chances of conceiving?”. The main aim of this study is to develop a validated model that estimates the chance of a live birth before they start their IVF non-donor cycle. Methods: A logistic regression model was developed based on the retrospective study of 737 IVF cycles. Each couple was characterized by 14 variables (woman’s and man’s age, duration of infertility, cause of infertility, woman’s and man’s body mass index (BMI), anti-Müllerian hormone (AMH), antral follicle count (AFC), woman’s and man’s ethnicity, woman’s and man’s smoking status and woman’s and man’s previous live children) and described with the outcome of the treatment \"Live birth\" or \"No live birth\". Results: The model results showed that from the 14 variables acquired before starting the IVF procedures, only male factor, man’s BMI, man's mixed ethnicity and level of AMH were statistically significant. The interactions between infertility duration and woman’s age, infertility duration and man’s BMI, AFC and AMH, AFC and woman’s age, AFC and woman’s BMI and AFC and disovulation were also statistically significant. The area under the receiver operating characteristic (AUROC) curve test for the discriminatory ability of the final prediction model is 0.700 (95% confidence interval (CI) 0.660–0.741). Conclusions: This model might result in a new validated decision support system to help physicians to manage couples’ expectations.",
"keywords": [
"IVF",
"ICSI",
"prediction model",
"live birth",
"assisted reproduction",
"logistic regression"
],
"content": "Introduction\n\nInfertility is defined as a disease of the reproductive system defined by the failure to achieve a clinical pregnancy after 12 months or more of regular unprotected sexual intercourse1. Taking this definition into account, it is a challenge to calculate the real prevalence or incidence rates for this condition2.\n\nAccording to the study performed by Boivin et al.3, the prevalence of infertility ranged from 3.5% to 16.7% in developed countries. Based on these authors’ estimates, 72.4 million women are currently infertile and, of these, 40.5 million are currently seeking infertility medical care3. Most recent Portuguese data estimate that 9.8% of couples are infertile4.\n\nInfertility is a multifactorial disease5. A male-related factor can be the cause, especially when there is a sperm or ejaculatory problem6,7 or it may be a female-related factor, namely when there is a disovulation or tubal dysfunction8,9. Women's age, ovarian reserve, weight, and lifestyle factors – such as alcohol consumption and exercise habits - have also been related to this5,10–16. Nonetheless, in up to 20% of cases, no cause can be found3,5.\n\nSeveral treatments have been proposed during the last century, with the most innovative bring in vitro fertilization (IVF), which was developed by Robert Edwards17 in 1978. After 41 years, the two most important medically assistant reproduction techniques (MAR) are in vitro fertilization (IVF) and its subtype intracytoplasmic sperm injection (ICSI)18. The number of IVF and ICSI treatments has been increasing across Europe19. According to the European Society of Human Reproduction and Embryology, the number of MAR cycles between 1997 and 2014 increased by 13%, reaching 776 556 cycles in Europe in 201519.\n\nDespite the increasing number of MAR treatments, IVF success is not guaranteed20. In healthy young couples, the probability of achieving a live birth is between 20% and 25% per month. This probability may increase by up to 60% with MAR techniques21. The study by Malizia et al. also corroborates this finding, indicating that between 38% and 49% of couples who start IVF remain childless, even after undergoing up to six IVF cycles22. In Portugal, the last report of MAR showed that the treatments success rate is around 25–30%23.\n\nMAR treatments are expensive, time-consuming, stressful, and may lead to anxiety, depression, or marital problems24–28. Also, these treatments have possible complications such as ovarian hyperstimulation syndrome, bleeding, and infection, as well as multiple or premature births29.\n\nThe success rates fluctuate between studies and are dependent on several factors30. If the chances of live birth are low and the risk or the cost is too high, the couple may consider other options such as adoption or to remain childless31. In other words, based on their specific probability of success, the couple may decide whether or not they proceed with the treatment. Therefore, the most common question asked by IVF patients is ‘‘What are my chances of conceiving?”\n\nThe answer to this tough question usually depends on the woman’s age and infertility diagnosis30. Nevertheless, many more parameters are known to affect IVF outcomes32 such as sperm quality, hormone doses, and physiological factors. Given this, a decision support system would be helpful to ensure that infertile couples are well informed regarding their chances of success with IVF33.\n\nThere have been various efforts to build prediction models to assist physicians in predicting MAR success30,34–37. To our knowledge, the first predictive model ever built in this context is from Templeton et al. in 1996 using a logistic regression model to predict the probability of live birth for an individual woman using the woman’s age, number of previous live birth or pregnancies not resulting in a live birth, whether these were a result of previous IVF treatment, female causes of infertility, duration of infertility and the number of previous unsuccessful IVF treatments30.\n\nThese authors found that the success of IVF decreased with female age and that women between 25 and 30 years were the most likely to have a live birth30.\n\nConsidering the evolution of technology and the importance of other variables, Nelson and Lawlor developed a new model in 2011, using the same mathematical techniques, but including other factors such as the most prevalent causes of infertility, the source of the egg (donor or patient’s own), type of hormonal preparation used (antioestrogen, gonadotrophin, or hormone replacement therapy), whether or not ICSI was used, and the number of previous cycles (1, 2 or 3)36.\n\nTaking into account that it is imperative to have validated models38, in 2014, Velde et al. used their cohort to validate Templeton and Nelson and Lawlor's model's performance. They found that Templeton’s model underestimated success rates while Nelson’s model overestimated it38.\n\nOther relevant models of note include the one developed by Marca et al. which predicts live birth in assisted reproduction based on serum anti-Müllerian hormone (AMH) and women’s age35, and the Mc Lernon et al. study37 that estimated the cumulative personal chance of a first live birth over a maximum of six complete cycles of IVF using data from 23 417 women in the UK. The estimates of the last study37 were only adjusted by the woman’s age, infertility duration, previous pregnancies, infertility causes, and type of treatment.\n\nNonetheless, when Leijdekkers performed an external validation of Mc Lernon’s model39 with Dutch women, it was found that there was an overestimation of the results and he decided to include biomarkers such as anti-Müllerian hormone (AMH), antral follicle count (AFC) and women’s body weight. Other studies indicate it is important to include covariables such as body mass index (BMI), ethnicity and ovarian reserve34 and corroborate that the use of variables such as BMI, AFC, AMH, ethnicity, and smoking status should be predictors in the model5,14,40,41.\n\nFurthermore, a machine learning approach was also proposed, which includes decision trees, genetic algorithms, and k nearest neighbors classifiers21,42–45.\n\nThe main aim of this study was to integrate all the variables of interest referred in the literature to develop a validated model that estimates the chance of live birth for couples before they start their IVF non-donor cycle.\n\n\nMethods\n\nThe present work is a retrospective study of data from IVF/ICSI cycles. The cycles were performed between 2012 and 2016 in the Centro de Infertilidade e Reprodução Medicamente Assistida (CIRMA) at Hospital Garcia de Orta, E.P.E., Almada, Portugal.\n\n739 couples were considered once they met the criterion of being fresh non-donor cycles or cycles with live birth or without frozen embryos available.\n\nThe study was approved by the Hospital’s Ethics Committee for Health, couples signed an informed consent document before treatment begins at the first infertility consultation. Patients were informed that, under complete anonymity, the data may be used in scientific papers for public presentation or publication46.\n\nThe IVF result for each couple in the study was classified as 1 (if, at least, one baby was born alive and survived for more than 1 month) and 0 (otherwise), which is the model’s dependent variable/primary outcome. This decision was based on other studies in this area30,34–37.\n\nIn terms of the baseline characteristics used to develop this model and based on evidence from the published literature in this area30,32 the woman’s and man’s age (years), duration of infertility (months), cause of infertility (categorised as tubal factor, endometriosis, disovulation, male factor, both female and male factor - depending on whether the cause of infertility underlies the woman or the man - multiple female factors, unexplained infertility or other), woman’s and man’s BMI (kg/m2), AMH (ng/mL), AFC (antral follicle count), woman’s and man’s ethnicity (categorised as Asian, Caucasian, Gipsy, Indian, Black or Mixed), woman’s and man’s smoking status (never, previous, present) and woman’s and man’s previous live children (yes or no) were considered. Antral follicle count was obtained by transvaginal ultrasound, and serum anti-Müllerian hormone levels were measured by blood analysis. There should be no bias associated with this study, however, possible sources of bias may arise from couples responses during the consultation and the entry of the data in the database.\n\nA summary statistical evaluation was performed for each variable. A univariate analysis was first performed. All continuous variables were compared using the t-student test and the categorical one using the Chi-square test. A p-value <0,05 was considered statistically significant47.\n\nA binary logistic regression48 was developed using the primary outcome as binary (no=0 and yes=1) to achieve the aim of this study. An automatic backward selection process based on the Wald statistic was used to determine the final and the best logistic regression model49. The model was based on the data from 737 couples (99,73%), because in 2 couples there was data missing (lack of BMI values in two couples) and as Iezzoni50 indicated, it is not recommended to input missing values in health data.\n\nAs recommended by Hosmer and Lemeshow48 the interactions between some variables were included, to increase the reliability of the model because as Fisher51 explained, the primary outcome can depend not only on one individual baseline characteristic but also on the relationship between baseline characteristics. The variables in interaction were: female’s age and infertility duration52, BMI and oligospermia53, AFC levels and women’s BMI13 and AFC and AMH levels40.\n\nThe model’s performance was measured with the discriminative power assessed using the area under the curve (AUC) value54,55 and the model was internally validated using the bootstrapping technique with 1000 iterations56. All statistical procedures were computed using IBM SPSS Statistics 25, and it was considered an α level of 0.05, which means that was assumed that there isn't statistical significance if the p-value of the test is above 0.05\n\nThe STROBE cross-sectional reporting guidelines were adopted in this article57.\n\n\nResults\n\nA total of 737 cycles were evaluated. The overall rate of at least one live birth was 31.4%.\n\nBaseline characteristics of couples are presented in Table 1 and Table 2.\n\nBMI – body mass index; AMH - anti-Müllerian hormone; AFC – antral follicle count\n\nThe average age for female participants was 34.04 years (Table 1), and 36.14 for male participants. Younger women and men were more likely to achieve a live birth. The same was seen for men and women with lower BMI and shorter infertility durations. However, these results were only statistically significant for women’s age, men’s age, AFC and AMH (p<0.05).\n\nTable 2 shows that most women and men had never smoked. Most men and women were Caucasian and the male factor, which means male infertility, is the most prevalent infertility cause in this data. Most men and women had no previous live births (90,1% and 87,4%). Chi-square test revealed that no discrete characteristics have statistically significant differences for live birth (p>0.05). For this reason, interactions on different characteristics were considered.\n\nIn Table 3, the binary logistic regression parameters computed with SPSS are presented. These parameters allow the assessment of the chance of living birth on couples before they start their IVF non-donor cycle based on the variables previously mentioned.\n\nBMI – body mass index; AMH - anti-Müllerian hormone; AFC – antral follicle count\n\nThe logistic regression results showed that from the 14 variables evaluated on the pre-treatment procedures, only male factor, the man’s BMI, mixed ethnicity for men, and the level of AMH were statistically significant. Apart from these variables, the interactions between infertility duration and women’s age, infertility duration and men’s BMI, AFC and AMH, AFC and women’s age, AFC and women's BMI and AFC and disovulation were also statistically significant (p-value less than 0.05). The interactions between infertility duration and woman’s age and man’s BMI; AFC and AMH, woman’s age, woman’s BMI and disovulation were also statistically significant.\n\nAccording to regression coefficients of the final model, an AMH unit increase raises the probability of IVF-ICSI success by 0.172 times, ceteris paribus. Similarly, the interaction between AFC and the woman’s BMI decreases the same probability by 0.003 times and interaction between AFC and the woman’s age increases it by 0.004 times. Male factor is the only infertility cause that enters individually into the final model raising the chances of success by 0.420 times.\n\nThis shows the ROC curve test for the discriminatory ability of the final prediction model is 0.700 (95% confidence interval (CI) 0.660–0.741).\n\n\nDiscussion\n\nSo far, providing an accurate prediction of the chances of achieving a live birth after FIV-ISCI treatments has not been an easy task32. In this study, a novel prediction model was built, including almost all clinical factors reported as important in the literature.\n\nTo our knowledge, this is the first model for live birth FIV-ICSI prediction that accounts for men’s ethnicity. It also includes variables such as BMI and AFC, which were not included in many models before34. This model includes all the characteristics that have been indicated as important in the literature to predict the chances of live birth in MAR5,30,32.\n\nThe Templeton model30 considers the existence of previous IVF cycles and has been externally validated by Nelson and Lawlor36. Therefore, both these models can be applied before IVF is started and predict the success of IVF/ICSI treatment. During this study, it was not possible to obtain information about the number of previous cycles per couple, and so this factor was not included.\n\nThis model supports earlier findings in this scientific area such as the importance of AMH and AFC to predict live birth as predicted by other studies35,40 and indirectly corroborates the importance of woman’s age to predict the result as shown by Templeton et al.30. Nonetheless, it is crucial to refer that the increase of women’s age, when in interaction with infertility duration, significantly decreases the probability of a living birth which is in accordance to Nelson’s study36.\n\nConcerning the main reason for treatment, both male factor and disovulation (with AFC interaction) are the only causes that emerged in the final model. These causes raised the chances of success, which might be explained with the IVF-ICSI technique itself once the problem of sperm and ovulation anomalies are overcome18. In other words, women with ovulation problems and men with sperm anomalies have higher chances of success with IVF-ICSI technique because sperm and oocytes are medically collected and interact directly, although in an in vitro environment18.\n\nAnother notable finding is that an increase of the man’s BMI (per si) or woman’s BMI (with AFC interaction) decreased the chances of live birth, which is in accordance with the scientific literature. Women who are overweight are known to have ovulatory problems, and increased risks of miscarriage10 and obesity may adversely affect male reproduction by endocrine, thermal, genetic, and sexual mechanisms10. For instance, increased testicular temperature can result from prolonged periods of sitting in men with excessive lower abdominal fat deposited in the lower abdomen58. Also, obesity tends to increase estrogen levels and reduce testosterone levels in men59.\n\nConcerning the man's ethnicity, we found that if the man is of mixed ethnicity, the odds of success are around 3.8. This finding must be interpreted cautiously as the sample is composed of 92% Caucasian males, with only 2.6% of males being of mixed ethnicity. Up to now, there doesn't seem to be any biological plausibility for this find, further researchers is required to validate this possble relationship between mixed ethnicity in men and increased success in IVF treatments.\n\nA limitation of our model is that it is restricted to pre-treatment stages. Thereby, the physician must explain to patients when using our model that their success probability invariably changes during the cycles process. The resulting probability of our model should be considered a baseline one.\n\nMany studies have accounted for the chances of live birth after IVF or ICSI treatment30,34–39. When compared with other studies, the most similar one is Dhillon et al.’s model34 due to the variables integrated into the model. However, in our study, there were no limitations on socioeconomic status since our data has been extracted from a public hospital with universal access60 and so the results can be generalized to people of all socioeconomic backgrounds. This is an advantage over the Dhillon et al.’s model34, where it is estimated that 75% of couples paid for their treatment and therefore their model could not be generalized to all social classes.\n\nPredictive ability of the prediction models on medicine has been assessed by the AUC54,55. In general, AUC for prediction models in reproductive medicine is rather low, ranging between 0.59 and 0.6455. Table 4 shows the values of models’ AUC predicting the chances of live birth for couples undergoing IVF-ICSI. Although McLernon et al.37 have the highest value of the area under the receiver operating characteristic (AUROC) curve, that value decreased on Leijedekkers validation for 0.6239. The model developed in this study has an AUC of 0.700, which is the second-highest value, and so has a comparable discriminatory ability with these previous models.\n\nTaking into account that Coppus et al.’s systematic review concluded that prediction models in reproductive medicine would be limited to an AUC value of 0.65 due to the relatively homogeneous group of subfertile patients55, this study can be considered to improve upon the current available models.\n\nThe present model predicts the specific probability of a live birth based on easily acquirable couple characteristics before starting a treatment. It might help patients to understand the limitations of an IVF/ICSI in their particular case and also physicians to compare different treatment strategies. Furthermore, it might support institutions to predict the probability of the need for treatment repetition based on the specific characteristics of each couple.\n\nWe intend in the near future to perform an external validation with the developed model with a new dataset from Centro de Infertilidade e Reprodução Medicamente Assistida (CIRMA). Follow this, it is also planned to perform an external geographical validation to check if there is a geographical influence on chances of live birth for couples undergoing IVF-ICSI.\n\n\nConclusions\n\nMany couples with fertility problems ask themselves if they should undergo an IVF-ICSI treatment. Our novel model provides an estimated probability of their chances of live birth. This is the first model with Portuguese data and takes in account the important variables described in literature such as AFC, AMH, and woman's age. This tool may help physicians to shape couples' expectations conceding them the opportunity to plan their treatments and to prepare both emotionally and financially to them and so we are developing a user-friendly interface to help physicians in their clinical practice.\n\n\nData availability\n\nThe data that support the findings of this study are available on request from the author José Metello (jose.metello@hgo.min-saude.pt). The data cannot be made publicly available in accordance to paragraph d), number 2, Article 9 of the Regulation no. 2016/679 of the European Parliament and the Council of 27 April 2016 and due to the sensitive data containing information that could compromise the privacy of research participants. Informed consent was provided by patients who participated in the study for the use of their data for scientific purposes only and to safeguard their anonymity.",
"appendix": "References\n\nZegers-Hochschild F, Adamson GD, Dyer S, et al.: The international glossary on infertility and fertility care, 2017. Hum Reprod. 2017; 32(9): 1786–1801. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGurunath S, Pandian Z, Anderson RA, et al.: Defining infertility — a systematic review of prevalence studies. Hum Reprod Update. 2011; 17(5): 575–588. PubMed Abstract | Publisher Full Text\n\nBoivin J, Bunting L, Collins JA, et al.: International estimates of infertility prevalence and treatment-seeking: potential need and demand for infertility medical care. Hum Reprod. 2007; 22(6): 1506–1512. PubMed Abstract | Publisher Full Text\n\nSilva-Carvalho JL, Santos A: Estudo Afrodite: Caracterização da infertilidade em Portugal (Vol. 1. Estudo na Comunidade). Porto Fac Med da Univ do. 2009; 74.\n\nVander Borght M, Wyns C: Fertility and infertility: Definition and epidemiology. Clin Biochem. 2018; 62: 2–10. PubMed Abstract | Publisher Full Text\n\nYing LY, Wu LH, Loke AY: Gender differences in experiences with and adjustments to infertility: A literature review. Int J Nurs Stud. 2015; 52(10): 1640–52. PubMed Abstract | Publisher Full Text\n\nCarlsen E, Giwercman A, Keiding N, et al.: Evidence for decreasing quality of semen during past 50 years. BMJ. 1992; 305(6854): 609–613. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGarcía D, Brazal S, Rodríguez A, et al.: Knowledge of age-related fertility decline in women: A systematic review. Eur J Obstet Gynecol Reprod Biol. 2018; 230: 109–118. PubMed Abstract | Publisher Full Text\n\nKovanci E, Carson SA: Female infertility and the evaluation of an infertile couple. Clin Gynecol Second Ed. 2015; 865–883.\n\nBest D, Bhattacharya S: Obesity and fertility. Horm Mol Biol Clin Investig. 2015; 24(1): 1–6. PubMed Abstract | Publisher Full Text\n\nNorman RJ: Impact of obesity on female fertility and fertility treatment. Womens Health (Lond). 2008; 4(2): 183–194. PubMed Abstract | Publisher Full Text\n\nFreeman EW, Gracia CR, Sammel MD, et al.: Association of anti-mullerian hormone levels with obesity in late reproductive-age women. Fertil Steril. 2007; 87(1): 101–6. PubMed Abstract | Publisher Full Text\n\nFreour T, Masson D, Mirallie S, et al.: Active smoking compromises IVF outcome and affects ovarian reserve. Reprod Biomed Online. 2008; 16(1): 96–102. PubMed Abstract | Publisher Full Text\n\nKünzle R, Mueller MD, Hänggi W, et al.: Semen quality of male smokers and nonsmokers in infertile couples. Fertil Steril. 2003; 79(2): 287–291. PubMed Abstract | Publisher Full Text\n\nMargolis R, Wright L: Better Off Alone Than With a Smoker: The Influence of Partner's Smoking Behavior in Later Life. J Gerontol B Psychol Sci Soc Sci. 2016; 71(4): 687–697. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDonarelli Z, Lo Coco G, Gullo S, et al.: Infertility-related stress, anxiety and ovarian stimulation: can couples be reassured about the effects of psychological factors on biological responses to assisted reproductive technology? Reprod Biomed Soc Online. 2016; 3: 16–23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJohnson MH: Robert Edwards: the path to IVF. Reprod Biomed Online. 2006; 12(2): 234–253.\n\nChow S, Cheung S, Parrella A, et al.: Intracytoplasmic sperm injection: state of the art in humans. Reproduction. 2004; 1741–7899.\n\nNyboe Andersen A, Goossens V, Bhattacharya S, et al.: Assisted reproductive technology and intrauterine inseminations in Europe, 2005: results generated from European registers by ESHRE: ESHRE. The European IVF Monitoring Programme (EIM), for the European Society of Human Reproduction and Embryology (ESHRE). Hum Reprod. 2009; 24(6): 1267–1287. PubMed Abstract | Publisher Full Text\n\nDe Geyter C, Calhaz-Jorge C, Kupka MS, et al.: ART in Europe, 2014: results generated from European registries by ESHRE: The European IVF-monitoring Consortium (EIM) for the European Society of Human Reproduction and Embryology (ESHRE). Hum Reprod. 2018; 33(9): 1586–1601. PubMed Abstract | Publisher Full Text\n\nPegah Hafiz BN, Nematollahi M, Boostani R: Predicting implantation outcome of in vitro fertilization and intracytoplasmic sperm injection using data mining techniques. Int J Fertil Steril. 2017; 11(3): 184–190. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMalizia BA, Hacker MR, Penzias AS: Cumulative live-birth rates after in vitro fertilization. Obstet Gynecol Surv. 2009; 64(5): 315–316. Publisher Full Text\n\nConselho Nacional de Procriação Medicamente Assistida: Relatório da Atividade Desenvolvida pelos Centros de PMA em 2015. 2015; 27: 2015.\n\nAssociação Portuguesa de Fertilidade. [Accessed: 24-Jan-2019]. Reference Source\n\nKaraca N, Karabulut A, Ozkan S, et al.: Effect of IVF failure on quality of life and emotional status in infertile couples. Eur J Obstet Gynecol Reprod Biol. 2016; 206: 158–163. PubMed Abstract | Publisher Full Text\n\nMaroufizadeh S, Karimi E, Vesali S, et al.: Anxiety and depression after failure of assisted reproductive treatment among patients experiencing infertility. Int J Gynecol Obstet. 2015; 130(3): 253–256. PubMed Abstract | Publisher Full Text\n\nValoriani V, Lotti F, Lari D, et al.: Differences in psychophysical well-being and signs of depression in couples undergoing their first consultation for assisted reproduction technology (ART): an Italian pilot study. Eur J Obstet Gynecol Reprod Biol. 2016; 197: 179–185. PubMed Abstract | Publisher Full Text\n\nHeredia M, Tenías JM, Rocio R, et al.: Quality of life and predictive factors in patients undergoing assisted reproduction techniques. Eur J Obstet Gynecol Reprod Biol. 2013; 167(2): 176–180. PubMed Abstract | Publisher Full Text\n\nPaulson R: in vitro Fertilization. UpToDate. 2018; 379–386. Reference Source\n\nTempleton A, Morris JK, Parslow W: Factors that affect outcome of in-vitro fertilisation treatment. Lancet. 1996; 348(9039): 1402–1406. PubMed Abstract | Publisher Full Text\n\nKarpel L, Frydman N, Hesters L, et al.: [Talking about adoption during IVF]. Gynecol Obstet Fertil. 2007; 35(3): 232–239. PubMed Abstract | Publisher Full Text\n\nZarinara A, Zeraati H, Kamali K, et al.: Models Predicting Success of Infertility Treatment: A Systematic Review. J Reprod Infertil. 2016; 17(2): 68–81. PubMed Abstract | Free Full Text\n\nKhalifa M: Clinical Decision Support: Strategies for Success. Procedia Comput Sci. 2014; 37: 422–427. Publisher Full Text\n\nDhillon RK, McLernon DJ, Smith PP, et al.: Predicting the chance of live birth for women undergoing IVF: a novel pretreatment counselling tool. Hum Reprod. 2016; 31(1): 84–92. PubMed Abstract | Publisher Full Text\n\nLa Marca A, Nelson SM, Sighinolfi G, et al.: Anti-Müllerian hormone-based prediction model for a live birth in assisted reproduction. Reprod Biomed Online. 2011; 22(4): 341–349. PubMed Abstract | Publisher Full Text\n\nNelson SM, Lawlor DA: Predicting live birth, preterm delivery, and low birth weight in infants born from in vitro fertilisation: a prospective study of 144,018 treatment cycles. PLoS Med. 2011; 8(1): e1000386. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcLernon DJ, Steyerberg EW, Te Velde ER, et al.: Predicting the chances of a live birth after one or more complete cycles of in vitro fertilisation: population based study of linked cycle data from 113 873 women. BMJ. 2016; 355: i5735. PubMed Abstract | Publisher Full Text | Free Full Text\n\nte Velde ER, Nieboer D, Lintsen AM, et al.: Comparison of two models predicting IVF success; the effect of time trends on model performance. Hum Reprod. 2014; 29(1): 57–64. PubMed Abstract | Publisher Full Text\n\nLeijdekkers JA, Eijkemans MJC, van Tilborg TC, et al.: Predicting the cumulative chance of live birth over multiple complete cycles of in vitro fertilization: an external validation study. Hum Reprod. 2018; 33(9): 1684–1695. PubMed Abstract | Publisher Full Text\n\nKeane K, Cruzat VF, Wagle S, et al.: Specific ranges of anti-Mullerian hormone and antral follicle count correlate to provide a prognostic indicator for IVF outcome. Reprod Biol. 2017; 17(1): 51–59. PubMed Abstract | Publisher Full Text\n\nDechanet C, Anahory T, Mathieu Daude JC, et al.: Effects of cigarette smoking on reproduction. Hum Reprod Update. 2011; 17(1): 76–95. PubMed Abstract | Publisher Full Text\n\nJurisica I, Mylopoulos J, Glasgow J, et al.: Case-based reasoning in IVF: prediction and knowledge mining. Artif Intell Med. 1998; 12(1): 1–24. PubMed Abstract | Publisher Full Text\n\nGuh RS, Wu TCJ, Weng SP: Integrating genetic algorithm and decision tree learning for assistance in predicting in vitro fertilization outcomes. Expert Syst Appl. 2011; 38(4): 4437–4449. Publisher Full Text\n\nGüvenir HA, Misirli G, Dilbaz S, et al.: Estimating the chance of success in IVF treatment using a ranking algorithm. Med Biol Eng Comput. 2015; 53(9): 911–920. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMilewska AJ, Jankowska D, Cwalina U: Prediction of Infertility Treatment Outcomes Using Classification Trees. Stud LOGIC Gramm Rhetor. 2016; 47(60): 7–19. Publisher Full Text\n\ncnpma: Conselho Nacional de Procriação Medicamente Assistida - Modelos de Consentimento Informado. [Accessed: 13-Aug-2019]. Reference Source\n\nDaniel W, Cross C: Biostatistics: A Foundation for Analysis in the Health Sciences. 10th ed. Hoboken: John Wiley & Sons, 2013. Reference Source\n\nHosmer DW, Lemeshow S: Applied Logistic Regression. 2nd ed. NY: John Wiley & Sons, Inc, 2000. Publisher Full Text\n\nMarill T, Green DM: On the Effectiveness of Receptors in Recognition Systems. IEEE Trans Inf theory. 1963; 9(1): 11–17. Publisher Full Text\n\nIezzoni L: Risk Adjustment for Measuring Health Care Outcomes. 4th ed. 2012.\n\nFisher RA: The Arrangement of Field Experiments. J Minist Agric Gt Britain. 1926; 33: 503–513. Reference Source\n\nHart RJ: Physiological Aspects of Female Fertility: Role of the Environment, Modern Lifestyle, and Genetics. Physiol Rev. 2016; 96(3): 873–909. PubMed Abstract | Publisher Full Text\n\nHajshafiha M, Ghareaghaji R, Salemi S, et al.: Association of body mass index with some fertility markers among male partners of infertile couples. Int J Gen Med. 2013; 6: 447–451. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSwets JA: Measuring the accuracy of diagnostic systems. Science. 1988; 240(4857): 1285–1293. PubMed Abstract | Publisher Full Text\n\nCoppus SF, Van Der Veen F, Opmeer BC, et al.: Evaluating prediction models in reproductive medicine. Hum Reprod. 2009; 24(8): 1774–1778. PubMed Abstract | Publisher Full Text\n\nEfron B, Tibshirani JT: An Introduction to the Bootstrap. 1st ed. 1993. Reference Source\n\nvon Elm E, Altman DG, Egger M, et al.: The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008; 61(4): 344–349. PubMed Abstract | Publisher Full Text\n\nHammoud AO, Meikle AW, Reis LO, et al.: Obesity and male infertility: a practical approach. Semin Reprod Med. 2012; 30(6): 486–495. PubMed Abstract | Publisher Full Text\n\nSchneider G, Kirschner MA, Berkowitz R, et al.: Increased estrogen production in obese men. J Clin Endocrinol Metab. 1979; 48(4): 633–638. PubMed Abstract | Publisher Full Text\n\nAssembleia da República: Lei n.o 56/79 - Diário da República n.o 214/1979, Série I de 1979-09-15. 1979; 2357–2363. Reference Source"
}
|
[
{
"id": "68000",
"date": "31 Jul 2020",
"name": "Charalampos Siristatidis",
"expertise": [
"Reviewer Expertise Assisted reproduction",
"mmethodology",
"artificial intelligence"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAuthors developed a logistic regression model based on the retrospective study of 737 IVF cycles. They concluded that the model might result in a new validated decision support system to help physicians to manage couples’ expectations concerning their possibility to have a live birth.\nThe introduction section is unreasonably extended, without providing the necessary information only, leading to the rationale of the study. Important similar models using more sophisticated networks, such as artificial intelligence, are missing.\nLanguage and grammar need revision. The sample size is small. A proper power calculation is missing. Some important parameters that have been tested in other studies that contribute to live birth are missing. Limitations are not reported.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "6092",
"date": "22 Dec 2020",
"name": "Bruna Estácio da Veiga",
"role": "Author Response",
"response": "Lisbon, November the 3rd 2020 We submitted an original research article entitled “Predicting the chances of live birth for couples undergoing IVF ICSI: a novel instrument to advise patients and physicians before treatment” for consideration by F1000Research journal that has been reviewed by two experts in the field, suggesting revisions to this paper. We want to thank their comments and we have done the following revisions accordingly for the first reviewer (all modifications can be seen through the “Track Changes” option): Point 1: The introduction section is unreasonably extended, without providing the necessary information only, leading to the rationale of the study. Important similar models using more sophisticated networks, such as artificial intelligence, are missing. Response: We followed the reviewer’s advice by modifying the introduction section, including more information about AI in predictive models, considering the recent study of Qiu et.al (reference 32) based on supervised learning algorithms. Point 2: Language and grammar need revision. Response: The reviewer asked for English changes and we reviewed the entire text. Point 3: The sample size is small. A proper power calculation is missing. Response: Considering the size of the sample available and the reviewer’s assessment of the model’s evaluation, we added a table (Table 4) that lists 5 performance metrics (i.e., accuracy, F‑score, precision, sensitivity and sensibility) and reflects our model performance, in order to respond to the reviewer’s observation. Particularly about the sample size, we still achieved a ROC curve with a 95% Confidence Interval (CI) that is not very wide (“The ROC curve (…) had an AUC equal to 0.700 (95% CI 0.660–0.741)”). Point 4: Some important parameters that have been tested in other studies that contribute to live birth are missing. Response: Although other important parameters are found in other studies, the 14 variables tested in our study were all the variables collected by CIRMA and made available for this study. Therefore, we can not follow the reviewer’s suggestion and include more variables, because we did not collect and select the couple’s data. Point 5: Limitations are not reported. Response: We thank the reviewer for this final comment and we tried to enrich the discussion section by adding more clinical limitations found in our model to the limitation reported before: “A limitation of our model is that it is restricted to pre-treatment stages. Thereby, the physician must explain to patients when using our model that their success probability invariably changes during the cycles process. That is, the resulting probability of our model should be considered a baseline one. Taking into account that this study was performed with data from only one medical center, one limitation of these models is that only an internal validation was made. Moreover, the large temporal spectrum of the data (2012 to 2016) could mean that treatments made with older technology may have different results than those made with more recent ones (namely, vitrification procedures or culture media).”. Sincerely, The authors."
}
]
},
{
"id": "69512",
"date": "07 Sep 2020",
"name": "Jichun Tan",
"expertise": [
"Reviewer Expertise Assisted reproduction"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper focused on an interesting topic of \"What are my chances of conceiving?\". However, the aim of this study is not clear. The logistic regression model is not enough to establish a novel prediction model. To the best of our knowledge, similar studies with larger sample size, more clinical variables, and advanced calculating models have been reported to predict the rates of live birth. In addition, more latest studies should be added in the introduction section.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "6093",
"date": "22 Dec 2020",
"name": "Bruna Estácio da Veiga",
"role": "Author Response",
"response": "Lisbon, November the 3rd 2020 We submitted an original research article entitled “Predicting the chances of live birth for couples undergoing IVF ICSI: a novel instrument to advise patients and physicians before treatment” for consideration by F1000Research journal that has been reviewed by two experts in the field, suggesting revisions to this paper. We want to thank their comments and we have done the following revisions accordingly for the second reviewer (all modifications can be seen through the “Track Changes” option): Point 1: This paper focused on an interesting topic of \"What are my chances of conceiving?\". However, the aim of this study is not clear. Response: We followed the reviewer’s advice by modifying the abstract and introduction section, reinforcing the aim of our study: “The main aim of this study is to develop a validated model that estimates the chance of a live birth before the start of an IVF/ICSI non-donor cycle.” and “The main aim of this study was to find a pre-treatment predictor for achieving a live birth before an IVF/ICSI treatment.”, respectively. Point 2: The logistic regression model is not enough to establish a novel prediction model. To the best of our knowledge, similar studies with larger sample size, more clinical variables, and advanced calculating models have been reported to predict the rates of live birth. Response: Considering the size of the sample available and the reviewer’s assessment of the model’s evaluation, we added a table (Table 4) that lists 5 performance metrics (i.e., accuracy, F‑score, precision, sensitivity and sensibility) and reflects our model performance, in order to respond to the reviewer’s observation. Particularly about the sample size, we still achieved a ROC curve with a 95% Confidence Interval (CI) that is not very wide (“The ROC curve (…) had an AUC equal to 0.700 (95% CI 0.660–0.741)”). Although other important parameters are found in other studies, the 14 variables tested in our study were all the variables collected by CIRMA and made available for this study. Therefore, we can not follow the reviewer’s suggestion and include more variables, because we did not collect and select the couple’s data. Point 3: In addition, more latest studies should be added in the introduction section. Response: We thank the reviewer for this final comment and we tried to enrich the introduction section by adding more information about AI in predictive models, considering the recent study of Qiu et.al (reference 32) based on supervised learning algorithms. Sincerely, The authors."
}
]
}
] | 1
|
https://f1000research.com/articles/8-1585
|
https://f1000research.com/articles/8-2002/v1
|
26 Nov 19
|
{
"type": "Clinical Practice Article",
"title": "Case report: Treating a combination of hidradenitis suppurativa and psoriasis with different therapeutic approaches",
"authors": [
"Eleftheria Tampouratzi",
"Theodora Kanni",
"John Katsantonis",
"Theodora Douvali",
"Eleftheria Tampouratzi",
"John Katsantonis",
"Theodora Douvali"
],
"abstract": "Hidradenitis suppurativa and psoriasis are considered chronic inflammatory diseases suggesting the existence of common pathogenetic pathways. We present two cases of comorbid psoriasis and hidradenitis suppurativa, treated with certolizumab pegol and brodalumab due to failure of response to other conventional therapies. Monoclonal antibody therapies have revolutionized the treatment of chronic inflammatory disorders such as psoriasis and hidradenitis suppurativa. Given the good clinical response to anti-IL-17 and anti-tumor necrosis factor agents in patients undergoing psoriasis and hidradenitis treatment, investigations on this direction could represent the starting point in new therapeutic approach for revolutionary treatment in these difficult-to-treat diseases.",
"keywords": [
"hidradenitis suppurativa",
"psoriasis",
"certolizumab",
"brodalumab"
],
"content": "Introduction\n\nHidradenitis suppurativa (HS) and psoriasis are considered chronic inflammatory diseases suggesting the existence of common pathogenetic links1–3. Patients with psoriasis and HS have elevated levels of tumor necrosis factor (TNF) and interleukin-17 (IL-17) in lesional tissues, which has been the justification for selective targeting of these inflammatory pathways4–7. We present two cases of comorbidity of psoriasis and HS treated with certolizumab pegol and brodalumab due to the peculiarities of treatment with other therapies.\n\n\nCase report\n\nThe first patient, a 27-year-old Caucasian woman, presented with extensive psoriasis covering her head, trunk, lower limbs over a period of 5 years, concomitant psoriatic arthritis with axial joint involvement (manifestations of hierolagonitis) over the previous 2 years and moderate HS-stage II (according to the Hurley staging system) on the axillae (Figure 1a, b, c, d, e) with considerable pain, discomfort and substantial negative effect on quality of life over the last year, despite the limited extent of the lesions. The patient didn’t have a positive family history for the above diseases and the molecular control for HLA-B27 was negative. Previous treatments with topical corticosteroids and methotrexate for one year were not effective and treatment with apremilast for 8 months didn’t offer clinical improvement. The patient underwent comprehensive laboratory investigations, including complete blood cell count, chemistry panel, tuberculosis (Quantiferon-TB Gold test), human immunodeficiency virus and hepatitis B and C screening and chest x-Ray. Since all these examinations revealed values within normal limits and because of the patient’s desire for childbirth, she was treated with certolizumab pegol (CZP). The initial dose was 400mg, followed by 400mg every 2 weeks. Treatment with CZP significantly improved psoriasis and psoriatic arthritis at week 8 and HS at week 12 (Figure 1f–i). She continues treatment 9 months after and at 3 months follow-up is fully controlled.\n\n(a–e) Psoriatic and HS lesions of first patient before treatment with certolizumab pegol. (f–j) Psoriatic and HS lesions of first patient after treatment with certolizumab pegol.\n\nThe second patient, a 42-year-old Caucasian man, was referred to our hospital’s dermatological department with multiple, itchy, scaly, red-gray psoriatic plaques covering almost all his body: scalp, arms, trunk, thighs (Figure 2a–d) for the previous 6 months, over a history of 10 years psoriatic disease. The patient also experienced concomitant psoriatic arthritis with peripheral joint involvement and dactylitis discomfort over the previous 10 years, with moderate HS-stage II appearing on the groin area in the previous year. The above diseases had a negative impact factor on his quality of life. The patient’s family history was positive: his mother and sister were also suffering from psoriasis. The patient had until recently received almost all the available therapies related to his diseases: cyclosporine for 2 years interrupted due to urea and creatinin increase (examinations restored after discontinuation), methotrexate and golimumab for 3 years with improvement only in psoriatic arthritis, adalimumab ustekinumab and secukinumab, with a partial response. After a complete laboratory examination, with results in normal limits, the patient started therapy with brodalumab. The initial dose was 210 mg at weeks 0, 1, 2 followed by 210 mg every 2 weeks. His psoriasis and psoriatic arthritis were highly improved at week 8 (Figure 2 e–h), as was HS at week 16. He has continued treatment for 1 year; at 3 months follow-up he reported improvement in of his quality of life.\n\n(a–d) Psoriatic lesions of second patient before treatment with brodalumab. (e–h) Psoriatic lesions of second patient after treatment with brodalumab.\n\n\nDiscussion\n\nMonoclonal antibody therapies have revolutionized the treatment of chronic inflammatory disorders such as psoriasis and HS. CZP is a TNF inhibitor that does not have a fragment crystallizable (Fc) region, which is normally present in a complete antibody and therefore it does not cause antibody-dependent cell-mediated cytotoxicity8–10. In contrast to other whole-antibody anti-TNFs, CZP crosses the placenta only by passive diffusion and could therefore be considered as the first-line choice of treatment for women who wish to become pregnant. Since CZP is an anti-TNF drug, therapies which have good clinical response in both psoriasis/psoriatic arthritis and HS, it was chosen as the treatment of choice in our case since it also has a safe profile for possible future pregnancy.\n\nBrodalumab is a monoclonal antibody against human IL-17 receptor A (IL-17RA). Given its efficacy in psoriasis and its mechanism of action in psoriatic arthritis and HS, due to the patient’s non response to all the available treatment options it was decided its use on the above combination diseases11–14.\n\nIt is well known that psoriasis and HS likely share immunopathogenetic pathways, including involvement of IL-17 and TNF. Given the good clinical response to anti-IL 17 and anti-TNF drugs in psoriasis and HS treatment, investigations into this direction could represent a starting point for a new therapeutic approach for revolutionary treatment of two difficult to treat diseases.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nConsent\n\nWritten informed consent for publication of their clinical details and clinical images was obtained from the patients.",
"appendix": "References\n\nPatel M, Cohen JM, Wriight NA, et al.: Epidemiology of concomitant psoriasis and hidradenitis suppurativa (HS): experience of a tertiary medical center. J Am Acad Dermatol. 2015; 73(4): 701–702. PubMed Abstract | Publisher Full Text\n\nGiuseppe P, Nicola P, Valentina C, et al.: A Case of Moderate Hidradenitis Suppurativa and Psoriasis Treated with Secukinumab. Ann Dermatol. 2018; 30(4): 462–464. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKridin K, Shani M, Schonmann Y, et al.: Psoriasis and Hidradenitis Suppurativa: A Large-scale Population-based Study. J Am Dermatol. 2018; pii: S0190-9622(18)32962-1. PubMed Abstract | Publisher Full Text\n\nFrew JW, Hawkes JE, Krueger JG: A systematic review and critical evaluation of inflammatory cytokine associations in hidradenitis suppurativa [version 1; peer review: 2 approved, 1 approved with reservations]. F1000Res. 2018; 7: 1930. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFrew JW, Hawkes JE, Krueger JG: Topical, systemic and biologic therapies in hidradenitis suppurativa: pathogenic insights by examining therapeutic mechanisms. Ther Adv Chronic Dis. 2019; 10: 2040622319830646. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKanni T, Tzanetakou V, Savva A, et al.: Compartmentalized Cytokine Responses in Hidradenitis Suppurativa. PLoS One. 2015; 10(6): e0130522. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLynde CW, Poulin Y, Vnder R, et al.: Interleukin 17A: toward a new understanding of psoriasis pathogenesis. J Am Acad Dermatol. 2014; 71(1): 141–150. PubMed Abstract | Publisher Full Text\n\nPorter C, Armstrong-Fisher S, Kopotsa T, et al.: Certolizumab pegol does not bind the neonatal Fc receptor (FcRn): Consequences for FcRn-mediated in vitro transcytosis and ex vivo human placental transfer. J Reprod Immunol. 2016; 116: 7–12. PubMed Abstract | Publisher Full Text\n\nDattola A, Cannizzaro MV, Mazzeo M, et al.: Certolizumab Pegol in the Treatment of Psoriasis and Psoriatic Arthritis: Preliminary Real-Life Data. Dermatol Ther (Heidelb). 2017; 7(4): 485–492. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChimenti MS, Saraceno R, Chiricozzi A, et al.: Profile of certolizumab and its potential in the treatment of psoriatic arthritis. Drug Des Devel Ther. 2013; 7: 339–348. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKimmel G, Chima M, Kim HJ, et al.: Brodalumab in the treatment of moderate to severe psoriasis in patients when previous anti-interleukin 17A therapies have failed. J Am Acad Dermatol. 2019; 81(3): 857–859. PubMed Abstract | Publisher Full Text\n\nAttia A, Abushouk AI, Ahmed H, et al.: Safety and Efficacy of Brodalumab for Moderate-to-Severe Plaque Psoriasis: A Systematic Review and Meta-Analysis. Clin Drug Investig. 2017; 37(5): 439–51. PubMed Abstract | Publisher Full Text\n\nTchero H, Hrlin C, Bekara F, et al.: Hidradenitis Suppurativa: A Systematic Review and Meta-analysis of Therapeutic Interventions. Indian J Dermatol Venereol Leprol. 2019; 85(3): 248–257. PubMed Abstract\n\nBilal J, Riaz IB, Kamal MU, et al.: A Systematic Review and Meta-analysis of Efficacy and Safety of Novel Interleukin Inhibitors in the Management of Psoriatic Arthritis. J Clin Rheumatol. 2018; 24(1): 6–13. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "57163",
"date": "10 Dec 2019",
"name": "Thrasyvoulos Tzellos",
"expertise": [
"Reviewer Expertise Hidradenitis suppurativa",
"atopic dermatitis",
"biologics",
"Evidence based medicine"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI suggest changing the title to: \"Treating co-existence of hidradenitis suppurativa and psoriasis\"\nIntroduction:\nPlease change \"We present two cases of comorbidity of psoriasis and HS” to ”We present two cases of co-existence of psoriasis and HS”.\n\nFor case report 1:\nPlease provide a severity assessment for psoriasis.\n\nThe authors only refer to “extensive”. It would be important to report PASI or another measure of severity assessment.\n\nFor case report 2:\nPlease provide a severity assessment for psoriasis as for case 1.\n\nAlso it reported negative impact on quality of life. Please provide a measure if available. For example VAS pain 6 or DLQI 10.\n\nIs the background of the cases’ history and progression described in sufficient detail? Partly\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the conclusion balanced and justified on the basis of the findings? Yes",
"responses": [
{
"c_id": "6180",
"date": "22 Dec 2020",
"name": "Theodora Kanni",
"role": "Author Response",
"response": "Dear Prof. Tzellos, We revised our manuscript according to your comments. All your comments were taken into account and please find below the answers: We changed the title according to your suggestion. In the introduction section, we changed the term comorbidity with the term co-existence. Finally, we provide severity assessment before and after treatment for both psoriasis and hidradenitis, as well as DLQI score for the impact on the quality of life."
}
]
},
{
"id": "57162",
"date": "11 Dec 2019",
"name": "Georgios Nikolakis",
"expertise": [
"Reviewer Expertise HS",
"sebocytes",
"acne",
"melanoma",
"allergy"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI have prepared a PDF file with most of the points I think need to be addressed in order to make this case acceptable for indexing - please find the file here. For the second case we have really no proof, even a single photo, showing that brodalumab led to improvement of HS. Both Psoriasis and HS need to be assessed using both descriptive terms but also objective and validated scoring systems, to quantify the improvement.\nMoreover, the improvement of HS under certolizumab pegol is not clear for me, since I cannot tell that inflammatory lesions (nodules, abscesses or sinus tracts) have decreased after therapy.\nSince I believe that this case report can add to the current literature, opening ways for more anti-inflammatory treatments for HS, I think that updating the documentation accordingly and providing some proof for the improvement of HS will make the manuscript acceptable for indexing.\n\nIs the background of the cases’ history and progression described in sufficient detail? Partly\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? No\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the conclusion balanced and justified on the basis of the findings? No",
"responses": [
{
"c_id": "6182",
"date": "22 Dec 2020",
"name": "Theodora Kanni",
"role": "Author Response",
"response": "Dear Dr. Nikolakis, We revised our manuscript according to your comments. We provide severity assessment before and after treatment for both cases. The clinical improvement for psoriasis and hidradenitis is estimated using the PASI (Psoriasis Area Severity Index) score, BSA (Body Surface Area), and IHS4 (International Hidradenitis Suppurativa Severity Scoring System) score, while the impact on the quality of life is estimated with the DLQI (Dermatology Life Quality Index) score. Regarding your comment about the photographic documentation of HS improvement of the second patient, the patient denied taking photos. The location of his HS lesions is on the groin area and we did not have the patient’s consent for photographic documentation."
}
]
},
{
"id": "71707",
"date": "28 Sep 2020",
"name": "Kevin K. Wu",
"expertise": [
"Reviewer Expertise Psoriasis",
"HS",
"atopic dermatitis",
"biologics",
"epidemiology."
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThank you for these two interesting cases. One major thing missing from this paper are objective measurements of improvement after starting the respective therapies. Simply stating \"His psoriasis and psoriatic arthritis were highly improved at week 8 (Figure 2 e–h), as was HS at week 16\" or \"Treatment with CZP significantly improved psoriasis and psoriatic arthritis at week 8 and HS at week 12 (Figure 1f–i).\" does not give the reader an objective measurement of how much better the patient's disease process became following therapy. Did you measure PASI/IGA scores? Did the patients improve based on their Hurley or HISCR scores? This paper should only be accepted for indexing after including some essential, objective outcome measures. If these measures cannot be obtained, then this manuscript should be rejected for indexing.\n\nIs the background of the cases’ history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? No\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the conclusion balanced and justified on the basis of the findings? Yes",
"responses": [
{
"c_id": "6181",
"date": "22 Dec 2020",
"name": "Theodora Kanni",
"role": "Author Response",
"response": "Dear Dr. Wu, We revised our manuscript according to your comments. We provide severity assessment before and after treatment for both cases. The clinical improvement for psoriasis and hidradenitis is estimated using the PASI (Psoriasis Area Severity Index) score, BSA (Body Surface Area), and IHS4 (International Hidradenitis Suppurativa Severity Scoring System) score, while the impact on the quality of life is estimated with the DLQI (Dermatology Life Quality Index) score."
}
]
}
] | 1
|
https://f1000research.com/articles/8-2002
|
https://f1000research.com/articles/9-514/v1
|
04 Jun 20
|
{
"type": "Research Article",
"title": "Computational screening for potential drug candidates against the SARS-CoV-2 main protease",
"authors": [
"Bruno Silva Andrade",
"Preetam Ghosh",
"Debmalya Barh",
"Sandeep Tiwari",
"Raner José Santana Silva",
"Wagner Rodrigues de Assis Soares",
"Tarcisio Silva Melo",
"Andria Santos Freitas",
"Patrícia González-Grande",
"Lucas Sousa Palmeira",
"Luiz Carlos Junior Alcantara",
"Marta Giovanetti",
"Aristóteles Góes-Neto",
"Vasco Ariston de Carvalho Azevedo",
"Preetam Ghosh",
"Debmalya Barh",
"Sandeep Tiwari",
"Raner José Santana Silva",
"Wagner Rodrigues de Assis Soares",
"Tarcisio Silva Melo",
"Andria Santos Freitas",
"Patrícia González-Grande",
"Lucas Sousa Palmeira",
"Luiz Carlos Junior Alcantara",
"Marta Giovanetti",
"Aristóteles Góes-Neto",
"Vasco Ariston de Carvalho Azevedo"
],
"abstract": "Background: SARS-CoV-2 is the causal agent of the current coronavirus disease 2019 (COVID-19) pandemic. They are enveloped, positive-sense, single-stranded RNA viruses of the Coronaviridae family. Proteases of SARS-CoV-2 are necessary for viral replication, structural assembly, and pathogenicity. The approximately 33.8 kDa Mpro protease of SARS-CoV-2 is a non-human homologue and is highly conserved among several coronaviruses, indicating that Mpro could be a potential drug target for Coronaviruses. Methods: Herein, we performed computational ligand screening of four pharmacophores (OEW, remdesivir, hydroxychloroquine and N3) that are presumed to have positive effects against SARS-CoV-2 Mpro protease (6LU7), and also screened 50,000 natural compounds from the ZINC Database dataset against this protease target. Results: We found 40 pharmacophore-like structures of natural compounds from diverse chemical classes that exhibited better affinity of docking as compared to the known ligands. The 11 best selected ligands, namely ZINC1845382, ZINC1875405, ZINC2092396, ZINC2104424, ZINC44018332, ZINC2101723, ZINC2094526, ZINC2094304, ZINC2104482, ZINC3984030, and ZINC1531664, are mainly classified as beta-carboline, alkaloids, and polyflavonoids, and all displayed interactions with dyad CYS145 and HIS41 from the protease pocket in a similar way as other known ligands. Conclusions: Our results suggest that these 11 molecules could be effective against SARS-CoV-2 protease and may be subsequently tested in vitro and in vivo to develop novel drugs against this virus.",
"keywords": [
"SARS-CoV-2",
"protease",
"virtual screening",
"pharmacophore",
"inhibitors",
"natural compounds"
],
"content": "Introduction\n\nCoronaviruses (CoVs) are enveloped, positive-sense, single-stranded RNA viruses of the Coronaviridae family1. Based on their antigenic properties, they were classified into three main groups2: i) alpha-CoVs, responsible for gastrointestinal disorders; ii) beta-CoVs, which include: (a) Bat coronavirus (BCoV), (b) human severe acute respiratory syndrome (SARS) virus, (c) Middle Eastern respiratory syndrome (MERS) virus; and iii) gamma-CoVs, which mainly infect avian species. The most well-known of these coronaviruses is the SARS-CoV, responsible for causing an outbreak in 2002–20033 and MERS-CoV, causing severe respiratory symptoms, which was identified in 20124.\n\nIn December 2019, a series of unusual pneumonia cases caused by a novel coronavirus, recently renamed as SARS-CoV-2, was identified in Wuhan, China5,6. The disease caused by SARS-CoV-2 is now called COVID-19, and displays vast pathophysiological aspects, which include symptoms, such as fever and coughing, and severe acute respiratory failure7. Since the infection crossed geographical barriers, the World Health Organization (WHO) declared a pandemic situation in March 2020, reaching a worldwide mortality rate of approximately 3%6.\n\nThe SARS-CoV-2 ORF1ab code for polyprotein 1ab (pp1ab), where the main protease Mpro is found, which is similar to a key enzyme in the processing of the picornavirus family polyprotein. The protease Mpro, digests more than 11 conserved sites starting from its autolytic cleavage in pp1ab, and is a protein with extreme functional importance in the viral cycle8. Due to its great importance in the coronavirus cycle, the Mpro sequence of SARS-CoV-2 shows more than 90% similarity with the enzymes of other coronaviruses9 and shares 96% identity with SARS-CoV. Although Mpro is conserved among SARS-CoVs, it has a loop that can make it difficult for an inhibitor to access the catalytic pocket, and mutations in this loop can generate drug resistance10. Thus, even though Mpro is one of the most conserved SARS-CoV group proteins, point mutational aspects can lead to a possible drug resistance, so that a wide range of inhibitor options is necessary for the treatment of COVID-19.\n\nORF1ab is characteristic of members of the Coronaviridae family11 and is equivalent to two-thirds of the SARS-CoV-2 virus genome12. Each of these ORFs encodes a polyprotein (pp), which, when cleaved by proteases contained in the sequence, will generate 11 proteins (pp1a) and 5 proteins (pp1ab), respectively. The functions associated with these proteins are related to the virus replication processes and the modulation of the immune response in the host, among other essential functions for the development of the pathogen within the host cell6.\n\nVirus resistance to drugs can lead to the emergence of new epidemics, such as influenza A virus (IAV). In this case, two drug classes have been related: M2 channel inhibitors (amantadine and rimantadine) and neuraminidase inhibitors (NAIs; oseltamivir, zanamivir, peramivir, and Laninamivir). Both drug classes act by inhibiting proteins that are located in the viral envelope, and this region is in greater contact with the external environment and is prone to suffer from greater evolutionary pressure and, consequently, mutations13. Drug resistance can occur when rapid viral replication is not repressed completely14. In contrast, virus proteases play a crucial role during virus replication and, therefore, they are a good target for drug discovery15.\n\nDuring viral replication, proteases are necessary for the assembly of the viral structure, and there have already been suggested to have relationships with the mechanism of infection and pathogenicity of SARS-CoV-25,16. Proteases are enzymes found in all cellular organisms and viruses and are classified according to their catalytic nature. Proteases are divided into four groups: serine, cysteine, aspartyl and metalloproteases. Different types of proteases can perform the same activity through different catalytic mechanisms15, and a protease commonly has a binding site and a catalytic site that are very close in the protein structure15. Furthermore, proteases are present in several types of viruses and are widely found in human viruses17.\n\nIn coronaviruses, pp1 is essential for the replication of the virus, as it encodes the protease Mpro, which is also called the \"main protease\"18,19. Mpro is classified as a chymotrypsin-like cysteine protease (3CLpro), EC: 3.4.19.12,9,18, and the Mpro protease of SARS-CoV-2, which has a mass of approximately 33.8 kDa19, is characterized by a self-cleavage protein20,21. It consists of a homodimer subdivided into two protomers (A and B) that have three distinct domains22. The first and second domains have antiparallel β-sheets while the third domain contains five α-helices forming a globular group, which is connected in parallel with the domain-II through a loop region19. The Mpro of SARS-CoV-2 has a catalytic cleft, consisting of a Cys-His dyad in the place of the protease substrate interaction, which is situated between domains -I and -II19. It also has non-canonic specificity to the substrate in the C-terminal portion. Furthermore, there is no homologue of Mpro in the human genome19,23, and it is highly conserved amongst coronaviruses24. Therefore, Mpro is a potential target for studying inhibitors.\n\nAntiviral therapy considers three main approaches for the control and avoidance of viral infections: (a) vaccination, (b) stimulation of host resistance mechanisms, and (c) antiviral chemotherapy. Antivirals are drugs that inhibit certain virus-specific events, such as binding to host cells, which is how SARS-Cov-2 binds to ACE2 and TMPRSS225, and MERS binds to the DPP415 receptor26. Antiviral chemotherapy can involve interfering with any or all of these viral replication steps. Most antiviral drugs are primarily targeted to the synthesis of nucleic acids in viruses. As viral replication and host cell processes are closely linked, one of the main problems of viral therapy would be to find a drug capable of being selectively toxic only for the virus. Antivirals are frequently more effective in prevention than in the treatment itself, and are ineffective in eliminating latent or non-replicating viruses27. In addition, when selecting an antiviral drug, viral resistance must also be considered since it is one of the main causes of therapeutic failure.\n\nThe main classes of antiviral drugs used in clinical therapy to treat systemic viral infections include: a) synthetic nucleosides (e.g. acyclovir, famciclovir, ganciclovir, valacyclovir, and valganciclovir; b) pyrophosphate analogs (e.g. foscarnet); c) drugs for syncytial virus and influenza A (e.g. amantadine and rimantadine hydrochloride and ribavirin); d) nucleoside reverse transcriptase inhibitors (NRTI; e.g. abacavir, didanosine, emtricitabine, stavudine, lamivudine, zidovudine, tenofovir in combination with NRTI); e) non-nucleoside reverse transcriptase inhibitors (NNRTI; e.g. delavirdine, efavirenz, nevirapine); and f) protease inhibitors (e.g. amprenavir, atazanavir, darunavir, fosamprenavir, lopinavir and ritonavir, nelfinavir mesylate, saquinavir mesylate, ritonavir, indinavir sulfate and tipranavir)28,29\n\nComputational studies of inhibitors that may reduce viral replication is a fast way for proposing drug candidates that can contribute to a reduction in severity and spread of the disease. Moreover, the use of antiviral compounds can assist in the prophylaxis of SARS-CoV-2 and reduce its spread30. Therefore, screening for potential viral protease inhibitors may assist in the selection of new drugs with antiviral potential for SARS-CoV-2.\n\n\nMethods\n\nFor this study, we employed both ligand-based virtual screening (LBVS) and receptor-based virtual screening (RBVS) approaches, considering 50,000 structures of natural compounds from the ZINC Database, which has more than 900 million structures deposited, and includes millions of drug-like compounds that can be obtained for in vitro and in vivo tests31. The ZINC molecules that were downloaded were those restricted to absorption, distribution, metabolism, excretion and toxicity characteristics (ADMET) for drug likeness: no more than 5 hydrogen bond donors, no more than 10 hydrogen bond acceptors, molecular weight between 160 and 500 Daltons and logP between -0.4 and 5.632,33. For LBVS, we defined four known drugs divided in the following groups: 1) peptide-like crystallographic ligands (N3 and OEW); and 2) repurposed drugs (remdesivir (nucleoside) and hydroxychloroquine) for chemical comparison with our database.\n\nCrystallographic ligand structures were obtained from their corresponding PDB files 6LU7 (N3) and 6Y7M (OEW). Additionally, these structures were used for re-docking validations. In the LBVS process, we used a simple run with vROCS (OpenEye)34 for generating queries with the pharmacophoric map with the stereochemical characteristics for each known ligand. Another option for pharmacophore generation and searching is the free software PharmaGist35. Afterwards, we submitted each ligand query for searching similar pharmacophore-like molecules using the Tanimoto Combo algorithm36,37 with a cutoff of 1.0, which returned the best 1,000 hits for each round. This procedure was repeated three times for each query, and, subsequently, redundant structures were discarded, generating, in the end, a total of 4,000 similar molecules for the docking experiment.\n\nConsidering PDB validation indices as crystallographic resolution, Ramachandran outliers, clash score, and release date, we selected the structure 6LU7 for RBVS, which is complexed with the peptide-like inhibitor N319. Furthermore, 6LU7 and 6Y7M38 were used for re-docking validations with its corresponding crystallographic inhibitors.\n\nThe best LBVS hits were submitted to molecular docking calculations with 6LU7 structure using Autodock 4.2 virtual screening protocol39. Ligand structures were prepared for virtual screening using Raccoon plugin40 for Autodock Tools40 according to the standard protocol37, as well as the 6LU7 structure. The gridbox was defined on the active site region, considering the amino acids THR 190, GLU 166, GLN 189, GLY 143, HIS 163, HIS 164, CYS 145, PHE 140, and with accordance with previous studies with the crystallographic structure of the SARS-CoV-2 main protease19,38,41. Each docking run was performed three times using the following specifications: flexible docking and Lamarckian Genetic Algorithm with 2,500,000 generations. Afterwards, the 10 best docking hits were selected using the Autodock Tools script summarize_results4.py, which can classify the best hits according to their lowest energy clustering conformations and root mean square deviation (RMSD) values. The results were organized according to the ligand pharmacophore relationship with the known structures in Table 2. Docking and re-docking results were evaluated at each docking position inside the 6LU7 active site using Pymol 2.142 and UCSF Chimera 1.1443 in order to confirm molecule interactions with the amino acids within the protease active site. Furthermore, 2D interaction maps were generated by Discovery Studio 201944. Another option for 2D map generation is the LigPlot+ software45.\n\n\nResults\n\nDifferent pharmacophoric characteristics were generated for each known ligand (Figure 1), which allowed us to find molecules included in different chemical classifications and amplify the number of possible drug candidates. Table 1 shows the pharmacophoric characteristics for each known 6LU7 inhibitor, which permitted us to find natural ligands with pharmacophore-like regions. Additionally, we used the ADMET characteristics for molecular weight and LogP that are important for molecular druggability.\n\n(A) OEW, (B) N3, (C) remdesivir and (D) hydroxychloroquine. In red spheres: hydrogen acceptors; blue spheres: hydrogen donors; yellow spheres: hydrophobic; and green spheres: aromatic.\n\nHb.A. = hydrogen acceptor; Hb.D. = hydrogen donor; M.W. = molecular weight.\n\nLigand based virtual screening and docking calculations of ZINC database compounds revealed the 40 best pharmacophore-like ligands that belong to different chemical classes, namely beta-carboline alkaloids, indole alkaloids, lupin alkaloids, harmala alkaloids, polyflavonoids, anthracenes, angular pyranocoumarins, and flavonoid-3-O-glycosides. Table 2 shows the detailed results on the average affinity energies, ZINC identification, and chemical classification of each selected ligand.\n\n*Re-docked crystallographic structures.\n\n**Repeated ligand between OEW and Remdesivir pharmacophores.\n\nFor selecting the best pharmacophore-like drug-candidates, we considered evaluating lower affinity energy values, as well as interactions with residues of the active site within the target. As can be seen in Figure 2A, all pharmacophore-like OEW ligand molecules formed a complex with the active pocket of 6LU7. The three best OEW ligands (ZINC1845382, ZINC1875405, ZINC2092396) are shown in complex with COVID-19 protease in Figure 2B–D with the detailed 2D interaction map. In this case, these top three hits are included in the beta-carboline alkaloid class.\n\n(A) SARS-CoV-2 main protease complexed with the 10 best hits for OEW pharmacophore molecules. Protomer A is represented in marine blue surface and protomer B in dark pink surface. ZINC1845382 in cyan (B), ZINC1875405 in dark pink (C) and ZINC2092396 in purple (D) inside 6LU7 binding site and their 2D interaction maps with pocket amino acids.\n\nThe intermolecular interactions carried out by ligand ZINC1845382 exhibited a hydrogen bond with the residue of the active PHE140 protease site. The catalytic residues CYS145 and HIS41 represented interactions of the type π, π-π stacked and π-alkyl with the entire beta-carboline group, which was composed of three hydrophobic rings. The remaining residues were of the π-sigma type, hydrogen-carbon acceptors, and halogen acceptors from residues THR25, THR26, as well as other residues from the active site GLU166, GLN189, GLY143, HYS164, respectively.\n\nLigand ZINC1875405 represented two hydrogen bonding interactions with residues THR25 and PHE140. Additionally, four more polar interactions of the type π-π stacked, π-aquil, aquil and π-sulfur with residues HIS41, MET49, CYS145 and MET165, respectively, were formed. The other interactions were of hydrophobic van der Waals type.\n\nLigand ZINC2092396 interacted by hydrogen interaction with the residue PHE140, π and π-alkyl with CYS140, π-π stacked and π-alkyl HIS41, and van der Waals with GLN189, GLY143, HIS164, GLU166. Other interactions occurred with hydrogen bonds by the ligand nitrobenzene group with the ASN142 residue and a π-sulfur interaction of the beta-carboline group with MET165 residue.\n\nThe remdesivir pharmacophore-like search returned two beta-carboline alkaloids (ZINC2104424 and ZINC1875405), as well as one polyflavonoid (ZINC44018332), which interacted with the COVID-19 main protease active pocket showing affinity energies below -10.0 kcal/Mol. Figure 3A–D shows the details of all ligand interactions, as well as the top three molecules interaction maps.\n\n(A) SARS-CoV-2 main protease complexed with 10 best hits for remdesivir pharmacophore molecules. Protomer A is represented in green surface, and protomer B in orange surface. ZINC2104424 in cyan (B), ZINC1875405 in wheat (C) and ZINC44018332 in violet (D) inside 6LU7 binding site and their 2D interaction maps with pocket amino acids.\n\nLigand ZINC2104424 also occupied the region of the active site (Figure 3B), showing polar interactions π, π-alkyl, π-π stacked and π-sulfur types from beta-carboline with HIS41, MET49, CYS145, and MET165 amino acids. Moreover, an interaction of THR26 halogen with the ligand fluorobenzene group also occurred. Other hydrophobic interactions were van der Waals, mostly with residues of the active site: PHE140, GLY143, HIS163, HIS164, GLU166 and GLN189.\n\nLigand ZINC1875405 (Figure 3C) displayed three hydrogen interactions with the indole group, and two oxygen interactions from a nitrobenzene of THR25, PHE140 AND GLN166, respectively. Several van der Waals-type hydrophobic interactions were found with GLY143, HIS164 and GLN189 amino acids. Furthermore, four polar interactions (π, π-alkyl, π-π stacked and π-sulfur) with residues HIS41, MET49, CYS 145 and MET165, respectively, were also retrieved.\n\nLigand ZINC2092396 (Figure 3D) exhibited two hydrogen interactions with HIS163 and THR26 by its hydroxyl from the flavonoid nucleus, as well as four more π-donor hydrogen bonding interactions with residues TYR54, PHE140, GLY143 and GLU166. Besides, three π-alkyl and π-sulfur interactions made with MET49, CYS145 and MET165 were also retrieved. Other hydrophobic interactions were of van der Waals type.\n\nFigure 4 shows the interactions between 6LU7 active sites and the three best hits from derived molecules of hydroxychloroquine pharmacophore (ZINC2101723, ZINC2094526, ZINC2094304). These complexes displayed affinity energies varying from -10.2 kcal/Mol to -9.6 kcal/Mol, and all the ligands were classified as beta-carboline alkaloid derivatives.\n\n(A) Best hits for hydroxychloroquine pharmacophore. Protomer A is represented using a violet surface, and protomer B in marine blue surface. ZINC2101723 in yellow (B), ZINC2094526 in red (C), and ZINC2094304 in dark blue (D) inside 6LU7 binding site and their 2D interaction maps with pocket amino acids.\n\nThe beta-carboline group of the ligand ZINC2101723 (Figure 4B) formed four π-alkyl, alkyl and π-sulfur type interactions with HIS41, MET49, CYS145 and MET165 residues, as well as other hydrophobic interactions from its naphthalene and beta-carboline groups with the active site amino acids PHE140, GLY143, HIS163 E164, GLU166 and GLN189. Ligand ZINC2094526 (Figure 4C) displayed a hydrogen bond interaction with PHE140 by its nitrobenzene group. Five polar interactions (π-sigma, π-aquil, π-π stacked and π-sulfur) were observed with residues THR25, HIS41, MET49, CYS145 and MET165. For ligand ZINC2094304 (Figure 4D), two hydrogen bonds with residues PHE140 and GLU166 by its nitrobenzene group were formed. In addition, this ligand formed four polar interactions (π-π stacked, π-alkyl, alkyl and π-sulfur) with residues HIS41, MET49, CYS145 and MET165, respectively. Other van der Waals type interactions could also be identified.\n\nThe N3 pharmacophore revealed one beta-carboline alkaloid (ZINC2101723) and two polyflavonoids (ZINC2094526 and ZINC2094304). This group displayed affinity energies ranging from -9.8 kcal/Mol to -10.1 kcal/Mol. In Figure 5, the best complex interactions with the protease, as well as their positions inside the binding pocket are depicted.\n\n(A) COVID-19 main protease is represented in cyan (protomer A) and dark salmon (protomer B). The best complexes are formed by the alkaloid ZINC2101723 in pink (B) and two polyflavonoids ZINC2094526 in marine blue (C) and ZINC2094304 in lemon green (D), and their 2D interaction maps with pocket amino acids are shown below each complex.\n\nLigand ZINC2104482 (Figure 5B) formed a large number of hydrophobic interactions (14 van der Waals interactions), surrounding the active site amino acid, such as GLY143, HIS164, GLU166 and GLN189. Furthermore, this ligand formed three π-alkyl and alkyl bonds with HIS41, MET49, CYS145 residues. Ligand ZINC3984030 (Figure 5C) exhibited three hydrogen bonds with THR26, TYR54 and GLU166 residues by OH groups of flavonoid nuclei. A π-donor hydrogen bond interaction of the GLY143 residue was also observed. Moreover, three polar interactions (π-π stacked, π-alkyl and π-sulfur) were identified with HIS41, CYS145 and MET165. The rest of the interactions were van der Waals type. Ligand ZINC1531664 (Figure 5D) showed a hydrogen bond by its OH group TYR54. Additionally, four π-donor hydrogen bond and hydrogen carbon bond interactions with residues PHE140, GLY143, GLU166 also occurred. Two polar interactions of the type π-alkyl and π-sulfur were observed with MET49, CY145, and MET165, and the other hydrophobic interactions were of van der Waals type.\n\nCrystallographic ligands N3 and OWE were re-docked with their respective Mpro structures 6LU7 and 6Y7M. As can be seen in Figure 6A, both N3 and OEW molecules bound into similar positions in comparison to their original crystallographic forms. The Figure 6B depicts the best clustering conformations graph for N3 with a free energy of binding ranging from -1.83 kcal/Mol to -9.7 kcal/Mol. Figure 6C shows the superposition between the crystalized and re-docked N3 structure. Even though N3 is peptide-like with 13 routable bonds, it presented an RMSD of 1.94 Angstroms for its best conformation (Table 2). OEW re-docking is shown in Figure 6D in the same way as for N3 where both the crystallized and docked structures bound into the same pocket. Conformational population of OEW clustering results returned a free energy of binding ranging from -7.0 kcal/Mol to -11.5 kcal/Mol but, on the other hand, the structure with binding energy of -8.86 kcal/Mol exhibited the smallest RMSD value (Figure 6E). Additionally, OEW presented an RMSD of 1.97 Å in comparison to its crystallized form (Figure 6F).\n\n(A) crystallographic (hot pink) and re-docked (green) N3 inhibitor of the 6LU7 SARS-CoV-2 Mpro inside its binding pocket; (B) N3 docking best clustering conformations; (C) aligned N3 crystallized (yellow) and re-docked (cyan). (D) crystallographic (pink) and re-docked (purple) OEW inhibitor of the 6LU7 SARS-CoV-2 Mpro inside its binding pocket; (E) OEW docking best clustering conformations; (F) aligned OEW crystallized (green) and re-docked (red).\n\n\nDiscussion\n\nDocking results revealed 39 pharmacophore-like natural ligands, which can be used as drug candidates for inhibiting SARS-CoV-2 main protease activity. Furthermore, we ranked the three best candidates for each known ligand pharmacophore as the best potential drug molecules (and totaling 12 molecules) for in vitro and in vivo assays purposes, but not excluding the other 28 molecules. For these cases, ligands are included in two most expressive chemical classes: β-carboline alkaloids and polyflavonoids. Additionally, all ligands exhibiting better affinity energies than the known drugs was used as references for construction of pharmacophoric characteristics: OEW6, remdesivir46, hydroxychloroquine47, and N319.\n\nThe groups of OEW and hydroxychloroquine pharmacophores presented their three most promising ligands classified as β-carboline alkaloids. This class of molecules is reported by different authors with antiviral activities. According to Gonzalez et al.43, β-carboline Alkaloids are widely distributed in nature, and its derivatives exhibited activity against Herpes Simplex Viruses by blocking virus replication. Additionally, Gonzalez et al.43 demonstrated the action of these alkaloids in dengue virus RNA replication. Furthermore, several other studies suggest alkaloid activity against viral proteases48–50. Similarly, remdesivir pharmacophore revealed two β-carboline alkaloids (ZINC2104424 and ZINC1875405). In addition, we detected a polyflavonoid (ZINC44018332) as one probable active molecule from a different class against SARS-CoV-2 main protease, and several authors have already described flavonoid activity as viral protease inhibitors51–53, as well as antiviral molecules acting in different target classes52,54–56. N3 pharmacophore displayed two flavonoids as the best molecules and just one β-carboline alkaloid. These results indicate that both classes of molecules could be explored for in vitro and in vivo tests to evaluate their potential antiviral activities for not only SARS-CoV-2 but also for other viruses of medical interest.\n\nOther classes of molecules were found in our screening for protease activity that were previously described in antiviral studies: anthracenes57, angular pyranocoumarin58,59, and flavonoid-3-O-glycoside60. Interaction maps of these complexes are available as Extended data61.\n\nAll the known ligands (OEW, remdesivir, hydroxychloroquine and N3), which were used for validating our computational screening, exhibited worse affinity energies in docking calculations (ranging from -7.8 kcal/Mol to -5.2 kcal/Mol) than the screened natural compounds (ranging from -10.6 kcal/Mol to -9.1 kcal/Mol). Moreover, all the 40 selected ligands docked inside Mpro active site, as previously described in several antiviral studies, and interacted in the region of connection between domains I and II with amino acids HIS41 and CYS14518,19,22,38,41,62.\n\nNovel Mpro ketoamide inhibitors were recently proposed, including the OEW ligand (ligand 13b) that was used in our study, and the authors detected a reduction in RNA replication in human cells infected with SARS-CoV-2, and also described binding interactions with its main protease. Besides, the same study indicated a ketoamide as a probable drug candidate against this virus38.\n\nIn a recent study, authors have proposed the peptidomimetic molecule N3 as a drug candidate against COVID-19, and described its binding interactions with the crystallographic structures of SARS-CoV-2 and other viral proteases. Their study reported that N3 can bind in all the active pockets from the main proteases of HCoV-NL63, SARS-CoV, and MERS-CoV63.\n\nOther molecules have also been tested as antivirals for effectiveness in inhibiting SARS-CoV-2 replication in cell culture. Two drugs exhibited a promising inhibitory effect: remdesivir, an experimental drug developed for the treatment of Ebola virus infection41,64, and hydroxychloroquine, a drug known for its effectiveness in the treatment of malaria and autoimmune diseases41. Remdesivir is an adenosine triphosphate analogue initially described in the literature in 2016 as a potential treatment for Ebola65, and this drug has been indeed considered as a potential treatment for SARS-CoV2,64. Notably, remdesivir has demonstrated antiviral activity in the treatment of MERS and SARS66 in animal models, both of which are caused by coronaviruses67. Pharmacophore models are widely used in medicinal chemistry with the aim of amplifying the number of drug candidates, and according to this definition, they are represented by a 3D arrangement of abstract features instead of chemical groups68. Remdesivir is a nucleotide analogue with capacity to inhibit RNA polymerase (Table 1): this molecule displayed almost the same pharmacophoric features (Figure 1) as for N3 and OEW, and, besides, both of them have already tested experimentally. Additionally, as can be seen in Table 2, the best hit is ZINC1875405, which was found in both OEW and remdesivir pharmacophore searching, and this could be explained by their similar characteristics. Hydroxychloroquine is an aminoquinoline-like chloroquine69. It is a drug commonly prescribed for the treatment of uncomplicated malaria, rheumatoid arthritis, chronic discoid lupus erythematosus, and systemic lupus erythematosus70. Chloroquine and hydroxychloroquine have been investigated for the treatment of SARS-CoV-271, and they have been reported to have direct antiviral effects, such as inhibition of flaviviruses, retroviruses (like HIV), and many coronaviruses. Additionally, hydroxychloroquine is capable of inhibiting the zika virus NS2B-NS3 protease, and exhibited good viral replication blocking in infected JEG3 cells in concentration of 80 µM of hydroxychloroquine72. Furthermore, the use of chloroquine and its analogues can be corroborated by a recent study showing that, with EC50 of 1.13 µmol/L and selectivity index (SI) greater than 88, chloroquine can effectively inhibit SARS-CoV-2 at the cellular level73. Its effectiveness in the human body for SARS-CoV-2 infection; however, has not yet been clinically proven. Another in silico study with chloroquine detected its interactions with viral NSP-3B type protease74.\n\nThe co-crystallized molecules N3 and OEW are both peptide analogs, which presented RMSD values of 1.94 Å and 1.97 Å in re-docking experiments, respectively. Generally, docking validations protocols use co-crystallized ligands, to test the accuracy of the program to predict the correct ligand docking poses in comparison to known conformations, and its RMSD varies 1.5 or 2 Å depending on ligand size for being considering acceptable75. The number of studies using protein-peptide docking has been increasing rapidly, followed by the number of applied drug design programs and models. On the other hand, the use of RMSD validations with experimental structures is not always the best criterion of docking success, once it can be influenced by resolution quality, as well as the number of peptide residues76. Thus, we can consider that N3 and OEW docking validations are in acceptable RMSD ranges.\n\n\nConclusions\n\nIn our study, we compared the pharmacophores of four well-tested human coronavirus (including SARS-Cov-2) main protease drug candidates to 50,000 structures of natural compounds from the ZINC Database. The three best molecules selected for each pharmacophore class are mainly classified as β-carboline alkaloids, and polyflavonoids. The best ligand-SARS-CoV-2 complexes exhibited better affinity energies in comparison to drug molecules used in this study. Furthermore, all the screened molecules bonded between domains -I and -II and formed interactions with the catalytic residues HIS41 and CYS145 in similar positions as previously described from other authors in viral protease inhibitor studies. Altogether, we propose these compounds as possible SARS-CoV-2 protease inhibitors, which can be used for subsequent in vitro and in vivo tests for finding novel drug candidates.\n\n\nData availability\n\nStructures of natural compounds were downloaded from the ZINC Database.\n\nCrystal structures of COVID-19 main protease were downloaded from the Protein Data Bank, accession numbers 6LU7 (in complex with N3) and 6Y7M (with OEW).\n\nHarvard Dataverse: Replication Data for: Computational screening for potential drug candidates against SARS-CoV-2 main protease. https://doi.org/10.7910/DVN/GYFXA061.\n\nThis project contains the following extended data:\n\n2D interaction maps of all OEW pharmacophore-like ligands (PNG).\n\n2D interaction maps of all Remdesivir pharmacophore-like ligands (PNG).\n\n2D interaction maps of all Hydroxychloroquine pharmacophore-like ligands (PNG).\n\n2D interaction maps of all N3 pharmacophore-like ligands (PNG).\n\nExtended data are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).",
"appendix": "Acknowledgements\n\nWe would like to thank The OpenEye Science for the OpenEye Software license, which made possible the ligand based virtual screening experiments.\n\nThis research was developed with the help of the Núcleo de Biologia Computacional e Gestão de Informações Biotecnológicas- NBCGIB \", with resources from FINEP / MCT, CNPQ and FAPESB and from Universdade Estadual de Santa Cruz – UESC, represented by Dr. Eduardo Almeida Costa. Additionally, the authors would like to acknowledge the SENAI CIMATEC Center for Industrial Innovation, with support from BG Brazil and the Brazilian Authority for Oil, Gas and Biofuels (ANP), for the provision and operation of computational facilities and the commitment to invest in Research & Development.\n\n\nReferences\n\nCui J, Li F, Shi ZL: Origin and evolution of pathogenic coronaviruses. Nat Rev Microbiol. 2019; 17(3): 181–192. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchoeman D, Fielding BC: Coronavirus envelope protein: current knowledge. Virol J. 2019; 16(1): 69. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPeiris JS, Guan Y, Yuen KY: Severe acute respiratory syndrome. Nat Med. 2004; 10(12 Suppl): S88–97. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZaki AM, Van Boheemen S, Bestebroer TM, et al.: Isolation of a novel coronavirus from a man with pneumonia in Saudi Arabia. N Engl J Med. 2012; 367(19): 1814–20. PubMed Abstract | Publisher Full Text\n\nBenvenuto D, Giovanetti M, Ciccozzi A, et al.: The 2019-new coronavirus epidemic: Evidence for virus evolution. J Med Virol. 2020; 92(4): 455–459. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWu F, Zhao S, Yu B, et al.: A new coronavirus associated with human respiratory disease in China. Nature. 2020; 579(7798): 265–269. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZheng M, Gao Y, Wang G, et al.: Functional exhaustion of antiviral lymphocytes in COVID-19 patients. Cell Mol Immunol. 2020; 17(5): 533–535. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJin Z, Du X, Xu Y, et al.: Structure of Mpro from COVID-19 virus and discovery of its inhibitors. Nature. 2020. PubMed Abstract | Publisher Full Text\n\nBzowka M, Mitusinska K, Raczynska A, et al.: Molecular Dynamics Simulations Indicate the COVID-19 Mpro Is Not a Viable Target for Small-Molecule Inhibitors Design. bioRxiv. 2020. Publisher Full Text\n\nBzówka M, Mitusińska K, Raczyńska A, et al.: Structural and Evolutionary Analysis Indicate That the SARS-CoV-2 Mpro Is a Challenging Target for Small-Molecule Inhibitor Design. Int J Mol Sci. 2020; 21(9): pii: E3099. PubMed Abstract | Publisher Full Text\n\nCheng VC, Lau SK, Woo PC, et al.: Severe acute respiratory syndrome coronavirus as an agent of emerging and reemerging infection. Clin Microbiol Rev. 2007; 20(4): 660–94. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGuo YR, Cao QD, Hong ZS, et al.: The origin, transmission and clinical therapies on coronavirus disease 2019 (COVID-19) outbreak - an update on the status. Mil Med Res. 2020; 7(1): 11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHussain M, Galvin HD, Haw TY, et al.: Drug resistance in influenza A virus: the epidemiology and management. Infect Drug Resist. 2017; 10: 121–134. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcKeegan KS, Borges-Walmsley MI, Walmsley AR: Microbial and viral drug resistance mechanisms. Trends Microbiol. 2002; 10(10 Suppl): S8–14. PubMed Abstract | Publisher Full Text\n\nSharma A, Gupta SP: Fundamentals of viruses and their proteases. In: Viral Proteases and Their Inhibitors. 2017; 1–24. Publisher Full Text | Free Full Text\n\nZhang T, Wu Q, Zhang Z: Probable Pangolin Origin of SARS-CoV-2 Associated with the COVID-19 Outbreak. Curr Biol. 2020; 30(7): 1346–1351.e2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPatel S: A critical review on serine protease: Key immune manipulator and pathology mediator. Allergol Immunopathol (Madr). 2017; 45(6): 579–591. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAnand K, Palm GJ, Mesters JR, et al.: Structure of coronavirus main proteinase reveals combination of a chymotrypsin fold with an extra alpha-helical domain. EMBO J. 2002; 21(13): 3213–24. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJin Z, Du X, Xu Y, et al.: Structure of Mpro from COVID-19 virus and discovery of its inhibitors. bioRxiv. 2020. Publisher Full Text\n\nMeng T, Cao H, Zhang H, et al.: The insert sequence in SARS-CoV-2 enhances spike protein cleavage by TMPRSS. bioRxiv. 2020. Publisher Full Text\n\nKang S, Yang M, Hong Z, et al.: Crystal structure of SARS-CoV-2 nucleocapsid protein RNA binding domain reveals potential unique drug targeting sites. bioRxiv. 2020. Publisher Full Text\n\nYang H, Yang M, Ding Y, et al.: The crystal structures of severe acute respiratory syndrome virus main protease and its complex with an inhibitor. Proc Natl Acad Sci U S A. 2003; 100(23): 13190–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang P, Gao Q, Wang T, et al.: Evaluation of recombinant nucleocapsid and spike proteins for serological diagnosis of novel coronavirus disease 2019 (COVID-19). medRxiv. 2020. Publisher Full Text\n\nXu X, Chen P, Wang J, et al.: Evolution of the novel coronavirus from the ongoing Wuhan outbreak and modeling of its spike protein for risk of human transmission. Sci China Life Sci. 2020; 63(3): 457–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHoffmann M, Kleine-Weber H, Schroeder S, et al.: SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor. Cell. 2020; 181(2): 271–280.e8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZumla A, Chan JF, Azhar EI, et al.: Coronaviruses - drug discovery and therapeutic options. Nat Rev Drug Discov. 2016; 15(5): 327–47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCrumpacker CS: Use of Antiviral Drugs to Prevent Herpesvirus Transmission. N Engl J Med. 2004; 350(1): 67–8. PubMed Abstract | Publisher Full Text\n\nTer Heine R, Mulder JW, Van Gorp EC, et al.: Intracellular and plasma steady-state pharmacokinetics of raltegravir, darunavir, etravirine and ritonavir in heavily pre-treated HIV-infected patients. Br J Clin Pharmacol. 2010; 69(5): 475–83. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBalayan T, Horvath H, Rutherford GW: Ritonavir-Boosted Darunavir Plus Two Nucleoside Reverse Transcriptase Inhibitors versus Other Regimens for Initial Antiretroviral Therapy for People with HIV Infection: A Systematic Review. AIDS Res Treat. 2017; 2017: 2345617. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMitjà O, Clotet B: Use of antiviral drugs to reduce COVID-19 transmission. Lancet Glob Health. 2020; 8(5): e639–e640. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIrwin JJ, Sterling T, Mysinger MM, et al.: ZINC: A free tool to discover chemistry for biology. J Chem Inf Model. 2012; 52(7): 1757–68. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGhose AK, Viswanadhan VN, Wendoloski JJ: A knowledge-based approach in designing combinatorial or medicinal chemistry libraries for drug discovery. 1. A qualitative and quantitative characterization of known drug databases. J Comb Chem. 1999; 1(1): 55–68. PubMed Abstract | Publisher Full Text\n\nLipinski CA, Lombardo F, Dominy BW, et al.: Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv Drug Deliv Rev. 2001; 46(1–3): 3–26. PubMed Abstract | Publisher Full Text\n\nHawkins PCD, Skillman AG, Nicholls A: Comparison of shape-matching and docking as virtual screening tools. J Med Chem. 2007; 50(1): 74–82. PubMed Abstract | Publisher Full Text\n\nSchneidman-Duhovny D, Dror O, Inbar Y, et al.: PharmaGist: a webserver for ligand-based pharmacophore detection. Nucleic Acids Res. 2008; 36(Web Server issue): W223–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRácz A, Bajusz D, Héberger K: Life beyond the Tanimoto coefficient: Similarity measures for interaction fingerprints. J Cheminform. 2018; 10(1): 48. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBajusz D, Rácz A, Héberger K: Why is Tanimoto index an appropriate choice for fingerprint-based similarity calculations? J Cheminform. 2015; 7: 20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang L, Lin D, Sun X, et al.: Crystal structure of SARS-CoV-2 main protease provides a basis for design of improved α-ketoamide inhibitors. Science. 2020; 368(6489): 409–412. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorris GM, Dallakyan S: AutoDock — AutoDock.2013. Reference Source\n\nForli S, Huey R, Pique ME, et al.: Computational protein-ligand docking and virtual drug screening with the AutoDock suite. Nat Protoc. 2016; 11(5): 905–919. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang M, Cao R, Zhang L, et al.: Remdesivir and chloroquine effectively inhibit the recently emerged novel coronavirus (2019-nCoV) in vitro. Cell Res. 2020; 30(3): 269–71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchrödinger L: The PyMol Molecular Graphics System, Versión 1.8. Thomas Holder. 2015. Reference Source\n\nPettersen EF, Goddard TD, Huang CC, et al.: UCSF Chimera - A visualization system for exploratory research and analysis. J Compu Chem. 2004; 25(13): 1605–12. PubMed Abstract | Publisher Full Text\n\nDiscovery Studio: Dassault Systemes BIOVIA, Discovery Studio Modelling Environment, Release 4.5. Accelrys Softw Inc. 2015. Reference Source\n\nLaskowski RA, Swindells MB: LigPlot+: Multiple ligand-protein interaction diagrams for drug discovery. J Chem Inf Model. 2011; 51(10): 2778–2786. Publisher Full Text\n\nMartinez MA: Compounds with Therapeutic Potential against Novel Respiratory 2019 Coronavirus. Antimicrob Agents Chemother. 2020; 64(5): pii: e00399–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGonzalez MM, Cabrerizo FM, Baiker A, et al.: β-Carboline derivatives as novel antivirals for herpes simplex virus. Int J Antimicrob Agents. 2018; 52(4): 459–68. PubMed Abstract | Publisher Full Text\n\nAhmad I, Fatima I, Afza N, et al.: Urease and serine protease inhibitory alkaloids from Isatis tinctoria. J Enzyme Inhib Med Chem. 2008; 23(6): 918–21. PubMed Abstract | Publisher Full Text\n\nul Qamar MT, Mumtaz A, Ashfaq UA, et al.: Potential of plant alkaloids as dengue ns3 protease inhibitors: Molecular docking and simulation approach. Bangladesh J Pharmacol. 2014; 9(3): 262–7. Publisher Full Text\n\nPowers CN, Setzer WN: An In-Silico Investigation of Phytochemicals as Antiviral Agents Against Dengue Fever. Comb Chem High Throughput Screen. 2016; 19(7): 516536. Publisher Full Text\n\nQamar MT, Ashfaq UA, Tusleem K, et al.: In-silico identification and evaluation of plant flavonoids as dengue NS2B/NS3 protease inhibitors using molecular docking and simulation approach. Pak J Pharm Sci. 2017; 30(6): 2119–37. PubMed Abstract\n\nShimizu JF, Lima CS, Pereira CM, et al.: Flavonoids from Pterogyne nitens Inhibit Hepatitis C Virus Entry. Sci Rep. 2017; 7(1): 16127. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHawas UW, Abou El-Kassem LT, Shaher F, et al.: In vitro inhibition of Hepatitis C virus protease and antioxidant by flavonoid glycosides from the Saudi costal plant Sarcocornia fruticosa. Nat Prod Res. 2019; 33(23): 3364–3371. PubMed Abstract | Publisher Full Text\n\nKaul TN, Middleton E Jr, Ogra PL: Antiviral effect of flavonoids on human viruses. J Med Virol. 1985; 15(1): 71–9. PubMed Abstract | Publisher Full Text\n\nGonzález-Búrquez MD, González-Díaz FR, García-Tovar CG, et al.: Comparison between in Vitro Antiviral Effect of Mexican Propolis and Three Commercial Flavonoids against Canine Distemper Virus. Evidence-based Complement Altern Med. 2018; 2018; 7092416. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDai W, Bi J, Li F, et al.: Antiviral Efficacy of Flavonoids against Enterovirus 71 Infection in vitro and in Newborn Mice. Viruses. 2019; 11(7): pii: E625. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTomlinson SM, Watowich SJ: Anthracene-based inhibitors of dengue virus NS2B-NS3 protease. Antiviral Res. 2011; 89(2): 127–35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarnard DL, Xu Z, Stowell VD, et al.: Coumarins and pyranocoumarins, potential novel pharmacophores for inhibition of measles virus replication. Antivir Chem Chemother. 2002; 13(1): 39–59. PubMed Abstract | Publisher Full Text\n\nHassan MZ, Osman H, Ali MA, et al.: Therapeutic potential of coumarins as antiviral agents. Eur J Med Chem. 2016; 123: 236–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBehbahani M, Sayedipour S, Pourazar A, et al.: In vitro anti-HIV-1 activities of kaempferol and kaempferol-7-O-glucoside isolated from Securigera securidaca. Res Pharm Sci. 2014; 9(6): 463–9. PubMed Abstract | Free Full Text\n\nAndrade B: \"Replication Data for: Computational screening for potential drug candidates against SARS-CoV-2 main protease\". Harvard Dataverse, V1. 2020. http://doi.org/10.7910/DVN/GYFXA0\n\nRen Z, Yan L, Zhang N, et al.: The newly emerged SARS-Like coronavirus HCoV-EMC also has an “Achilles’’ heel\": Current effective inhibitor targeting a 3C-like protease”. Protein Cell. 2013; 4(4): 248–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDai W, Zhang B, Su H, et al.: Structure-based design of antiviral drug candidates targeting the SARS-CoV-2 main protease. Science. 2020; pii: eabb4489. PubMed Abstract | Publisher Full Text | Free Full Text\n\nde Wit E, Feldmann F, Cronin J, et al.: Prophylactic and therapeutic remdesivir (GS-5734) treatment in the rhesus macaque model of MERS-CoV infection. Proc Natl Acad Sci U S A. 2020; 117(12): 6771–6776. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWarren TK, Jordan R, Lo MK, et al.: Therapeutic efficacy of the small molecule GS-5734 against Ebola virus in rhesus monkeys. Nature. 2016; 531(7594): 381–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGordon CJ, Tchesnokov EP, Feng JY, et al.: The antiviral compound remdesivir potently inhibits RNA-dependent RNA polymerase from Middle East respiratory syndrome coronavirus. J Biol Chem. 2020; 295(15): 4773–4779. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSheahan TP, Sims AC, Graham RL, et al.: Broad-spectrum antiviral GS-5734 inhibits both epidemic and zoonotic coronaviruses. Sci Transl Med. 2017; 9(396): pii: eaal3653. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKaserer T, Beck KR, Akram M, et al.: Pharmacophore models and pharmacophore-based virtual screening: Concepts and applications exemplified on hydroxysteroid dehydrogenases. Molecules. 2015; 20(12): 22799–832. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFurst DE: Pharmacokinetics of hydroxychloroquine and chloroquine during treatment of rheumatic diseases. Lupus. 1996; 5 Suppl 1: S11–5. PubMed Abstract | Publisher Full Text\n\nShippey EA, Wagler VD, Collamer AN: Hydroxychloroquine: An old drug with new relevance. Cleve Clin J Med. 2018; 85(6): 459–467. PubMed Abstract | Publisher Full Text\n\nDevaux CA, Rolain JM, Colson P, et al.: New insights on the antiviral effects of chloroquine against coronavirus: what to expect for COVID-19? Int J Antimicrob Agents. 2020; 105938.PubMed Abstract | Publisher Full Text | Free Full Text\n\nKumar A, Liang B, Aarthy M, et al.: Hydroxychloroquine Inhibits Zika Virus NS2B-NS3 Protease. ACS Omega. 2018; 3(12): 18132–41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang Q, Zhang Y, Wu L, et al.: Structural and Functional Basis of SARS-CoV-2 Entry by Using Human ACE2. Cell. 2020; pii: S0092-8674(20)30338-X. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWu C, Liu Y, Yang Y, et al.: Analysis of therapeutic targets for SARS-CoV-2 and discovery of potential drugs by computational methods. Acta Pharm Sin B. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHevener KE, Zhao W, Ball DM, et al.: Validation of molecular docking programs for virtual screening against dihydropteroate synthase. J Chem Inf Model. 2009; 49(2): 444–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCiemny M, Kurcinski M, Kamel K, et al.: Protein-peptide docking: opportunities and challenges. Drug Discov Today. 2018; 23(8): 1530–7. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "75868",
"date": "09 Dec 2020",
"name": "Murtaza Tambuwala",
"expertise": [
"Reviewer Expertise Inflammation",
"cancer",
"and infections research"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\n\"Computational screening for potential drug candidates against the SARS-CoV-2 main protease\" is overall a timely interesting review focused on COVID pandemic.\nHowever, since the literature on COVID research is being updated so rapidly this review seems to have used older references and some latest and relevant papers have been missed out.\n\nBelow I have provided suggestions for several such manuscripts that need to be included and discussed:\nThe Structural Basis of Accelerated Host Cell Entry by SARS-CoV-21. The Importance of Research on the Origin of SARS-CoV-22. Questions concerning the proximal origin of SARS‐CoV‐23. Probing 3CL protease: Rationally designed chemical moieties for COVID-194. COVID-19 pandemic: an overview of epidemiology, pathogenesis, diagnostics and potential vaccines and therapeutics5. COVID-19: Underpinning Research for Detection, Therapeutics, and Vaccines Development6.\nI am confident that this review will attract a good level of readership from both the public can the research community.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "75869",
"date": "09 Dec 2020",
"name": "Alaa A A Aljabali",
"expertise": [
"Reviewer Expertise viral nanotecnology",
"biotechnology and drug delviery"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript is well written, balanced, and follows the scientific logic and flow very nicely. The work is very novel and holds great potential. The approach is very useful and holds great potential in using the docking system to maybe identify new antiviral candidates in this pandemic.\nSome minor comments:\nPlease list the software used for generating the docking work.\n\nCould you please improve the figure axis of 6 bars for clarity (B and E)?\n\nCould you please add a short paragraph on the novelty of this approach of molecular docking?\n\nCould you please indicate if any of the reported drugs have been used in clinical trials originated from the docking approach?\n\nWould it be possible to show the pharmacophore score for both acceptors and donners for ZINC?\n\nSome key reference should be included in the manuscript:\nCOVID-19 pandemic: an overview of epidemiology, pathogenesis, diagnostics and potential vaccines and therapeutics1. COVID-19: Underpinning Research for Detection, Therapeutics, and Vaccines Development2. The Structural Basis of Accelerated Host Cell Entry by SARS-CoV-23. Questions concerning the proximal origin of SARS‐CoV‐24.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6200",
"date": "21 Dec 2020",
"name": "Bruno Andrade",
"role": "Author Response",
"response": "The manuscript is well written, balanced, and follows the scientific logic and flow very nicely. The work is very novel and holds great potential. The approach is very useful and holds great potential in using the docking system to maybe identify new antiviral candidates in this pandemic. We would like to thank you very much for your description of our work. We improved the paper based on your recommendations. Some minor comments: Please list the software used for generating the docking work. The docking protocol is described in the Methods section (Docking Studies subsection) Could you please improve the figure axis of 6 bars for clarity (B and E)? Figure 6 (B and E) is part of Autodock 4 results, visualized at Autodock Tools program. This program doesn’t give us a good resolution docking energy graph. We improved the figure quality using an image editor program in this new version. Thank you for your suggestion. Could you please add a short paragraph on the novelty of this approach of molecular docking? Our focus in this work was using pharmacophore modeling in order to find new possible ligand Mpro inhibitors, as well as repurpose approved drugs for SARS-CoV-2. In this case, we used a common docking approach for confirming our pharmacophore predictions. All docking protocol is described in the Methods section. Thank you very much for your suggestion. Could you please indicate if any of the reported drugs have been used in clinical trials originated from the docking approach? All molecules used for pharmacophore generation were tested in vitro, in vivo, and in silico. We included four citations for the in silico tests in the discussion section. Would it be possible to show the pharmacophore score for both acceptors and donners for ZINC? We used OpenEye vROCS program, which considers the Tanimoto Combo (TC) scores for selecting the most similar pharmacophore-like molecules. We included a column in table 2, showing the TC score for each best-selected ligand. Thank you very much for this suggestion. Some key reference should be included in the manuscript: COVID-19 pandemic: an overview of epidemiology, pathogenesis, diagnostics and potential vaccines and therapeutics1. We included the suggested reference in this new version. Thank you. COVID-19: Underpinning Research for Detection, Therapeutics, and Vaccines Development2. The Structural Basis of Accelerated Host Cell Entry by SARS-CoV-23. Questions concerning the proximal origin of SARS‐CoV‐24. We included the suggested reference in this new version. Thank you. References included 1. Amawi H, Abu Deiab GI, A Aljabali AA, Dua K, Tambuwala MM. COVID-19 pandemic: An overview of epidemiology, pathogenesis, diagnostics and potential vaccines and therapeutics. Therapeutic Delivery. 2020. 2. Aljabali AAA, Bakshi HA, Satija S, Metha M, Prasher P, Ennab RM, et al. COVID-19: Underpinning Research for Detection, Therapeutics, and Vaccines Development. Pharm Nanotechnol. 2020; 3. Hagar M, Ahmed HA, Aljohani G, Alhaddad OA. Investigation of some antiviral N-heterocycles as COVID 19 drug: Molecular docking and DFT calculations. Int J Mol Sci. 2020; 4. C S, S DK, Ragunathan V, Tiwari P, A S, Brindha Devi BD. Molecular docking, validation, dynamics simulations, and pharmacokinetic prediction of natural compounds against the SARS-CoV-2 main-protease. J Biomol Struct Dyn. 2020; 5. Naik VR, Munikumar M, Ramakrishna U, Srujana M, Goudar G, Naresh P, et al. Remdesivir (GS-5734) as a therapeutic option of 2019-nCOV main protease–in silico approach. J Biomol Struct Dyn. 2020; 6. Peele KA, Potla Durthi C, Srihansa T, Krupanidhi S, Ayyagari VS, Babu DJ, et al. Molecular docking and dynamic simulations for antiviral compounds against SARS-CoV-2: A computational study. Informatics Med Unlocked. 2020;"
}
]
}
] | 1
|
https://f1000research.com/articles/9-514
|
https://f1000research.com/articles/9-1494/v1
|
21 Dec 20
|
{
"type": "Software Tool Article",
"title": "Bleach correction ImageJ plugin for compensating the photobleaching of time-lapse sequences",
"authors": [
"Kota Miura"
],
"abstract": "During the capturing of the time-lapse sequence of fluorescently labeled samples, fluorescence intensity exhibits decays. This phenomenon is known as 'photobleaching' and is a widely known problem in imaging in life sciences. The photobleaching can be attenuated by tuning the imaging set-up, but when such adjustments only partially work, the image sequence can be corrected for the loss of intensity in order to precisely segment the target structure or to quantify true intensity dynamics. We implemented an ImageJ plugin that allows the user to compensate for the photobleaching to estimate the non-bleaching condition with choice of three different algorithms: simple ratio, exponential fitting, and histogram matching methods. The histogram matching method is a novel algorithm for photobleaching correction. This article presents details and characteristics of each algorithm based on application to actual image sequences.",
"keywords": [
"Fluorescence microscopy",
"photobleaching",
"bleach correction",
"histogram matching",
"restoration",
"time series",
"Fiji",
"ImageJ"
],
"content": "Introduction\n\nIn biological fluorescence microscopy, cells are irradiated with excitation light that causes the emission of fluorescence from protein markers. This irradiation is necessary to detect the position of proteins but it also has a side effect. Emitted fluorescence gradually decreases by time. This is because, with a certain probability, fluorophores irreversibly lose the ability to fluoresce. This effect is called “photobleaching” and is a widely-known problem in bioimage analysis. Photobleaching not only degrades the visual quality of the results but also interferes with the measurement of molecular kinetics and the quality of the segmentation of target objects. Since bleaching attenuates total signal intensity even when the density of labeled protein is unchanged, precise estimation of the amount of protein or the boundary of the biological structure becomes difficult.\n\nTo overcome the problem of photobleaching, improvements can be made either before or after the experiment. Before the experiment, one can tune instruments and sample environment, such as careful choice of the fluorophore, use of anti-fading reagents, decreasing the power of laser irradiation, increasing the detector gain, and increasing the time interval of capturing1. After the experiment, time-lapse sequence image data can be processed to compensate for the loss of intensity. To do this, the amount of the fluorescence loss is estimated and then the none-bleached condition is restored by image processing. We call such restoration procedure “bleach correction”. Several different algorithms have been developed and used by various researchers.\n\nA conventional method for correcting the bleaching has been done by multiplying the inverse of the ratio of intensity loss compared to a reference image frame. Estimation of bleach ratio could either be calculated directly2–4 or by fitting exponential equation5,6. We call the first method simple ratio method and the second method exponential fitting method.\n\nThe simple ratio method seems to be the most widely used method. It has been applied for improving the deconvolution results in 3D stacks7 and in the fluorescence recovery after the photobleaching (FRAP) technique. In FRAP literature, this method is called double normalization2,3. In ImageJ, this method was implemented as a macro by Jens Rietdorf and has been available since 2004 (https://www.embl.de/eamnet/html/bleach_correction.html).\n\nIn the exponential fitting method, it also compensates the loss by multiplying the inverse of the bleach ratio, but the method first fits the exponential equation to the bleaching curve and uses the fitted parameters for deriving the bleach ratio at each time point8,9. This method has been implemented as an ImageJ plugin and is available as a part of MBF ImageJ for Microscopy bundle (https://imagej.net/MBF_Plugin_Collection). A plugin PixBleach allows three different exponential decay models to be used for the correction (http://bigwww.epfl.ch/algorithms/pixbleach/)10.\n\nThe histogram matching method is based on a strategy that is different from the above two methods and a novel algorithm introduced in this paper. Instead of correcting the fluorescence intensity based on the average intensity of the image at each time point, the histogram matching algorithm11 is used to restore comparable intensity histogram distribution of each frame by taking the first frame as the reference.\n\nWe implemented an ImageJ plugin that allows bleach correction with a choice from these three different algorithms. In this article, we explain each of these algorithms in detail and compare their characteristics.\n\n\nMethods\n\nIn the case of the simple ratio and exponential fitting methods, the mean intensity of 3D stack from each time point was calculated to estimate the bleach ratio. For the histogram matching method, pixel intensity histogram was generated from the 3D stack at each time point.\n\nWe consider an i-th frame image Ii = Ii(x, y) in an image sequence and correct its loss of fluorescence emission. We assume that the mean intensity I¯ is constant through the time-lapse sequence if not for the photobleaching. Then the ratio of the mean intensity of i-th frame I¯i to that of the first frame I¯0 is the ratio of none-photobleached fluorophores in i-th frame. We could then estimate the true pixel intensity of the i-th frame by the equation below.\n\n\n\nIb is the value of the background intensity. This value is estimated independently by measuring the none-fluorescence region within the image, or by measuring a blank image with all the image acquisition conditions being the same but without sample.\n\nThe bleach correction tool included in MBF ImageJ bundle curve-fits total intensity value of each time frame with an exponential decay curve. This decay curve is then used to estimate the bleach ratio at each time point to calculate the true fluorescence intensity12. We implemented a similar capability for processing three-dimensional time series. In this case, the mean intensity of each time point, the average of 3D stack pixel values, was first fitted by an exponential equation to estimate the background intensity.\n\n\n\nValues a, b and c are estimated by this curve fitting. The original image is then subtracted by the estimated background value c. The background-subtracted image was then fitted again with the single exponential equation. Using the estimated parameters from this second fitting, which we now call them a', b', and c', the ratio of bleaching was determined and then it’s inverse was multiplied to the background-subtracted image.\n\n\n\nHistogram matching algorithm modifies pixel values of an image to match its histogram shape to a reference image histogram13. We used the histogram of the first frame image H0(p) as a reference and matched the histogram of i-th image frame Hi(p). p is the pixel value that is 0 ≤ p ≤ 255 in 8-bit image and is 0 ≤ p ≤ 65535 in 16-bit image. The cumulative distribution function of histogram CDFi(p) is used for the actual calculation.\n\n\n\nSince we take the first frame of the time-lapse sequence as the reference, we use CDF0(p) as the reference CDF. We then match the rest of CDF to the reference CDF0(p) by\n\nwhere CDF0−1 is the inverse function of CDF0 and p' is the pixel value updated after matching a pixel value p in the original image.\n\nFor running this plugin with Fiji, any entry-level laptop or desktop machine is sufficient. If one needs to work on a huge image stack, then one should make sure that the RAM has a capacity twice the size of the file size of the image stack. In case of Mac, Mac OS X 10.4 or higher is required to run Fiji.\n\n\nResults\n\nThe sample image sequence was a time series of three-dimensional stacks14,15. The mean intensity showed an overall decrease by time, accompanied by repetitive small peaks (Figure 1, top-left). These small peaks corresponded to single time points, as each peak represented the spherical shape of the yeast cell. Mean intensity was low at the top slice, high at the cell equator, and then low again at the bottom slice.\n\nFor the original and the histogram matching results, values were subtracted by 68 to set the y-axis range to a comparable level to the other two curves. Black curves are the mean intensity of each image and red curves are the mean intensity of each time point.\n\nTo correct for the bleaching using the simple ratio method, we first determined background intensity. An arbitrary area outside the cell was selected and the mean intensity of the full sequence was measured. The mean intensity of the background was 68.3 ± 0.5. The sample sequence was then corrected for bleaching by the simple ratio method using this background intensity (Figure 1, top-right).\n\nWe further examined how the level of background intensity affects the correction results (Figure 2). We used four different background values, 64, 68, 70, and 72. A slight difference in the background values caused a large difference in the resulting curves. When the background value was set large, the correction resulted in a curve with an increasing trend. When the value is set small the corrected curve showed a decreasing trend. When the measured background intensity 68 was used, drift was minimal.\n\nWith the exponential fitting method, the estimation of background value is already a part of the algorithm. In the sample image, the estimated background intensity was 73. The image sequence was subtracted by this background value, fitted with a single exponential equation and then the bleaching was corrected (Figure 1, bottom-left). The mean intensity of the corrected image stack was mostly constant but with a slightly decreasing trend and also with small fluctuations. The decreasing trend was due to a fact that a single exponential equation was not perfectly fitting to the bleaching process seen in the sample. Since the exponential fitting assumes an idealized decrease in fluorescence by time, fluctuations present in the original image sequence were preserved in the corrected image sequence (Figure 1, red curves).\n\nThe histogram matching method does not need background intensity estimation in its algorithm. For the purpose of comparing this method with the other two methods, the original stack was first subtracted by a constant background intensity value 68. The correction resulted in a stably constant mean intensity time series (Figure 1, bottom-right). It was the most stable result compared to the other two methods in terms of the fluctuation of intensity after the correction (Figure 1 red curves).\n\n\nDiscussion\n\nAll three methods do correct bleaching of fluorescence but the results of correction showed some difference. The simple ratio method has the capacity to correct time series with abrupt changes in intensity because the bleaching ratio is calculated for each time point. On the other hand, the quality of the correction is heavily dependent on the estimated value of background intensity. A small deviation of estimated value from the true background value causes wrong correction results (Figure 2).\n\nThe exponential fitting method has an assumption that the bleaching process follows a single exponential decay. It is known that in some cases bleaching time course is a double exponential decay8. In this respect, one must consider before using this method if there is any good reason to model the bleaching process of the sample as a single exponential decay. Otherwise, the goodness of fit needs to be evaluated for the proper use of this method. In addition, the exponential fitting method ignores small perturbations in the intensity such as abrupt changes in the emission of fluorescence. Such changes can be caused by small fluctuations in the power of the excitation light or in the slight variations in the timings of the shutter controlling the light path. The simple ratio method deals better with such changes. However, such a non-flexible nature of the exponential fitting method can become an advantage in some other occasions. For example, if the change in the intensity is due to the synthesis of GFP molecules by cell, the simple ratio method will wrongly correct such true increase and mask the biological event, but the exponential fitting method will achieve a better correction as long as the bleaching is known to be a single exponential decay.\n\nThe histogram matching method is robust when it is difficult to measure background intensity. This can happen when the whole image frame is filled with sample. For example, image data with packed cells in the image frame hinder the estimation of background intensity. Since histogram matching does not require the input of background intensity, the correction will be straight forward even with such cases. Moreover, this method is especially suited as a preprocessing for segmentation since strictly constant mean intensity in the corrected image sequence affords an optimal condition for segmenting objects.\n\nThe limitation of using the histogram matching method is that it assumes a stable distribution of fluorescence. If object under observation undergo changes in localization, e.g. signal changes from diffuse to spots during cell surface receptor internalization, or in shape, e.g. cell spreading, we expect that histogram shape will change as well. If we take an example of cell surface receptor internalization, the formation of spots by the aggregation of protein is expected to create a new peak in the high pixel values while decreasing the height of the existing peak in low pixel values of the histogram. Applying histogram matching to this time course would result in the wrong correction by forcing the histogram shape to become constant over time.\n\n\nConclusions\n\nAll thee methods correct bleaching, but have specific limitations of each. With the simple ratio method, background intensity should be accurately estimated. The exponential fitting method relies heavily on the model. With the histogram fitting method, object shape and pattern should be constant. For choosing an appropriate method, these limitations, and the known details of the observed biological event should be taken into account. In the future, the following features are planned to be added to the plugin: first, a fitting method with a double exponential equation; second, the background estimation method for the simple ratio method will be considered and will be added as a helper function; third, the current histogram matching method uses the first frame as the reference frame. Tolerance to changes in shape and pattern becomes higher if the reference frame is updated for every time frame so that the matching is done between neighboring time points.\n\n\nData availability\n\nFor the development of the plugin, the sample time-lapse sequence of fluorescently labeled yeast cells was kindly provided by Boryana Petrova and Christian Häring (Cell Biology and Biophysics Unit, EMBL Heidelberg). These sequences are three-dimensional time-lapse movies, taken with eight optical sections for each time point14.\n\nZenodo: A sample image data for ImageJ Bleach Correction Plugin, http://doi.org/10.5281/zenodo.406011115.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).\n\n\nSoftware availability\n\nThe Bleach correction plugin is contained within the download package of Fiji (https://imagej.net/Fiji) and can be used directly by launching the software package and accessing the menu item [Image > Adjust > Bleach Correction] without any further additional installation. For using it in ImageJ 1.× (ImageJ (https://imagej.nih.gov/ij/), the source code should be downloaded, compiled, and installed locally.\n\nSource code (version 2.0.3) available from: https://github.com/fiji/CorrectBleach\n\nArchived source code as at time of publication: http://doi.org/10.5281/zenodo.5870116\n\nLicense: GNU General Public License version 2.",
"appendix": "Acknowledgements\n\nWe are grateful to Boryana Petrova and Christian Häring (Cell Biology and Biophysics unit, EMBL Heidelberg) for providing yeast image sequences. We thank Johannes Schindelin for suggestions on coding and including the plugin in the Fiji project.\n\nThis publication was supported by COST Action NEUBIAS (CA15124), funded by COST (European Cooperation in Science and Technology\n\n\nReferences\n\nWaters JC: Live-cell fluorescence imaging. Methods Cell Biol. 2007; 81(06): 115–40. PubMed Abstract | Publisher Full Text\n\nPhair RD, Misteli T: Kinetic modelling approaches to in vivo imaging. Nat Rev Mol Cell Biol. 2001; 2(12): 898–907. PubMed Abstract | Publisher Full Text\n\nStavreva DA, McNally JG: Fluorescence recovery after photobleaching (FRAP) methods for visualizing protein dynamics in living mammalian cell nuclei. Methods Enzymol. 2004; 375(1976): 443–55. PubMed Abstract | Publisher Full Text\n\nMcNally JG: Quantitative FRAP in analysis of molecular binding dynamics in vivo. Methods Cell Biol. 2008; 85(08): 329–51. PubMed Abstract | Publisher Full Text\n\nWüstner D, Faergeman NJ: Chromatic aberration correction and deconvolution for UV sensitive imaging of fluorescent sterols in cytoplasmic lipid droplets. Cytometry A. 2008; 73(8): 727–44. PubMed Abstract | Publisher Full Text\n\nKolin Dl, Costantino S, Wiseman PW: Sampling effects, noise, and photobleaching in temporal image correlation spectroscopy. Biophys J. 2006; 90(2): 628–39. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarkham J, Conchello JA: Artefacts in restored images due to intensity loss in three-dimensional fluorescence microscopy. J Microsc. 2001; 204(Pt 2): 93–8. PubMed Abstract | Publisher Full Text\n\nFüreder-Kitzmüller E, Hesse J, Ebner A, et al.: Non-exponential bleaching of single bioconjugated Cy5 molecules. Chem Phys Lett. 2005; 404(1–3): 13–18. Publisher Full Text\n\nVicente NB, Zamboni JED, Adur JF, et al.: Photobleaching correction in fluorescence microscopy images. J Phys Conf Ser. 2007; 90: 012068. Publisher Full Text\n\nWüstner D, Larsen AL, Faergeman NJ, et al.: Selective visualization of fluorescent sterols in Caenorhabditis elegans by bleach-rate-based image segmentation. Traffic. 2010; 11(4): 440–454. PubMed Abstract | Publisher Full Text\n\nGopinath S, Wen Q, Thakoor N, et al.: A statistical approach for intensity loss compensation of confocal microscopy images. J Microsc. 2008; 230(Pt 1): 143–59. PubMed Abstract | Publisher Full Text\n\nAndrews DW: The MBF ImageJ collection. 2005. Reference Source\n\nBurger W, Burge MJ: Digital Image Processing: An Algorithmic Introduction using Java. Springer, 2007. Reference Source\n\nPetrova B, Dehler S, Kruitwagen T: Quantitative Analysis of Chromosome Condensation in Fission Yeast. Mol Cell Biol. 2013; 33(5): 984–998. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMiura K: A sample image data for ImageJ Bleach Correction Plugin. type: dataset. 2020. http://www.doi.org/10.5281/zenodo.4060111\n\nSchindelin J, Rueden C, Miura K, et al.: CorrectBleach: upgrade with Exponential fitting method (Version v2.0.3). Zenodo. 2016. http://www.doi.org/10.5281/zenodo.58701"
}
|
[
{
"id": "76514",
"date": "18 Jan 2021",
"name": "Fabrice P Cordelières",
"expertise": [
"Reviewer Expertise Bio-image analysis"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper presents a software tool by Kota Miura aimed at correcting image datasets for the effect of photobleaching. Shaped as an ImageJ/Fiji plugin, it provides three different methods: the first is based on a simple ratio, the second on exponential fitting, the third on histogram matching. The latter method is novel to the field. The author describes both the theoretical approaches on which the tools has been built on, gives characterization details and finally concludes on the application field of each method.\nAlthough the manuscript meets the expectations about what a software paper should be, minor suggestions and questions have been raised when reading the paper.\nImproving the reader’s experience:\nIn spite of the full dataset being available through Zenodo, no image is actually presented within the manuscript. A selection of the original and processed images could be presented alongside the graphs, for visual inspection.\n\nThe conclusion section clearly states the different fields of application/restriction use for each single method. A visual transcript, as a decision tree might help the end-user decide which method to try in her/his first attempts.\n\nAbout the methods section:\nSimple ratio: the author makes the reader aware of the caution to be applied when estimating the background, and its impact on the correction when overestimated. It remains unclear to me how the plugin gets this background estimate: is it automatically estimated (if so, how?) or should the user place a region of interest over the image?\n\nExponential fitting: Would it be possible to use background estimate from the exponential fitting method as a starting point to evaluate which value to set when using the simple ratio method? The background estimate is higher than the one used for simple ratio method, but it could be of interest to investigate if a value could be automatically derived from this estimate, if possible/relevant.\n\nAbout the discussion section:\nThe author discusses the effect of physiological fluctuation of intensity on the efficiency of the correction algorithms, pointing out the simple ratio method as the most appropriate in such situations. Would pre-processing the data with smoothing algorithms such as applying a sliding average algorithm (with proper window size) help?\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "78015",
"date": "08 Feb 2021",
"name": "Arianne Bercowsky Rama",
"expertise": [
"Reviewer Expertise Light Sheet Microscopy",
"Bio-image analysis"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article presents three different methods for the correction of photo-bleaching as a post-processing step using an ImageJ/Fiji plugin. The mathematical details of how the corrections are done in 3-dimensional time-lapse data are explained. Moreover, an example data-set is provided to show the photo-bleach correction results when implementing each method. Limitations and examples for the three methods are described, guiding the user when to apply them. The three methods are: simple-ratio, exponential fitting and histogram matching. This last one is a novel algorithm introduced in this article.\nThe software tool described in this article by Kota Miura provides the information needed to understand how the Bleach correction ImageJ/Fiji plugin works. The article is easy to read and offers sufficient details to comprehend the stated methods and its applicability. A few questions were raised while reading the methods section:\nExponential fitting method: Once the background subtracted image is fitted a second time with the single exponential equation, you obtain the values a’, b’ and c’. It is not clear why it is then used the estimated value 'c’ since it seems it should be very close to 0.\n\nOperation: the author suggests that the RAM should have a capacity twice the size of the file size of the image stack. Is it because the new corrected image will be duplicated? For very large data-sets, can Virtual Stack be used?\n\nSimple ratio method and exponential fitting method: In these two methods the background plays an important role in the correction. What happens when the background is non-uniform? Is there a need of a prior step to correct for this in order to use these photo bleach correction methods?\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "78014",
"date": "09 Feb 2021",
"name": "Thomas Pengo",
"expertise": [
"Reviewer Expertise Image processing and analysis"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe author has addressed a very common issue in fluorescence microscopy experiments and has provided an easy to use tool that is integrated directly into Fiji. The manuscript is easy to read and complete.\nI particularly appreciate the effort to provide precise references to the software versions.\nIt would be nice to have an image illustrating the effects of photo-bleaching and the corresponding compensation algorithms. The author provides plots, but no images.\nI have a few minor comments. The page numbers refer to the PDF version, as appear in the first version 1 of the manuscript.\nPage 3, col.1\n\n“before or after the experiment”. Unclear, maybe add “imaging experiment”, or simply “imaging” or “image acquisition” or “image data acquisition”. The experiment, without context, is typically larger in scope and includes, for example, sample preparation.\n\n“fluorescence loss is estimated and then the none[sic]-bleached condition is restored” is a bit too optimistic. Apart from the typo (non-bleached), I would argue something like “fluorescence loss is estimated and compensated for through image processing” would convey a more accurate assessment.\nPage 3, col.2\n\n“none-photobleached” => “non-photobleached”\n\n“none-fluorescence” => “non-fluorescing”? or, better, “an empty region”\nPage 4, Figure 1. The plot was confusing at first glance as it mixes two dimensions in one. I’d suggest using either a 3D surface (t, z vs Intensity) or an average/max projection in z (t, Intensity mean over z). The latter is probably sufficient even though it hides the Intensity variation in z.\n\nPage 4, col. 1 There is no mention of the number of bins in the image. This is relevant for the following reason. CDF0 can be inverted only if strictly monotonic (CDF0(a) < CDF0(b) i.i.f. a0-1 will have two possible values c-1 and c. An easy fix would be to add a small constant to H everywhere (e.g. 1) after calculation.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "78013",
"date": "16 Feb 2021",
"name": "Anna H. Klemm",
"expertise": [
"Reviewer Expertise BioImage Analysis"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article describes the Fiji plugin Bleach correction, which integrates three methods for correcting intensity loss by bleaching. One of the methods, histogram matching, is introduced and novel for addressing bleaching. All methods are well explained. Methods are compared to each other not only by performance on the given use case but also with other use scenarios in mind. I especially liked that specific biological examples were given (e.g. synthesis of GFP) and how the methods would compare in these scenarios.\nIt was no problem to test the plugin with the sample data linked in the text and the description.\nStructure of the article. Methods and results could be fused, such that the explanation of a method goes directly along with presentation of the correction result on the sample data set. It should be referred to Fig 1 earlier in the text, e.g. Fig 1 top left to illustrate bleaching when introducing it in the very beginning. The chosen structure could be due to guidelines of the publisher and in this case it should be communicated to the publisher.\nOther minor suggestions:\nIntroduction:\nInclude a sentence about ImageJ.\n\nMethods:\nSimple ratio method: We could then estimate the true pixel intensity Iic of .. Include the information that the measurement is done per slice.\n\nHistogram matching method: Since the article has also educational character it could be good to include example cumulative histograms of a frame before and after correction using the histogram matching method. We used the histogram of the first frame (per slice?) image H0 (p) as a reference..\n\nResults:\nThe sample image sequence was a time series of three-dimensional stacks of a yeast cell.\n\nSimple ratio method: Add a brief definition of background to the introduction, and that, due to the different nature of source, it has a different bleaching behavior. I liked that the effect of the background intensity value was given much space in the article, including a figure.\n\nExponential fitting method: In the sample image, the estimated background intensity was 73. : In the sample image, the background intensity estimated by fitting was 73 (to stress that estimation by fitting is different from the estimation by manually selecting a background region and measuring the intensity within the selection). You might want to show the fit together with the curve (as done in the plugin output).\n\nFigure 1: Specify that it is a stack imaged over time also in the figure legend. Instead of a red line use red dots at the center slice to reflect that it is the average value per time point. Dots in the center slice position can also help to visually check whether the z-position was stable over the time of recording.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "77752",
"date": "22 Feb 2021",
"name": "Bram van den Broek",
"expertise": [
"Reviewer Expertise BioImage Analysis",
"Microscopy",
"Biophysics"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this paper Kota Miura describes an ImageJ plugin for fluorescence photobleaching correction. This tool has been built into Fiji for a number of years already, and is no doubt used by many researchers. The plugin comprises three different methods to compensate for photobleaching in time-lapse movies (2D/3D). All methods alter the pixel intensities of the output image, in an attempt to stabilize the signal.\nIt is praiseworthy that the author takes the effort to describe such an established tool. The article thus also serves as a manual.\nThe article is well-written in a very straightforward way, without unnecessary details. The three methods are independently explained to a level that the reader can understand the workflow, but the descriptions don’t go very deep. The discussion is to the point and mainly handles limitations of the implemented methods. The author provides adequate links to software, source code and sample data.\nWhile reading the paper I had several questions and a few concerns, which I will discuss below. I realize that there are a lot of points; some are very technical and some can be seen as feature requests. I do not expect the author to address/implement them all.\nAll example data in the paper is taken from one sample image of a single cell. This image is acquired in 3D, but the bleach correction is performed, or at least shown, in a ‘linearized’ 2D stack. This 2D representation causes periodic ‘peaks’ in the time trace. Perhaps it is the author’s intention to show that the tool works for 3D time-lapses, but these ‘peaks’ obscure the bleaching curve, are unnecessarily confusing the reader and hamper a thorough comparison of the methods. Luckily, the mean intensity of each time point is also provided (the red lines in figure 1). Without the black lines, all four panels can be merged into one panel, with different methods in different colors. Doing so will free up room for a few other bleaching correction examples. Ideally, such examples would reflect cases that highlight the positive/negative sides of each method. In my opinion that would really elevate the paper beyond a fancy manual. (I have imaged some data for this review that the author is free to use for this purpose, if desired.).\n\nThe output image in the Simple Ratio method has the background subtracted, where in the other methods the background remains. It would be better to keep this the same for all methods. Because the output image is not in 32-bit floating point format, this operation also creates a positive bias in the background noise (negative numbers are set to 0). Even more so, the calculation itself seems to be done in 8-bit or 16-bit integer format. This partially underlies the following issue:\n\nFigure 2 depicts corrected traces of a single cell surrounded by background, for varying input background estimation, using the Simple Ratio method. The figure seems to indicate that overestimating the background is much more severe than underestimating. However, the interesting part of the data is obviously the cell, and not the background. The corrected traces are mean intensities of the whole image. A flat trace can arise from the amplified signal of the bleaching cell alone (this is what we want), but also from an erroneously amplified background. With many more background pixels than ‘cell pixels’ this effect can be large even with moderate background increase. Exactly this turns out to be happening with the image stack presented here. I have reproduced the bleaching corrected data with the sample image used in the article, but now also separated the foreground (the cell) from background (using a simple Otsu threshold on a slice where the cell is in focus, halfway the time-lapse). The resulting time traces (total, cell, background) show that when the background set at 64, the background intensity increases by 45% while the cell intensity decreases by 35%: under-correction. With background at 72 the background intensity decreases by 42%, while the cell intensity increases by 29%: over-correction. (In the latter case the background gray values are almost zero everywhere, because the image remains 8-bit; this could be part of the problem.) When a correct background of 68.3 is set (or 69, because non-integer values seem to be rounded up) the cell intensity over time remains roughly constant. Because the traces as shown in Figure 2 (and Figure 1 as well!) are computed from the cell+background, which both vary, they do not fully convey the right message. Interestingly, the paragraph explaining the effect of background intensity correctly describes the result, but the words do not agree with the figure. Perhaps the author can instead show both cell and background traces (and then preferably only the mean/sum of all z planes vs time, to remove the spikes).\n\nI noticed that with the Simple Ratio and Exponential Fit methods it is possible to calculate the correction on a selection (ROI) in the image, yielding different results. Please add a paragraph on the effect, and when it is useful to do so.\n\nExponential method: The estimated background value c is subtracted from the original. Is this done in 32-bit floating point numbers, to prevent a positive bias in the background noise?\n\nExponential method: The text says “The original image is then subtracted by the estimated background value c. The background-subtracted image was then again with the single exponential equation.” Why is the data again fitted again after background subtraction? Am I missing something, or would it just lead to a = a’, b = b’ and c’=0? And if not, why is that? When running the plugin the exponential fit is displayed. Are the parameters in this fit plot a, b, c... or a’, b’, c'?\n\nExponential method: The author correctly states that bleaching sometimes (more often than not?) follows a double exponential curve, and that “the goodness of fit needs to be evaluated for the proper use of this method.” Unfortunately, the displayed R2 in the plot is not the same as goodness of fit, and is really not valid for a nonlinear regression. (See for instance Spiess and Neumeyer (20101). A real goodness of fit, like Chi2, involves knowing the uncertainty of the signal. I am not asking for that, but for judging the goodness of fit by eye it would be instructive to also show a residuals plot.\n\nHistogram Matching method: This method yields the best results in some cases, as described, but I also observe strange incorrect behavior on some of my (confocal) bleaching time-lapses (with no change in fluorescence localization, background almost zero): sudden steps, not a stable signal. I’m not sure what to make of that, but I suspect that it involves handling of the background, possibly because of integer rounding. Also with this method, I sometimes see the background increase (in steps), though the effect does not seem as severe as choosing a wrong background in the Simple Ratio case.\n\nThe Histogram Matching method is blazingly fast for 8-bit images, but extremely slow for 16-bit images. The difference at least 1000-fold.\n\nThe output images are all named DUP_[image name]. Renaming them would be nice, e.g. [image name]_BleachCorrected_[method].\n\nTextual changes:\nAbstract: We implemented an ImageJ plugin that allows the user to compensate for the photobleaching to estimate the non-bleaching condition with choice of three different algorithms – This sentence seems incorrect. Replace second to by and?\n\nIntroduction: A conventional method for correcting the bleaching has been done by multiplying the inverse of the ratio of intensity loss compared to a reference image frame – sentence not quite correct. Multiplying by the inverse of the relative intensity loss compared to a reference frame?\n\nIn the exponential fitting method, it also compensates the loss by multiplying by/with the inverse of the bleach ratio, but…\n\nMethods: In the case of the simple ratio and exponential fitting methods, the mean intensity of the 3D stack from each time point was calculated to estimate the bleach ratio. For the histogram matching method, the pixel intensity histogram was generated from the 3D stack at each time point.\n\nThe histogram matching algorithm modifies pixel intensities…\n\nConclusions: All three methods...\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1494
|
https://f1000research.com/articles/9-1279/v1
|
28 Oct 20
|
{
"type": "Software Tool Article",
"title": "Automated cell tracking using StarDist and TrackMate",
"authors": [
"Elnaz Fazeli",
"Nathan H. Roy",
"Gautier Follain",
"Romain F. Laine",
"Lucas von Chamier",
"Pekka E. Hänninen",
"John E. Eriksson",
"Jean-Yves Tinevez",
"Guillaume Jacquemet",
"Elnaz Fazeli",
"Nathan H. Roy",
"Gautier Follain",
"Romain F. Laine",
"Lucas von Chamier",
"Pekka E. Hänninen",
"John E. Eriksson",
"Jean-Yves Tinevez"
],
"abstract": "The ability of cells to migrate is a fundamental physiological process involved in embryonic development, tissue homeostasis, immune surveillance, and wound healing. Therefore, the mechanisms governing cellular locomotion have been under intense scrutiny over the last 50 years. One of the main tools of this scrutiny is live-cell quantitative imaging, where researchers image cells over time to study their migration and quantitatively analyze their dynamics by tracking them using the recorded images. Despite the availability of computational tools, manual tracking remains widely used among researchers due to the difficulty setting up robust automated cell tracking and large-scale analysis. Here we provide a detailed analysis pipeline illustrating how the deep learning network StarDist can be combined with the popular tracking software TrackMate to perform 2D automated cell tracking and provide fully quantitative readouts. Our proposed protocol is compatible with both fluorescent and widefield images. It only requires freely available and open-source software (ZeroCostDL4Mic and Fiji), and does not require any coding knowledge from the users, making it a versatile and powerful tool for the field. We demonstrate this pipeline's usability by automatically tracking cancer cells and T cells using fluorescent and brightfield images. Importantly, we provide, as supplementary information, a detailed step-by-step protocol to allow researchers to implement it with their images.",
"keywords": [
"Cell migration",
"Image analysis",
"StarDist",
"TrackMate",
"Deep-learning",
"Automated tracking"
],
"content": "Introduction\n\nThe study of cell motility typically involves recording cell behavior, using live-cell imaging, and tracking their movement over time1,2. To enable the analysis of such data, various software solutions have been developed3–9. However, despite the availability of these computational tools, manual tracking remains widely used among researchers due to the difficulty in setting up fully automated cell tracking analysis pipelines. Automated tracking pipelines share a typical workflow that starts with a segmentation strategy that identifies the objects to track in each image. Tracking algorithms are then used to link these objects between frames. One challenging aspect of an automated tracking pipeline is often achieving an accurate segmentation of the objects to track. One option to facilitate cell segmentation is to label their nuclei, using fluorescent dyes or protein markers. Nuclei can then be automatically segmented using intensity-based thresholding. However, this approach tends to become inaccurate when images are noisy or when the cells to track are very crowded10. Deep-Learning approaches have demonstrated their robustness against these two issues11. In this work, we present a new analysis workflow that builds upon a Deep-Learning segmentation tool and a cell tracking tool to achieve robust cell tracking in cell migration assays. We combine StarDist, a powerful deep learning-based segmentation tool, and TrackMate, a user-friendly tracking tool, into a tracking pipeline that can be used without requiring expertise in or specialized hardware for computing (Figure 1)12–15.\n\n\nMethods\n\nThe use of deep learning networks, such as StarDist, often requires the user to train or retrain a model using their images. While high-quality StarDist pre-trained models are readily available, they are likely to underperform when used on different data with, e.g., different staining, noise, and microscope type15. To train StarDist models, we took advantage of the ZeroCostDL4Mic platform, allowing researchers to train (and retrain), validate, and use deep learning networks15. Importantly, the ZeroCostDL4Mic StarDist 2D notebook can directly output a file containing all the nuclei's geometric center coordinates (named tracking files), that can be used as input for TrackMate (Figure 1). Therefore, our proposed pipeline can be divided into three parts (Figure 1; Extended data16). 1) First, a StarDist model is trained using the ZeroCostDL4Mic platform. This part needs to be performed only once for each type of data. 2) Second, the trained StarDist model is used to segment the object to track and generate Tracking files. 3) Finally, the tracking files can be used in TrackMate to track the identified objects.\n\nTraining a StarDist model requires a set of images and their corresponding masks (Figure 1 and Figure 2). Generating a training dataset is by far the most time-consuming part of the analysis pipeline presented here as it requires the manual annotations of the images to analyze (Extended data: Supplementary protocol16). For instance, to generate the training datasets presented in Figure 2, each cell/nuclei contour was drawn manually using the freehands selection tool in Fiji. The creation of a high-quality training dataset is a critical part of the process as it will impact the specificity and performance of the StarDist model. However, the generation of a training dataset is only required once per dataset type. If a StarDist model already exists for similar images it can be used to significantly accelerate the creation of the training dataset via semi-automated annotation (see Extended data: Supplementary protocol16).\n\n(A, B) Migration of MCF10DCIS.com, labeled with Sir-DNA, recorded using a spinning disk confocal microscope and automatically tracked. Examples of images used to train StarDist (A), and an example of results obtained using automated tracking are displayed (B, Video 1). The yellow square indicates a magnified ROI, where the local track of each nucleus is displayed. The full cell tracks are displayed on the left. Tracks are color-coded as a function of their maximum instantaneous velocity (blue slow, red fast tracks). (C–E) Migration of activated T cell plated on VCAM-1 or ICAM-1, recorded using a brightfield microscope and automatically tracked. Examples of images used to train StarDist (C) and an example of results obtained using automated tracking are displayed (D, Video 2). (E) Comparison of the migration of activated T cells on VCAM-1 or ICAM-1. Track mean speed and track straightness were quantified. Data are displayed as boxplots. *** p-value = <0.001, p-values were determined using a randomization test. (F, G) Cancer cells flowing in a microfluidic chamber, recorded live using a brightfield microscope and automatically tracked (Video 3). Examples of images used to train StarDist (F), and an example of results obtained using automated tracking are displayed (G). The full tracks shown here were color-coded as a function of their x coordinate.\n\nOne of our analysis pipeline's key features is that, once a StarDist model has been satisfactorily trained, movies of migrating cells can efficiently be processed in batch. Indeed, while individual tracking files can be analyzed one by one using the TrackMate graphical interface, we also provide a Fiji macro to analyze a folder containing multiple tracking files. Our batch processing macro will provide basic quantitative information for each track, including median and maximal speeds. If more information is needed, the tracking results generated by our script are directly compatible with the Motility lab website, where they can be further processed17.\n\nThe described image analysis pipeline is composed of a Jupiter notebook optimized to run in Google Colab (ZeroCostDL4Mic framework15) and a Python script that can run in Fiji14. A step-by-step protocol describing how to use our analysis pipeline is provided as Extended data16.\n\n\nUse case\n\nTo illustrate our analysis pipeline's functionality and flexibility, we first trained a StarDist model to analyze the behavior of breast cancer cells migrating collectively (Figure 2A; Extended data: Video 116). The cancer cell's nuclei were fluorescently labeled, and the cells imaged using fluorescence-based microscopy. The creation of the training dataset used in this example was greatly facilitated by the availability of a StarDist model, released by the StarDist creators, capable of segmenting fluorescent nuclei. In this case, the StarDist Fiji plugin was used to segment the location of nuclei in the training images, and all miss-annotations were manually corrected (Extended data: Supplementary protocol16).\n\nTo highlight that our pipeline can also be used to analyze brightfield images, we generated a StarDist model to track T cells migrating on ICAM-1 or VCAM-1 (Figure 2C–E; Extended data: Video 216). Importantly, automated analysis of these data could reproduce the results obtained via manual tracking19.\n\nFinally, we used our pipeline to automatically track non-adherent cancer cells flowing in a microfluidic chamber (Figure 2F and G; Extended data: Video 316). In this case, automated tracking is especially useful due to the very high number of frames to analyze. For the last two examples, no suitable pre-trained StarDist models were available. Therefore, to generate the training datasets, we manually annotated 20 images and trained a first StarDist model. This model was then used to accelerate the annotation of the rest of the training images.\n\nBreast cancer cell dataset. MCF10DCIS.com cells were described previously15,22. DCIS.COM lifeact-RFP cells were incubated for 2h with 0.5 µM SiR-DNA (SiR-Hoechst, Tetu-bio, Cat Number: SC007) before being imaged live for 14 h using a spinning-disk confocal microscope (1 picture every 10 min). The spinning-disk confocal microscope used was a Marianas spinning disk imaging system with a Yokogawa CSU-W1 scanning unit on an inverted Zeiss Axio Observer Z1 microscope (Intelligent Imaging Innovations, Inc.) equipped with a 20x (NA 0.8) air, Plan Apochromat objective (Zeiss).\n\nT cell dataset. Lab-Tek 8 chamber slides (ThermoFisher) were coated with 2 μg/mL ICAM-1 or VCAM-1 overnight at 4°C19. Activated primary mouse CD4+ T cells were washed and resuspended in L-15 media containing 2 mg/mL D-glucose. T cells were then added to the chambers, incubated 20 min, gently washed to remove all unbound cells, and imaged. Imaging was done using a 10x phase contrast objective at 37°C on a Zeiss Axiovert 200M microscope equipped with an automated X-Y stage and a Roper EMCCD camera. Time-lapse images were collected every 30 sec for 10 min using SlideBook 6 software (Intelligent Imaging Innovations).\n\nFlow chamber dataset. Cancer cells (500,000 cells/ml in PBS) were perfused at a speed of 300 µm/sec using a peristaltic pump (ISMATEC MS12/4 analogic) and a homemade tubing system (Ismatek 3-Stop tubes and Ibidi® tubings and connectors) in a microchannel (Ibidi® µ-slides400 LUER). Images were acquired with a brightfield microscope (Zeiss Laser-TIRF 3 Imaging System, Carl Zeiss) and a 10X objective.\n\nBox plots were generated using PlotsOfData23. Randomization tests were performed using the online tool PlotsOfDifferences24.\n\n\nConclusions\n\nHere we show that StarDist and TrackMate can be integrated seamlessly and robustly to automate cell tracking in fluorescence and brightfield images. We envision that this pipeline can also be applied to any circular or oval-shaped objects. However, we acknowledge that using brightfield images may not always work directly with our pipeline, especially if cells display complex and interchanging shapes, since StarDist is mostly designed to detect round or compact shapes. In this case, other tools, such as Usiigaci, could also be considered8. Still, brightfield images could also be artificially labeled using deep learning, transforming the brightfield dataset into a pseudo-fluorescence one, as can be done with ZeroCostDL4Mic already15. The pipeline described here is currently limited to the tracking of objects in 2D. However, a similar workflow can be applied to 3D datasets as both StarDist and TrackMate can accommodate 3D images12,13,25.\n\n\nData availability\n\nZenodo: Combining StarDist and TrackMate example 1 - Breast cancer cell dataset, http://doi.org/10.5281/zenodo.403497626\n\nZenodo: Combining StarDist and TrackMate example 2 - T cell dataset, http://doi.org/10.5281/zenodo.403492927\n\nZenodo: Combining StarDist and TrackMate example 3 - Flow chamber dataset, http://doi.org/10.5281/zenodo.403493928\n\nZenodo: Combining StarDist and TrackMate - Extended data, http://doi.org/10.5281/zenodo.409146716.\n\nThis project contains the following extended data:\n\nSupplementary protocol\n\nVideo 1: Automated tracking of breast cancer cell migrating collectively. MCF10DCIS.com cells, labeled with Sir-DNA, were recorded using a spinning disk confocal microscope and automatically tracked using StarDist and TrackMate. Local tracks are displayed.\n\nVideo 2: Automated tracking of T cell migrating on ICAM-1. Activated T cell plated ICAM-1 were recorded using a brightfield microscope and automatically tracked using StarDist and TrackMate. Local tracks are displayed.\n\nVideo 3: Automated tracking of cancer cells flowing in a microfluidic chamber. AsPC1 pancreatic cancer cells flowing in a microfluidic chamber were recorded live using a brightfield microscope and automatically tracked using StarDist and TrackMate. Local tracks are displayed.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).\n\n\nSoftware availability\n\nSource code available from: https://github.com/HenriquesLab/ZeroCostDL4Mic\n\nArchived source code at time of publication: http://doi.org/10.5281/zenodo.409147426\n\nLicense: MIT license.",
"appendix": "Acknowledgments\n\nThe Cell Imaging and Cytometry Core facility (Turku Bioscience, University of Turku, Åbo Akademi University, and Biocenter Finland), the Finnish Euro-Bioimaging Node, and Turku Bioimaging are acknowledged for services, instrumentation and expertise.\n\nThis publication was supported by COST Action NEUBIAS (CA15124), funded by COST (European Cooperation in Science and Technology).\n\n\nReferences\n\nJacquemet G, Baghirov H, Georgiadou M, et al.: L-type calcium channels regulate filopodia stability and cancer cell invasion downstream of integrin signalling. Nat Commun. 2016; 7: 13297. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJacquemet G, Morgan MR, Byron A, et al.: Rac1 is deactivated at integrin activation sites through an IQGAP1-filamin-A-RacGAP1 pathway. J Cell Sci. 2013; 126(Pt 18): 4121–4135. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDuChez BJ: Automated Tracking of Cell Migration with Rapid Data Analysis. Curr Protoc Cell Biol. 2017; 76: 12.12.1–12.12.16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCordelières FP, Petit V, Kumasaka M, et al.: Automated Cell Tracking and Analysis in Phase-Contrast Videos (iTrack4U): Development of Java Software Based on Combined Mean-Shift Processes. PLoS One. 2013; 8(11): e81266. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcQuin C, Goodman A, Chernyshev V, et al.: CellProfiler 3.0: Next-generation image processing for biology. PLoS Biol. 2018; 16(7): e2005970. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPiccinini F, Kiss A, Horvath P: CellTracker (not only) for dummies. Bioinformatics. 2016; 32(6): 955–957. PubMed Abstract | Publisher Full Text\n\nBarry DJ, Durkin CH, Abella JV, et al.: Open source software for quantification of cell migration, protrusions, and fluorescence intensities. J Cell Biol. 2015; 209(1): 163–180. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTsai HF, Gajda J, Sloan TFW, et al.: Usiigaci: Instance-aware cell tracking in stain-free phase contrast microscopy enabled by machine learning. SoftwareX. 2019; 9: 230–237. Publisher Full Text\n\nPijuan J, Barceló C, Moreno DF, et al.: In vitro Cell Migration, Invasion, and Adhesion Assays: From Cell Imaging to Data Analysis. Front Cell Dev Biol. 2019; 7: 107. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChenouard N, Smal I, de Chaumont F, et al.: Objective comparison of particle tracking methods. Nat Methods. 2014; 11(3): 281–289. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCaicedo JC, Roth J, Goodman A, et al.: Evaluation of Deep Learning Strategies for Nucleus Segmentation in Fluorescence Images. Cytometry A. 2019; 95(9): 952–965. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTinevez JY, Perry N, Schindelin J, et al.: TrackMate: An open and extensible platform for single-particle tracking. Methods. 2017; 115: 80–90. PubMed Abstract | Publisher Full Text\n\nSchmidt U, Weigert M, Broaddus C, et al.: Cell Detection with Star-Convex Polygons. In: Medical Image Computing and Computer Assisted Intervention – MICCAI 2018 (eds. Frangi, A. F., Schnabel, J. A., Davatzikos, C., Alberola-López, C. & Fichtinger, G.) (Springer International Publishing, 2018). 11071: 265–273. Publisher Full Text\n\nSchindelin J, Arganda-Carreras I, Frise E, et al.: Fiji: an open-source platform for biological-image analysis. Nat Methods. 2012; 9(7): 676–682. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChamier LV, Laine RF, Jukkala J, et al.: ZeroCostDL4Mic: an open platform to use Deep-Learning in Microscopy. bioRxiv. 2020; 2020.03.20.000133. Publisher Full Text\n\nJacquemet G, Roy NH, Follain G, et al.: Combining StarDist and TrackMate - Extended data. Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4091467\n\nWortel IMN, Dannenberg K, Berry JC, et al.: CelltrackR: an R package for fast and flexible analysis of immune cell migration data. 2019. Publisher Full Text\n\nJacquemet G: Video 1: Automated tracking of breast cancer cell migrating collectively. f1000research.com. Media.2020. http://www.doi.org/10.6084/m9.figshare.13122635.v1\n\nRoy NH, Kim SHJ, Buffone A Jr, et al.: LFA-1 signals to promote actin polymerization and upstream migration in T cells. J Cell Sci. 2020; 133(17): jcs248328. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJacquemet G: Video 2: Automated tracking of T cell migrating on ICAM-1. f1000research.com. Media.2020. http://www.doi.org/10.6084/m9.figshare.13122755.v1\n\nJacquemet G: Video 3: Automated tracking of cancer cells flowing in a microfluidic chamber. f1000research.com. Media.2020. http://www.doi.org/10.6084/m9.figshare.13122764.v1\n\nJacquemet G, Paatero I, Carisey AF, et al.: FiloQuant reveals increased filopodia density during breast cancer progression. J Cell Biol. 2017; 216(10): 3387–3403. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPostma M, Goedhart J: PlotsOfData—A web app for visualizing data together with their summaries. PLoS Biol. 2019; 17(3): e3000202. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoedhart J: PlotsOfDifferences – a web app for the quantitative comparison of unpaired data. bioRxiv. 2019; 578575. Publisher Full Text\n\nWeigert M, Schmidt U, Haase R, et al.: Star-convex Polyhedra for 3D Object Detection and Segmentation in Microscopy. 2020; 8. Publisher Full Text\n\nJacquemet G: Combining StarDist and TrackMate example 1 - Breast cancer cell dataset (Version 1) [Data set]. Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4034976\n\nRoy NH, Jacquemet G: Combining StarDist and TrackMate example 2 - T cell dataset (Version 1) [Data set]. Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4034929\n\nFollain G, Jacquemet G: Combining StarDist and TrackMate example 3 - Flow chamber dataset (Version 1) [Data set]. Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4034939\n\nJacquemet G, Fazeli E, Laine RF, et al.: Combining StarDist and TrackMate - Archived source code (Version v1). Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4091474"
}
|
[
{
"id": "73864",
"date": "11 Nov 2020",
"name": "Lachlan W. Whitehead",
"expertise": [
"Reviewer Expertise bioimage analysis"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article presents a pipeline for analyzing cell migration in a variety of contexts by combining several complimentary techniques. Utilising stardist for cell detection (provided cells are round/have nuclei staining), and trackmate for connecting the detected nuclei over time - a start-to-finish protocol is described allowing a microscopist with little image analysis knowledge to be able to quantify their experiment.\n\nThe authors do an admirable job of describing the required steps of the analysis pipeline, including an introduction to Jupyter Notebooks and the ZeroCostDL4Mic workflows to train a custom Stardist model. They also provide a FIJI macro for batch analysis, potentially saving a researcher many hours of human analysis time.\n\nIndeed, as all of these methods are published and validated already my only (small) criticism has to do with the training of the stardist models. As the authors rightly note, this is the most time consuming part of the analysis and as a result I would have liked to see some mention of how much manual time is required. For instance, while the article suggests training a small 20 image dataset and then using transfer learning to speed up the remaining annotation, the number of remaining images is not described. While the amount of training required is likely to vary across experiments, I think readers would benefit from knowing when considering this pipeline how many cells (rather than image fields of view) they are likely to be required to manually annotate. Perhaps for each dataset where a model was trained the authors could specify the size of the training dataset (in both images and number of cells).\n\nOverall, this is a clear description of several powerful tools being combined into a very useful and versatile workflow.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "6175",
"date": "21 Dec 2020",
"name": "Guillaume Jacquemet",
"role": "Author Response",
"response": "We thank the reviewer for his positive comments. The reviewer makes an excellent point. We have now added in the main text of the manuscript the size of the training dataset, including both the number of images as well as the total number of annotated cells. The training dataset size are as follow: Breast cancer cell dataset (72 paired images, 24500 labelled objects); T cell dataset (Training dataset after 5x augmentation: 209 paired images, 31200 labelled objects); Flow chamber dataset (Training dataset: 57 paired images, 3680 labelled objects). We estimate that it takes between 4-8h of work to annotate a completely new training dataset."
}
]
},
{
"id": "73865",
"date": "11 Nov 2020",
"name": "Ignacio Arganda-Carreras",
"expertise": [
"Reviewer Expertise Computer Vision",
"Bioimage Analysis."
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors propose the combination of open-source tools (namely StarDist and TrackMate) for the automatic tracking of cells in fluorescence and brightfield images in 2D. Moreover, they provide a step-by-step workflow to process videos in a batch mode using exclusively free tools. They evaluate its performance by comparing the results obtained using such a workflow with those of manual tracking from an already published public dataset.\nThe paper is very well written, concise and straight to the point, with a special emphasis on reproducibility.\nAs pointed out in the conclusions, one of the limitations of the proposed approach is inherent to the type of objects that StarDist can properly segment (basically round). However, that doesn't prevent this pipeline from being extremely useful for a broad spectrum of cell tracking problems. Moreover, the adaptation of the pipeline to 3D images seems pretty straightforward.\nSomething interesting that is not mentioned in the paper is how well the workflow would perform in the presence of cell divisions of apoptosis. That could be easily tested using some of the datasets from the Cell Tracking Challenge (http://celltrackingchallenge.net/2d-datasets/).\nMinor comments:\nPlease homogenize how you write \"deep learning\", which appear sometimes as \"Deep-Learning\" and sometimes as \"deep learning\".\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "6176",
"date": "21 Dec 2020",
"name": "Guillaume Jacquemet",
"role": "Author Response",
"response": "We thank the reviewer for his positive comments. The reviewer highlights an excellent point when asking about the ability of the pipeline to cope with cell division. This is indeed a critical concern when tracking cells for an extended period of time. Numerous cell divisions were actually detected in our DCIS.com test dataset. In this case, our pipeline worked very well, and division events were both recognized, and the tracks splits after divisions. This is due to two factors 1) we trained the StarDist model also to recognize mitotic cells and 2) we enabled track splitting in TrackMate. We have now added a sentence in the manuscript to reflect this."
}
]
},
{
"id": "73862",
"date": "16 Nov 2020",
"name": "Vinay Swaminathan",
"expertise": [
"Reviewer Expertise Cell migration",
"mechanobiology",
"Cancer."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this article, Fazeli et al. combine deep-learning segmentation tools with open access cell tracking platforms to show the feasibility for automated cell tracking and large population motility analysis. Using this approach, the authors here clearly describe and validate this process in fluorescent movies of collective cell migration as well as single cell bright-field cell migration datasets. While the description of methods is very clear (and special kudos to the authors for transparency and availability), and its applicability very apparent, a few other potential points to consider are as follows:\nA comparison on a given dataset between trackmate with StarDist based segmentation versus canonical segmentation methods of the nuclei. How much better does it perform? This could be in terms of accuracy/precision or time taken to analyze.\n\nRelated to point 1: The power of StarDist based nuclear segmentation is its ability in crowded environments as well as where there is poor signal in the fluorescent channels. Is there a way to explicitly show that when there is poor signal or bleaching during a movie, StarDist combined with trackmate does a better job?\n\nOne of the hardest objects to segment are bright-field objects which is why the field has been relying on hand tracking of cell migration movies taken with bright-field. I think the authors should underscore this and highlight that this method is overcoming this big challenge enabling larger population analysis.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "6177",
"date": "21 Dec 2020",
"name": "Guillaume Jacquemet",
"role": "Author Response",
"response": "We thank the reviewer for his positive comments. Regarding the performance of Deep Learning and StarDist for nuclear segmentation, this topic has been extensively covered by others, and I would recommend the excellent paper from Caicedo et al. (Cytometry, 2019). In addition, StarDist was demonstrated to perform very well on images containing dividing cells, extensive noise and nuclei deformation. This has also been our experience. We would also add that the main reason why we started to use StarDist was that we could not segment our cell migration movies with enough accuracy using intensity-based thresholding. Because of this issue, automated tracking was not feasible. We would argue that the best way to ensure that the results generated by automated tracking approach are of good quality are 1) the visual inspection of the tracks and 2) comparing the results obtained compared to manual tracking. We fully agree with the reviewer that the automatic tracking of brightfield movies is very challenging and that the field often relies on manual analysis. We have now added a few sentence to the text to highlight this further. We indeed hope that the protocol published here may help to alleviate some of this burden."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1279
|
https://f1000research.com/articles/9-1491/v1
|
21 Dec 20
|
{
"type": "Research Article",
"title": "Antibacterial activity in secondary metabolite extracts of heterotrophic bacteria against Vibrio alginolyticus, Aeromonas hydrophila, and Pseudomonas aeruginosa",
"authors": [
"Jarod Setiaji",
"Feli Feliatra",
"Hilwan Yuda Teruna",
"Iesje Lukistyowati",
"Indra Suharman",
"Zainal Abidin Muchlisin",
"Teuku Iskandar Johan",
"Jarod Setiaji",
"Hilwan Yuda Teruna",
"Iesje Lukistyowati",
"Indra Suharman",
"Zainal Abidin Muchlisin",
"Teuku Iskandar Johan"
],
"abstract": "Background: Disease causing bacteria such as Vibrio alginolyticus, Aeromonas hydrophila, and Pseudomonas aeruginosa present a problem for fish farming. Treatment to remove them are generally carried out using antibiotics which have side effects on fish, the environment and humans. However, the use of antibacterial compounds derived from heterotrophic bacteria serve as a good alternative for antibiotics. Therefore, this study aimed to explore antibacterial activity in the secondary metabolite extracts of heterotrophic bacteria against Vibrio alginolyticus, Aeromonas hydrophila, and Pseudomonas aeruginosa. Methods: Heterotrophic bacteria namely Bacillus sp. JS04 MT102913.1, Bacillus toyonensis JS08 MT102920.1, Bacillus cereus JS10 MT102922.1, Bacillus sp. JS11 MT102923.1, Pseudoalteromonas sp. JS19 MT102924.1, Bacillus cereus JS22 MT102926.1, and Bacillus sp. strain JS25 MT102927.1 were used in this study. The sequences of these bacteria have been deposited and are available from NCBI GenBank. Each heterotrophic bacterium was cultured on 6L nutrient broth for 8 days, and extracts produced using ethyl acetate to obtain their secondary metabolites. These extracts were tested for their phytochemical contents using FT-IR and also tested for their inhibitory property in pathogenic bacteria by agar diffusion method. Results: Phytochemical test results showed that the seven heterotrophic bacterial isolates produced terpenoid compounds. Based on the inhibitory test, the secondary metabolite extracts from Bacillus sp strain JS04 had the highest inhibitory effect on the growth of pathogenic bacteria namely, V. alginolyticus (17.5 mm), A. hydrophila (16.8 mm), and P. aeruginosa (17.3 mm). Conclusion: It was concluded that the secondary metabolite extracts of heterotrophic bacteria inhibit the growth of V. alginolyticus, A. hydrophila, and P. aeruginosa.",
"keywords": [
"antibacterial",
"fish pathogens",
"heterotrophic bacteria",
"secondary metabolites"
],
"content": "Introduction\n\nBacteria diseases in fish stocks constitute a major problem for fish farming, since they cause significant economic losses1–3. Common pathogenic bacteria that affect fish include Vibrio alginolyticus, Aeromonas hydrophila and Pseudomonas aeruginosa4–9. V. alginolyticus is a gram-negative bacteria which is an opportunistic pathogen in marine animals10–13. Bacterial diseases cause different fish infections such as, exophthalmia, ulcers, septicemia, and corneal damage14–16. Aeromonas hydrophila is found to be the main cause of the septicemia epidemic in freshwater fish17,18. Its outbreak causes tissue damage of the spleen, gills, and the fish's stomach19. A. hydrophila is found to frequently infect various fish species namely, catfish (Ictalurus punctatus)20, carp (Cyprinus carpio) and catfish (Pangasius hypophthalmus)21, tilapia (Oreochromis niloticus)22, salmon (Oncorhynchus masou masou)23, snapper (Lates calcarifer)24, striped snakehead (Channa striata)25, cod (Gadus macrocephalus), and tank goby (Glossogobius guris)26. Meanwhile, P. aeruginosa is found to infect freshwater and marine fish27,28, with infection being characterized by the expression of red spots due to bleeding, skin darkens, loose scales, protruding eyes, fin erosion29, behavioural changes due to disruption of locomotor activity30, and abnormal swimming31.\n\nBacteria disease treatment is generally carried out using antibiotics, however, these can have adverse effects on the fish and their environment32–37. The accumulation of antibiotics in the fish increase the risk of bacterial resistance38,39. Escherichia coli bacteria isolated from the digestive organs of catfish showed high resistance levels towards tetracycline, ampicillin, and chloramphenicol40. Therefore, it is necessary to explore natural compounds with antibacterial activity41. Sea water is a potential source of heterotrophic bacteria that produce antimicrobial compound42, and have probiotic activity43\n\nSea bacteria such as, Bacillus sp. B. cereus, B. toyonensis, and Pseudoalteromonas sp., are known to inhibit the growth of pathogenic bacteria namely, V. alginolyticus, A. hydrophila, and Pseudomonas sp43. They also produce antimicrobial compounds such as, Pseudoalteromonas44. Pseudoalteromonas piscicida produces antimicrobial substances that inhibit the growth of different pathogenic bacteria namely, Vibrio vulniosis45, Bacillus sp46, B. pumilus47, and B. subtilis48,49. Bacillus amyloliquefaciens shows antibacterial activity towards pathogenic bacteria such as, Aeromonas hydrophila, Vibrio harveyi, V. vulnificus, and V. parahaemolyticus50. Meanwhile, Bacillus subtilis shows antibacterial activity towards the pathogens Vibrio parahaemolyticus, V. vulnificus, and Aeromonas hydrophila51.\n\nHeterotrophic bacteria extracted from Riau sea waters were examined and found to able to inhibit the activity of pathogenic bacteria Aeromonas salmonicida, Edwarsiela tarda and Edwarsiela ictaluri as previously reported by Setiaji et al.52. However, the antibacterial activity of these heterotrophic bacteria extracted from Riau sea waters on the pathogenic bacterial namely, Vibrio alginolyticus, Aeromonas hydrophila, and Pseudomonas aeruginos have never been examined for its potential against pathogenic bacteria. Therefore, this study aims to explore antibacterial activity in secondary metabolite extracts of heterotrophic bacteria isolated from Riau sea water, against pathogenic bacteria namely, V. alginolyticus, A. hydrophila, and P. aeruginosa.\n\n\nMethods\n\nThe heterotrophic bacteria isolates were collected from sea waters in Sungai Pakning Bengkalis Regency Riau Province Indonesia (North latitude 01o21’36,8” and East longitude 102o09’34,1”). 1 liter of the sea water was collected at 50 cm depth by using a water sampler (Tiolan Lab, type: WSV-BIT22), then was transferred into a sample bottle and was put into a coolbox filled with ice at 15°C, before being transported by car for 1 hour to the laboratory. The heterotrophic bacteria was cultured using nutrient Agar (NA; Merck-1.05450.0500). The heterotrophic bacteria cultured were used for an antagonist test against pathogen bacteria. The antagonist test procedure is as follows, 1 ml of pathogenic inoculants (OD 600nm = 0.08–0.1) (OD measured with Thermo scientific, Genesys 10S UV-Vis) was added to 15 ml liquid nutrient Agar media at 50ºC, then homogenized, and poured into a petri dish to solidify. Furthermore, Oxytetracycline antibiotic disc paper (Oxoid, CT0041B, OT30 mcg) was used as the positive control, while 30 µl aquades (Kimiapedia id-1720602804) was dripped to a disc paper (Macherey-nagel, MN827 ATD) as the negative control. 30 µl heterotrophic bacterial isolate taken from bacteria culture in nutrient Broth (NB; Merck-1.05443.0500) was dripped to a disc paper and incubated at 30°C for 24 hours. The inhibitory power of heterotrophic bacterial isolate was measured from the diameter of clear zone formed around the disc. From the antagonist test, eight isolates with the best inhibition were collected, and the heterotrophic bacteria was identified using 16S rDNA technique43. The sequenced products were run through BLAST (NCBI Basic Local Alignment Search Tool) and registered to GenBank.\n\nThe pathogenic bacteria were obtained from the collection at the Marine Microbiology Laboratory of the Faculty of Fisheries and Marine Science, University of Riau, Indonesia. The heterotrophic and pathogenic bacteria were cultured on the nutrient Agar (NA; Merck-1.05450.0500). The cultured medium was sterilized in an autoclave at a pressure of 15 psi and 121°C for 15 minutes. After 1 hour at room temperature, the medium was inoculated by the heterotrophic bacteria and the pathogenic bacteria. Then the bacteria were incubated in an incubator (Memmert, Model 30–1060) at 30°C for 24 hours.\n\nPrevious studies showed that eight heterotrophic bacterial isolates possessed the potential to produce pathogens. Seven of these species were used in this study namely, Bacillus sp. JS04 MT102913.1, Bacillus toyonensis JS08 MT102920.1, Bacillus cereus JS10 MT102922.1, Bacillus sp. JS11 MT102923.1, Pseudoalteromonas sp. JS19 MT102924.1, Bacillus cereus JS22 MT102926.1 and Bacillus sp. strain JS25 MT102927.1 have been deposited in GenBank.\n\nThen, each bacterium was cultured in a 6 L nutrient Broth (NB; Merck-1.05443.0500) diluted with sea water of salinity 29 ppt and aerated for 8 days. After this, the bacteria were mixed with ethyl acetate (P.a) at ratio 1:1 and shaken vigorously to homogenize. Subsequent filtering was performed until a clear filtrate was obtained using funnel and filter paper (Whatman 41, no. 1441–125), and evaporated with a rotary evaporator (Cole Parmer, N-1300) at 50°C and a speed of 50 rpm. This allowed thick secondary metabolite extracts to be obtained53.\n\nPhytochemical test was conducted on the secondary metabolite extracts of heterotrophic bacteria, which included tests for alkaloid, terpenoid, flavonoid, phenolic, and saponin compounds54.\n\nMayer reagent was prepared by adding 1.36 g HgCl2 (Merck, 1.04419. 0050) to 60 mL distillled water and 5 g Ki (Meck 1.05043.1000) to 10 mL distilled water. Both solutions were then mixed with a further 20 mL distilled water. The Liebermen – burchad reagent was prepared by mixing 97% H2SO4 (Merck 1.00731.2500) and 100% CH3COOH (Merck 1.00063.2500).\n\nAlkaloids were tested for using using 10 mg heterotrophic bacteria extract and 250 µL Mayer reagent.\n\nThe terpenoid was tested using 10 mg heterotrophic bacteria extract, 10 drops of CH3COOH, and 3 drops of H2SO4.\n\nFlavonoid tests were performed using 10 mg heterotrophic bacteria extract added to 5 mL distilled water. This was then boiled before adding 0.05 g Mg (Merck 1.05815.1000) and 10 drops of 37% HCl (Merck 1.00317. 2500), the mixture was then shaken for one minute.\n\nPhenolic compounds were tested by using 10 mg heterotrophic bacteria extract combined with 500 µL 5% FeCl3 (Merck 1.03943.0250).\n\nSaponin compounds were tested for using 10 mg heterotrophic bacteria extract added to 5 mL distilled water which was then shaken for 1 minute. 150 µL 1N HCl (Merck 1.00317. 2500) was then added, and shaken for another minute.\n\nA positive alkaloid test was indicated by the formation of a white precipitate after adding Mayer regent. A positive terpenoid test was indicated by the formation of a red colour. A positive flavonoid test was indicated by a red colour change. Phenolic compounds were indicated by a blue colour change. Saponin compounds were indicated by a foam forming.\n\nMeanwhile, to determine the functional groups in secondary metabolite extracts, FT-IR (Shimadzu, IR prestige-21, IR solution software ver. 1.1) spectroscopy analysis was performed. This was conducted by crushing 1 mg of each extract, added to KBr (Merck-1.04950.0500), and mixed vigorously until homogenized. This mixture was then measured for infrared absorbance at 4500–450 cm wavelength.\n\nThe secondary metabolite extracts of heterotrophic bacteria obtained were tested on pathogenic bacteria namely, V. algynolyticus, A. hydrophila, and P. aeruginosa using agar diffusion method, and 6 mm disc paper (Macherey-nagel, MN827 ATD)55. The procedure is as follows, 1 ml of pathogenic inoculants (OD 600nm = 0.08–0.1) (OD measured with Thermo scientific, Genesys 10S UV-Vis) added to 15 ml liquid nutrient agar media at 50°C, then homogenized, and poured into a petri dish to solidify. Furthermore, Oxytetracycline antibiotic disc paper (Oxoid, CT0041B, OT30 mcg) was used as the positive control, while methanol disc paper was the negative control. The metabolite extracts were then dissolved in 1 mg / mL methanol (P.a) and incubated at 30°C for 24 hours. The inhibitory power of heterotrophic bacterial extracts was measured from the diameter of clear zone formed around the disc.\n\nThe data were subjected to one-way analysis of variance followed by the Post Hoc Tukey multiple range test using R 4.0 software (GNU General Public License), p<0.05 is considered to indicate a statistically significant difference.\n\n\nResults\n\nPhytochemical test results of the metabolite extracts when added to Lieberman-Burchard reagents produced a red colour indicating the presence of terpenoids in the seven isolates. Meanwhile, the test for alkaloid, flavonoid, phenolic, and saponin compounds gave negative results.\n\nBased on infrared spectrum analysis, the secondary metabolite extracts of Bacillus sp. strain JS04 contained O-H alcohol, C-H aldehyde, O-H carboxylic acid, and C=C alkene groups. Bacillus toyonensis JS08 contained C-H alkanes, C=N nitriles, C=O carbonyl, and C-N amine groups. Bacillus cereus JS10 contained C-H alkanes, O-H carboxylic acid, C=C alkenes, and C-H alkanes groups. Bacillus sp. JS11 contained O-H alcohol, C-H alkanes, O-H carboxylic acids, and C=O carbonyl groups. Pseudoalteromonas sp. JS19 contained alcohol O-H, C-H alkanes, C=O carbonyl, and C=C alkenes groups. Bacillus cereus JS22 contain O-H alcohol, C-H alkane, C=O carbonyl, and C=C alkene groups. Bacillus sp. JS25 contain C-H alkanes, O-H carboxylic acids, O-H alcohols, and C=C alkenes groups (Table 156).\n\nThe results showed that the seven heterotrophic bacterial isolates inhibited the growth of pathogenic bacteria. The extracts inhibitory activity against pathogenic bacteria are shown in Table 256. The average inhibition zone diameter of the extracts against pathogenic bacteria namely, V. alginolyticus, A. hydrophila, and P. aeruginosa ranges from 9.3 to 17.5 mm, 9.3 to 16.8 mm, and 8.5 to 17.3 mm, respectively. This inhibitory zone activity was indicated by the presence of clear zones formed around the disc paper. The largest inhibition zone diameter of the extracts against the growth of pathogenic bacteria was derived from isolates of Bacillus sp. strain JS04 (17.5 mm) on V. alginolyticus, 17.3 mm on P. aeruginosa, and 16.8 mm on A. hydrophila.\n\nMean values with different superscripts in the same columns were significantly different (p < 0.05).\n\n\nDiscussion\n\nPhytochemical test results showed that the seven heterotrophic bacterial isolates produced terpenoids, which consist 5 carbon atoms or isoprene (C5) units. Microbes carry out biosynthesis by producing isopentyl pyrophosphate and dimethyl allyl pyrophosphate for terpenoid formation57. A significant relationship between terpenoids gene expression and isoprene production in Bacillus subtilis has previously been reported58.\n\nInfrared spectrum analysis provided information about the detected compounds in the mixture59. Metabolite extracts showed the presence of hydroxyl, aldehyde, carboxylic acid, alkene, alkane, carbonyl, and amine functional groups in these extracts. This indicated that the seven bacterial isolates produced terpenoids, while the functional groups contained in the terpenoids were namely, O-H hydroxyl, C-H aliphatic, carbonyl, C-H cyclic, and carboxylic acid60.\n\nThe result of inhibitory activity in the secondary metabolite extracts of Bacillus sp. strain JS04 showed the largest inhibition zone against the growth of pathogenic bacteria. The formation of clear zones on culture media indicated that heterotrophic bacteria produced terpenoid compounds for antibacterial purposes.\n\nThe terpenoid compounds contained several phytochemicals that possess antimicrobial activity61. For example, Terpenes and terpenoids have been reported to exert antimicrobial activity against a wide variety of bacteria, both Gram-positive and Gram-negative62. Terpenes cause membrane disruption through acting on lipophilic compound in the membrane63. Therefore, terpenoid compounds were able to prevent the forming of biofilm cell in the bacterium Streptococcus mutans60,64.\n\nThere are many antimicrobial compounds produced by sea bacteria especially from the Bacillus and Pseudoaltreromonas genus. For instance, B. pumilus produces antimicrobial compound against V. algynolyticus, V. anguillarum, Listeria monocytogenes and Staphylococcus aureus pathogens48. The Bacillus sp. from sea water produced chemical compound effective at preventing motility of V. Algynolyticus47. Bacillus subtilis produced antibacterial compound against Aeromonas hydrophila and Vibrio parahemolyticus pathogens49. The genus Pseudoalteromonas hosts 16 antimicrobial metabolite producers. To date, a total of 69 antimicrobial compounds are classified into alkaloids, polyketides, and peptides45. Furthermore, the bacterium Pseudoalteromonas rubra which was symbiotic with soft coral Sarcophyton sp. produced carotenoid pigments with antibacterial activity against Staphylococcus aureus65 and V. algynolyticus pathogens66.\n\n\nConclusion\n\nThe secondary metabolite extracts produced by the seven isolates of haterotrophic bacteria can inhibit the growth of pathogenic bacteria, namely V. alginolyticus, A. hydrophila, and P. aeruginosa. The secondary metabolite extracts of Bacillus sp. strain JS04 has the highest inhibitory activity against the growth of these three pathogenic bacteria.\n\n\nData availability\n\nFigshare: Antibacterial activity in secondary metabolite extracts of heterotrophic bacteria against Vibrio alginolyticus, Aeromonas hydrophila, and Pseudomonas aeruginosad Item. https://doi.org/10.6084/m9.figshare.12818798.v356\n\nThis project contains the following underlying data:\n\n- Data FT-IR activity in the secondary metabolite. Jarod Setiaji.pdf (Infrared spectrum of secondary metabolite extracts of heterotrophic bacteria)\n\n- Data Inhibitory activity in the secondary metabolite. Jarod Setiaji.xlsx (Inhibitory activity in the secondary metabolite extracts of heterotrophic bacteria against pathogenic bacteria)\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Acknowledgments\n\nThe authors are grateful to the laboratory assistants for supporting this research.\n\n\nReferences\n\nCai S, Cheng H, Pang H, et al.: AcfA is an essential regulator for pathogenesis of fish pathogen Vibrio alginolyticus. Vet Microbiol. 2018; 213: 35–41. PubMed Abstract | Publisher Full Text\n\nLiu W, Huang L, Su Y, et al.: Contributions of the oligopeptide permeases in multistep of Vibrio alginolyticus pathogenesis. MicrobiologyOpen. 2017; 6(5): e00511. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu H, Gu D, Cao X, et al.: Characterization of a new quorum sensing regulator luxT and its roles in the extracellular protease production, motility, and virulence in fish pathogen Vibrio alginolyticus. Arch Microbiol. 2012; 194(6): 439–452. PubMed Abstract | Publisher Full Text\n\nChen Y, Cai S, Jian J: Protection against Vibrio alginolyticus in pearl gentian grouper (Epinephelus fuscoguttatus × Epinephelus lanceolatu) immunized with acfA-deletion live attenuated vaccine. Fish Shellfish Immunol. 2019; 86: 875–881. PubMed Abstract | Publisher Full Text\n\nZhang D, Moreira GS, Shoemaker C, et al.: Detection and quantification of virulent Aeromonas hydrophila in channel catfish tissues following waterborne challenge. FEMS Microbiol Lett. 2016; 363(9): fnw080. PubMed Abstract | Publisher Full Text\n\nStratev D, Odeyemi OA: Antimicrobial resistance of Aeromonas hydrophila isolated from different food sources: A mini-review. J Infect Public Health. 2016; 9(5): 535–44. PubMed Abstract | Publisher Full Text\n\nLee J, Zhang L: The hierarchy quorum sensing network in Pseudomonas aeruginosa. Protein cell. 2015; 6(1): 26–41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZheng J, Tang XM, Wu RX, et al.: Identification and characteristics of aptamers against inactivated Vibrio alginolyticus. LWT–Food Sci Technol. 2015; 64(2): 1138–1142. Publisher Full Text\n\nMittal R, Khandwaha RK, Gupta V, et al.: Phenotypic characters of urinary isolate of Pseudomonas aeruginosa and their association with mouse renal colonization. Indian J Med Res. 2006; 123(1): 67–72. PubMed Abstract\n\nZhao Z, Liu J, Deng Y, et al.: The Vibrio alginolyticus T3SS effectors, Val1686 and Val1680, induce cell rounding, apoptosis and lysis of fish epithelial cells. Virulence. 2018; 9(1): 318–330. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYu Q, Li F, Wang Y, et al.: Isolation, identification and pathogenicity of Vibrio alginolyticus from marine cultured Trachinotus ovatus. in Beibu Gulf, Guangxi. Guangxi Sciences. 2018; 25(1): 68–73. Reference Source\n\nCai SH, Lu YS, Jian JC, et al.: Protection against Vibrio alginolyticus in crimson snapper Lutjanus erythropterus immunized with a DNA vaccine containing the ompW gene. Dis Aquat Organ. 2013; 106(1): 39–47. PubMed Abstract | Publisher Full Text\n\nAustin B: Vibrios as causal agents of zoonoses. Vet Microbiol. 2010; 140(3–4): 310–7. PubMed Abstract | Publisher Full Text\n\nWang YD, Wang YH, Hui CF, et al.: Transcriptome analysis of the effect of Vibrio alginolyticus. infection on the innate immunity-related TLR5–mediated induction of cytokines in Epinephelus lanceolatus. Fish Shellfish Immunol. 2016; 52: 31–43. PubMed Abstract | Publisher Full Text\n\nLuo G, Huang L, Su Y, et al.: flrA, flrB and flrC regulate adhesion by controlling the expression of critical virulence genes in Vibrio alginolyticus. Emerg Microbes Infect. 2016; 5(8): e85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRameshkumar P, Nazar AKA, Pradeep MA, et al.: Isolation and characterization of pathogenic Vibrio alginolyticus from sea cage cultured cobia (Rachycentron canadum (Linnaeus 1766)) in India. Lett Appl Microbiol. 2017; 65(5): 423–430. PubMed Abstract | Publisher Full Text\n\nXia H, Tang Y, Lu F, et al.: The effect of Aeromonas hydrophila infection on the non specific immunity of blunt snout bream (Megalobrama amblycephala). Cent Eur J Immunol. 2017; 42(3): 239–243. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChaix G, Roger F, Berthe T, et al.: Distinct aeromonas populations in water column and associated with Copepods from estuarine environment (Seine, France). Front Microbiol. 2017; 8: 1259. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAbdelhamed H, Ibrahim I, Baumgartner W, et al.: Characterization of histopathological and ultrastructural changes in Channel Catfish experimentally infected with virulent Aeromonas hydrophila. Front Microbiol. 2017; 8: 1519. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJubirt MM, Hanson LA, Hanson-Dorr KC, et al.: Potential for great egrets (Ardea alba) to transmit a virulent strain of Aeromonas hydrophila. among Channel Catfish (Ictalurus punctatus) culture ponds. J Wildl Dis. 2015; 51(3): 634–9. PubMed Abstract | Publisher Full Text\n\nRasmussen-Ivey CR, Hossain MJ, Odom SE, et al.: Classification of a hypervirulent Aeromonas hydrophila pathotype responsible for epidemic outbreaks in warm-water fishes. Front Microbiol. 2016; 7: 1615. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStratev D, Stoev S, Vashin I, et al.: Some varieties of pathological changes in experimental infection of carps (Cyprinus carpio) with Aeromonas hydrophila. Journal of Aquaculture Engineering and Fisheries Research. 2015; 1(4): 191–202. Publisher Full Text\n\nHan JE, Kim JH, Choresca C, et al.: Draft genome sequence of a clinical isolate, Aeromonas hydrophila SNUFPC-A8, from a Moribund Cherry Salmon (Oncorhynchus masou masou). Genome Announc. 2013; 1(1): e00133–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHamid R, Ahmad A, Usup G: Pathogenicity of Aeromonas hydrophila isolated from the Malaysian Sea against coral (Turbinariasp) and seabass (Lates calcarifer). Environ Sci Pollut Res Int. 2016; 23(17): 17269–76. PubMed Abstract | Publisher Full Text\n\nSamayanpaulraj V, Sivaramapillai M, Palani SN, et al.: Identification and characterization of virulent Aeromonas hydrophila Ah17 from infected Channa striata in river Cauvery and in vitro evaluation of shrimp chitosan. Food Sci Nutr. 2020; 8(2): 1272–1283. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCitterio B, Biavasco F: Aeromonas hydrophila virulence. Virulence. 2015; 6(5): 417–418. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMishra P, Jerobin SRSJ, Thomas J, et al.: Study on antimicrobial potential of neem oil nanoemulsion against Pseudomonas aeruginosa infection in Labeo rohita. Biotechnol Appl Biochem. 2014; 61(5): 611–619. PubMed Abstract | Publisher Full Text\n\nThomas J, Thanigaivel S, Vijayakumar S, et al.: Pathogenecity of Pseudomonas aeruginosa in Oreochromis mossambicus and treatment using lime oil nanoemulsion. Colloids Surf B Biointerfaces. 2014; 116: 372–377. PubMed Abstract | Publisher Full Text\n\nAltinok I, Kayis S, Capkin E: Pseudomonas putida infection in rainbow trout. Aquaculture. 2006; 261(3): 850–855. Publisher Full Text\n\nBaldissera MD, Souza CF, Santos RCV, et al.: Pseudomonas aeruginosa strain PAO1 infection impairs locomotor activity in experimentally infected Rhamdia quelen: Interplay between a stress response and brain neurotransmitters. Aquaculture. 2017a; 473: 74–79. Publisher Full Text\n\nBaldissera MD, Souza CF, Santos RCV, et al.: Blood brain barrier breakdown and myeloperoxidase activity in silver catfish experimentally infected with Pseudomonas aeruginosa. J Fish Dis. 2018; 41(2): 209–213. PubMed Abstract | Publisher Full Text\n\nRodriguez-Mozaz S, Vaz-Moreira I, Giustina SVD, et al.: Antibiotic residues in final effluents of European wastewater treatment plants and their impact on the aquatic environment. Environ Int. 2020; 140: 105733. PubMed Abstract | Publisher Full Text\n\nCao J, Zhang J, Ma L, et al.: Identification of fish source Vibrio alginolyticus and evaluation of its bacterial ghosts vaccine immune effects. Microbiologyopen. 2018; 7(3): e00576. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChuah LO, Effarizah ME, Goni AM, et al.: Antibiotic Application and Emergence of Multiple Antibiotic Resistance (MAR) in Global Catfish Aquaculture. Curr Environ Health Rep. 2016; 3(2): 118–127. PubMed Abstract | Publisher Full Text\n\nDi Cesare A, Luna GM, Vignaroli C, et al.: Aquaculture can promote the presence and spread of antibiotic-resistant enterococci in marine sediments. PLoS One. 2013; 8(4): e62838. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRomero J, Gloria C, Navarrete P: Antibiotics in aquaculture-use, abuse and alternatives. Health and Environment in Aquaculture. 2012; 159–202. Publisher Full Text\n\nDing C, He J: Effect of antibiotics in the environment on microbial populations. Appl Microbiol Biotechnol. 2010; 87(3): 925–41. PubMed Abstract | Publisher Full Text\n\nRasul MG, Majumdar BC: Abuse of antibiotics in aquaculture and it’s effects on human, aquatic animal and environment. The Saudi Journal of Life Sciences. 2017; 2(3): 81–88. Publisher Full Text\n\nOlesen I, Hasman H, Aarestrup MF: Prevalence of β-Lactamases among ampicillin-resistant Escherichia coli and Salmonella isolated from food animals in denmark. Microb Drug Resist. 2004; 10(4): 334–340. PubMed Abstract | Publisher Full Text\n\nDamayanti S, Kusdarwati R, Suprapto H: Bacterial resistance of Escherichia coli against antibiotics in Clarias batrachus digestion. AACL Bioflux. 2019; 12(6): 2195–2201. Reference Source\n\nSamuel L, Marian MM, Apun K, et al.: Characterization of Escherichia coli isolated from cultured catfish by antibiotic resistance and RAPD analysis. Int Food Res J. 2011; 18(3): 971–976. Reference Source\n\nAlekseeva BI, Danilovich KA, Valirievna KU: Antimicrobial Activity of Heterotrophic Bacterial Strains of Marine Origin. Undishapur Journal of Microbiology. 2013; 6(2): 166–175. Publisher Full Text\n\nSetiaji J: DNA Analysis of marine heterotrophic bacteria that are antagonistic towards pathogenic bacteria in fish. Proceedings of the 18th National Ocean Seminar. Implementation of marine and coastal resources research results in the framework of achieving national economic independence. Faculty of Engineering and Marine Sciences, Universitas Hang Tuah, Surabaya, Indonesia. 2018; c2.17-27. [In Indonesian].\n\nOffret C, Desriac F, Le Chevalier P, et al.: Spotlight on antimicrobial metabolites from the marine bacteria Pseudoalteromonas: chemodiversity and ecological significance. Mar Drugs . 2016; 14(7): 129. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRichards GP, Needleman DS, Watson MA, et al.: Whole-Genome Sequences of Two Pseudoalteromonas piscicida Strains, DE1-A and DE2-A, with Strong Antibacterial Activity against Vibrio vulnificus. Microbiol Resour Announc. 2019; 8(1): e01451–18. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXiu P, Liu R, Zhang D, et al.: Pumilacidin-Like Lipopeptides Derived from Marine Bacterium Bacillus sp. Strain 176 Suppress the Motility of Vibrio alginolyticus. Appl Environ Microbiol. 2017; 83(12): e00450-17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZidour M, Chevalier M, Belguesmia Y, et al.: Isolation and Characterization of Bacteria Colonizing Acartia tonsa Copepod Eggs and Displaying Antagonist Effects against Vibrio anguillarum, Vibrio alginolyticus and Other Pathogenic Strains. Front Microbiol. 2017; 8: 1919. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChakraborty K, Thilakan B, Raola VK: Polyketide family of novel antibacterial 7-O-methyl-5'-hydroxy-3'-heptenoate-macrolactin from seaweed-associated Bacillus subtilis MTCC 10403. J Agric Food Chem. 2014; 62(50): 12194–12208. PubMed Abstract | Publisher Full Text\n\nLi W, Tang XX, Yan X, et al.: A new macrolactin antibiotic from deep sea-derived bacteria Bacillus subtilis B5. Nat Prod Res. 2016; 30(24): 2777–2782. PubMed Abstract | Publisher Full Text\n\nChakraborty K, Thilakan B, Raola VK: Previously Undescribed Antibacterial Polyketides from Heterotrophic Bacillus amyloliquefaciens Associated with Seaweed Padina gymnospora. Appl Biochem Biotechnol. 2017a; 184(2): 716–732. PubMed Abstract | Publisher Full Text\n\nChakraborty K, Thilakan B, Raola VK: Antimicrobial polyketide furanoterpenoids from seaweed-associated heterotrophic bacterium Bacillus subtilis MTCC 10403. Phytochemistry. 2017b; 142: 112–125. PubMed Abstract | Publisher Full Text\n\nSetiaji J, Feliatra F, Teruna HY, et al.: Antimicrobial agents derived from heterotrophic bacteria againts pathogenic bacteria. IOP Conf Ser Earth Environ Sci. 2019; 348: 012029. Reference Source\n\nSateesh VN, Rathod JL: Selective isolation and antimicrobial activity of rare actinomycetes from mangrove sediment of Karwar. Journal of Ecobiotechnology. 2011; 3(10): 48–53. Reference Source\n\nDevi GK, Manivannan K, Thirumaran G, et al.: In vitro antioxidant activities of selected seaweeds from Southeast coast of India. Asian Pac J Trop Med. 2011; 4(3): 205–211. PubMed Abstract | Publisher Full Text\n\nFeliatra, Muchlisin ZA, Teruna HY, et al.: Potential of bacteriocins produced by probiotic bacteria isolated from tiger shrimp and prawns as antibacterial to Vibrio, Pseudomonas, and Aeromonas species on fish [version 1; peer review: 2 approved, 1 approved with reservations, 1 not approved]. F1000Res. 2018; 7: 415. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMuchlisin ZA, Setiaji J, Feliatra F, et al.: Antibacterial activity in secondary metabolite extracts of heterotrophic bacteria against Vibrio alginolyticus, Aeromonas hydrophila, and Pseudomonas aeruginosad Item. figshare. Dataset. 2020. http://www.doi.org/10.6084/m9.figshare.12818798.v3\n\nMoser S, Pichler H: Identifying and engineering the ideal microbial terpenoid production host. Appl Microbiol Biotechnol. 2019; 103(14): 5501–5516. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHess BM, Xue J, Markillie LM, et al.: Coregulation of Terpenoid Pathway Genes and Prediction of Isoprene Production in Bacillus subtilis Using Transcriptomics. PLoS One. 2013; 8(6): e66104. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKrishna GM, Muthukumaran M, Krisnhnamoorthy B, et al.: A Critical review on fundamental and pharmaceutical analysis of FT-IR spectroscopy. Int J Pharm. 2013; 3(2): 396–402. Reference Source\n\nWidyawati, Yanwirasti, Djong DH, et al.: Potential of terpenoid isolated fromMyrmecodia pendans AS antibacterial against Streptococcus mutans ATCC 2517. International Journal of Development Research. 2016; 6(11): 10350–10354. Reference Source\n\nBarbieri R, Coppo E, Marchese A, et al.: Phytochemicals for human disease: An update on plant-derived compounds antibacterial activity. Microbiol Res. 2017; 196: 44–68. PubMed Abstract | Publisher Full Text\n\nMahizan NA, Yang SK, Moo CL, et al.: Terpene Derivatives as a Potential Agent against Antimicrobial Resistance (AMR) Pathogens. Molecules. 2019; 24(14): 2631. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCowan MM: Plant products as antimicrobial agents. Clin Microbiol Rev. 1999; 12(4): 564–582. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGartika M, Pramesti HT, Kurnia D, et al.: A terpenoid isolated from sarang semut (Myrmecodia pendans) bulb and its potential for the inhibition and eradication of Streptococcus mutans biofilm. BMC Complement Altern Med. 2018; 18(1): 151. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWiguna AS, Kusmita L, Radjasa OK: Antibacterial activity test of carotenoid pigments from soft coral symbiont isolates from Sarcophyton sp. on the growth of Staphylococcus aureus ATCC 25923 bacteria. Indonesian Journal of Pharmaceutical Science and Technology. 2016; 3(3): 92–98. [In Indonesian]. Publisher Full Text\n\nSupardy NA, Ibrahim D, Nor SRM, et al.: Bioactive Compounds of Pseudoalteromonas sp. IBRL PD4.8 Inhibit Growth of Fouling Bacteria and Attenuate Biofilms of Vibrio alginolyticus FB3. Pol J Microbiol. 2019; 68(1): 21–33. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "76449",
"date": "14 Jan 2021",
"name": "Agung Damar Syakti",
"expertise": [
"Reviewer Expertise Analytical Environmental Chemistry"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nEvidence of Antibacterial activity in secondary metabolite extracts of heterotrophic bacteria against Vibrio alginolyticus, Aeromonas hydrophila, and Pseudomonas aeruginosa could be interesting for many applications in aquatic science. The finding may be useful for Indonesia’s fish farming activities. The study scientifically detailed enough. However, the author(s) should develop much better the occurrence of the specific spectrum thon FT-IR spectra related to the metabolite extracts. The authors should present the FTIR spectra. Thus, the manuscript could be accepted after such minor revision of the spectra figures addition showing the relative response of the functional groups in regard to the spectrum for target compound validation.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "76451",
"date": "26 Jan 2021",
"name": "Yuhanis Mhd Bakri",
"expertise": [
"Reviewer Expertise Natural product chemistry"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nOverall the authors have presented sound article with sufficient findings. Suggestion: Title could be amended to ethyl acetate extracts instead of secondary metabolite extracts. The word inhibitory power in method of inhibitory activity shall be amended appropriately. Isolate tests subtopic shall be amended to extraction. Functional groups investigation is rather general, although acceptable, a further study is needed to verify in length the chemical constituents of secondary metabolites extracts.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1491
|
https://f1000research.com/articles/9-1168/v1
|
24 Sep 20
|
{
"type": "Research Article",
"title": "Comparison of two lipid emulsions on interleukin-1β, interleukin-8 and fatty acid composition in infants post gastrointestinal surgery: a randomized trial",
"authors": [
"Meta Herdiana Hanindita",
"Roedi Irawan",
"I Dewa Gede Ugrasena",
"I. G. B. Adria Hariastawa",
"Roedi Irawan",
"I Dewa Gede Ugrasena",
"I. G. B. Adria Hariastawa"
],
"abstract": "Background: Nutritional support plays an essential role for recovery in infants who undergo gastrointestinal surgery. The current standard type of intravenous lipid emulsion (IVLE) used as parenteral nutrition is the mixture of medium-chain triglyceride (MCT) and long chain triglyceride (LCT) rich in ω-6. Studies showed that ω-6 is associated with higher level of proinflammatory cytokines, leading to increased mortality rate, morbidity rate, and postoperative recovery time. The latest generation of emulsion is a mixture of MCT, LCT, olive oil (OO), and fish oil (FO) which may optimize the ω6/ω3 ratio. This study aimed to compare the effect of MCT/LCT/OO/FO IVLE to standard IVLE on IL-1β, IL-8 and plasma fatty acid composition in infants who had undergone gastrointestinal surgery. Methods: A single-blind, randomised controlled, pretest-posttest design study was done in twelve subjects that were classified into two groups. Group 1 received standard IVLE, group 2 received MCT/LCT/OO/FO IVLE. The type of standard and MCT/LCT/OO/FO IVLE used in this study were Lipofundin 20% and SMOFlipid 20%, respectively, both administered for three consecutive days in 1-4 gram/kilogram/day. IL-1β and IL-8 were examined using ELISA while fatty acid composition was analyzed using gas chromatography tandem mass spectrometry (GC-MS). Statistical analyses were performed using SPSS for Mac 23. Results: No statistical difference was found in age, gender, birth weight and diagnosis, between both groups. Leukocyte level was significantly lower in MCT/LCT/OO/FO group 3 days after surgery (p=0.025). CRP level was lower in MCT/LCT/OO/FO group 3 days after surgery (p=0.01) and in changes within 3 days (p=0.016). There were no differences in IL-1β and IL-8 but ω-6 was higher in standard IVFE group on third day after surgery (p=0,048). Conclusion: MCT/LCT/OO/FO IVLE can significantly lower leukocyte, CRP and ω-6 levels and is comparable with standard IVLE on IL-1β & IL-8 levels in infants underwent gastrointestinal surgery.",
"keywords": [
"Parenteral Nutrition",
"Intravenous Lipid Emulsion",
"Interleukin-1Beta",
"Interleukin-8",
"Omega-3"
],
"content": "Introduction\n\nSurgical interventions may stimulate physiological inflammatory response as body’s attempt towards general recovery1. The balance of inflammatory response brings about good recovery, while excessive level of proinflammatory cytokines such as interleukin (IL) )-1β, IL-6, IL-8, and tumor necrosis factor (TNF)-α, may cause organ damage and severe complications, leading to the rise of postoperative mortality and morbidity rate2. De Mooij stated that IL-1β plays an important role in infection control, homeostasis, and tissue repair, while IL-8 plays an important role in inflammation and wound healing3–5.\n\nNutritional support is essential in wound healing and plays an important role in growth and development of an infant after undergoing gastrointestinal surgery6. Patients who could not receive oral and enteral nutrition for two days should be considered for parenteral nutrition7,8. The current standard type of fat emulsion used as parenteral nutrition is a mixture of medium-chain triglyceride (MCT) and soy oil enriched with long-chain triglyceride (LCT)9. This emulsion is rich in ω-6 and contains linoleic acid (LA, C18:2 ω -6) and also α-linolenic acid (ALA, C18:3 ω-3). Several studies showed that ω-6 is associated with impaired cell-mediated immunity and higher potential risk of elevated proinflammatory biomarkers and severe inflammatory response. These mechanisms may bring about the rise in mortality rate, morbidity rate and may also prolong the duration of treatment and postoperative recovery time10–12.\n\nThe latest generation of fat emulsion, SMOFlipid, is a mixture of MCT, LCT, olive oil (OO), and fish oil (FO), optimizing the ω6/ ω3 ratio. Some studies showed that OO exerts indirect anti-inflammatory effect by replacing ω-6 with oleic acid, while the addition of ω-3 from FO in soy-oil based fat emulsion may inhibit inflammatory reactions, i.e. reducing cytokine secretion and adhesion molecule expression and balancing the immune system, since it contains eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA)6,13–16. ω-3 may also act as a regulator of the immune system and mitigator of inflammation since it acts as a precursor for lipid mediator16. Our previous study showed lower levels of proinflammatory cytokines IL-6 and TNF-α in subjects who received the mixture of MCT/LCT/OO/FO compared with subjects who received MCT/LCT IVLE17. To our knowledge, there been no study regarding the effects of MCT/LCT/OO/FO IVLE on IL-1β and IL-8 levels and fatty acid composition in infants who undergo gastrointestinal surgery, compared to those who receive standard IVLE.\n\n\nMethods\n\nThis single-blind, randomized controlled, pretest-posttest design, parallel-group with 1:1 randomization study aimed to compare the effect of MCT/LCT/OO/FO IVLE to standard MCT/LCT IVLE on IL-1β, IL-8 levels and fatty acid composition in infants who had undergone gastrointestinal surgery. The primary outcomes of this study were IL-1β and IL-8 levels and fatty acid composition while the secondary outcomes were hemoglobin, leukocyte, C-reactive protein (CRP) and albumin levels. This study was conducted in April–July 2020. Our subjects were infants who had undergone gastrointestinal surgery at Soetomo General Hospital, Surabaya. Parents or legal guardian were recruited to the study through referrals from physicians who then contacted the research team. The number of subjects was determined based on the formula for calculating number of samples in non-comparative numerical analytical study, with type 1 error of 5% and type 2 error of 10%, and the minimum number samples for each group were 5 subjects18. Inclusion criteria for this study included subjects whose parents were willing for them to participate in this study, had undergone gastrointestinal surgery, and had received parenteral nutrition for at least 3 days. Exclusion criteria included subjects who had chronic diseases and subjects who were allergic to fish, egg, soy and/or nut proteins. Adverse effects were minimal or rare. However, if any harm was seen in the subjects, they would be recorded and reported at the end of the trial. Vital signs and allergic reaction signs were evaluated every 12 hours for all subjects.\n\nThis study was approved by the Ethical Committee of Dr. Soetomo General Hospital (No. 1922/KEPK/11/2020, March 27th 2020). Written informed consent obtained from the subjects’ parents or legal guardian.\n\nSubjects were randomly assigned to one of two IVLE groups following simple randomization procedures (computerized random numbers, https://www.random.org). Determination of whether a subject would get MCT/LCT standard IVLE or MCT/LCT/OO/FO was made by reference to a statistical series based on random sampling numbers drawn up by the primary investigator. Except the primary investigator and the pharmacist in charge, all subjects and staff were kept blind to IVLE assignment of the subjects. Eight-folded numbered papers were placed into opaque sealed envelopes to be chosen by the subjects’ parents or legal guardian. Investigators and pharmacy staff opened the envelope and used the lipid emulsion assigned to that patient. The trial is registered at ClinicalTrials.gov, number NCT04511299, registered on August 13th 2020. The protocol of this study can be seen at https://doi.org/10.17504/protocols.io.bknmkvc6. The flow of this research is shown on Figure 1.\n\nThe type of MCT/LCT IVLE used in this study was Lipofundin 20% which contained 50% coconut oil as the source of MCT and 50% soy oil as the source of LCT. Lipofundin 20% was given intravenously for three consecutive days after gastrointestinal surgery at 1–4 gram/kilogram/day dosing8. The type of MCT/LCT/OO/FO IVLE used in this study was Smoflipid 20%, which contained 30% soy oil as the source of LCT, 30% coconut oil as the source of MCT, 25% olive oil, and 15% fish oil. SMOFlipid® was given for three consecutive days in 1–4 gram/kilogram/day dosing8. The lipids used in this study were obtained from the manufacturers: Bbraun Indonesia (Lipofundin 20% I and Fresenius Kabi Indonesia (Smoflipid 20%). The comparison of the standard fat emulsion and ω-3-enriched fat emulsion is shown on Table 1.\n\nBefore surgery, blood samples of subjects were drawn (3–4 cm3) in order to measure their IL-1β, IL-8 levels, fatty acid composition, also hemoglobin, leukocyte, CRP and albumin levels. After surgery, subjects were assigned to either MCT/LCT IVLE or MCT/LCT/OO/FO IVLE for three consecutive days. On 3rd day (72 hours) after surgery, blood samples of subjects were drawn again in order to measure the same outcomes post-treatment. Once the samples arrive to the laboratory, samples are allowed to clot for 30 minutes at room temperature before centrifugation for 15 minutes at 1000 x g. L-1β and IL-8 levels in the serum were examined using the Quantikine HS ELISA by R&D Systems (Catalog Number HSLB00D and HS800) at wavelengths of 540 nm and 650 nm, respectively. Fatty acid composition was analyzed using gas chromatography tandem mass spectrometry (GC-MS). This examination was measuring the levels of free fatty acid in human serum quantitatively, including the arachidonic acid (AA)/EPA ratio, EPA, DHA, and AA.This method consists of two techniques, namely gas chromatography, which is a separation technique based on the degree of polarity and vapor point and mass spectrophotometry, which is a quadupole scanning spectrometer that can measure masses of 7-250 atomic mass units. Reagents used in this study were FAME Standard Mix (Supelco), GLC Nonadecanoic ISTD (Supelco), N-Hexane MS grade (Merck), Chloroform (Merck), Methanol Hyper Grade (Merck), Capillary Column and Helium Gas for GCMS.\n\nStatistical analyses performed in this study were Mann-Whitney U-test, Fishers’ Exact test, independent sample t-test and chi-square test using SPSS for Mac 23.0. The analyses of the IL-1β, IL-8 levels, fatty acid composition, also hemoglobin, leukocyte, CRP and albumin levels will be done by Mann-Whitney U-test or independent sample T-test as appropriate to test the significance between the two groups. A p-value less than or equal to 0.05 will be considered statistically significant. The analysis of subjects’ characteristic will be done by Mann-Whitney U-test, independent sample T-test, Fisher’s exact test or chi-square test as appropriate to test the significance between the two groups.\n\n\nResults\n\nThis study enrolled 12 subjects at Soetomo General Hospital Surabaya who had undergone gastrointestinal surgery and met the inclusion and exclusion criteria. The recruitment flow of the subjects is shown on Figure 2. The subjects were classified into two groups: group 1 received intravenous MCT/LCT lipid emulsion, and group 2 received intravenous MCT/LCT/OO/FO lipid emulsion. Subject characteristics are shown in Table 2.\n\n*Mann-Whitney test\n\n** Fishers’ Exact test\n\n*** Independent Sample Test\n\n****Chi-square test\n\nNo statistical difference was found in age, gender, birth weight, and diagnosis between both groups. De-identified subject characteristics, alongside all parameters measured in this study, are available as Underlying data19.\n\nMean IL-1β levels among subjects are depicted in Table 3; no difference was found in IL-1β levels between both groups before surgery (p = 0.873) and on day three after surgery (p = 0.873). We also did not find any difference in changes in IL-1β levels within 3 days (p = 0.906) in both groups.\n\n*Independent sample t-test.\n\n**Mann-Whitney U-test.\n\na Difference between groups before surgery.\n\nb Difference between groups 3 days post-surgery.\n\nc Difference between groups on changes within 3 days.\n\nFurthermore, there was no significant difference in mean IL-8 levels between both groups before surgery (p = 0.688) and on day three after surgery (p = 0.494), and no difference in IL-8 levels changes within 3 days (p = 0.837) in both groups.\n\nThe analysis of fatty acid composition is shown on Table 4. No significant differences were observed in ω6/ω3 ratio, AA/EPA ratio, and EPA, DHA, and AA levels between both groups. Nevertheless, ω-6 level was significantly lower in MCT/LCT/FO/OO IVLE group on third day after surgery group (p=0,048) compared to the standard IVLE.\n\n*Independent Sample Test\n\n**Mann-Whitney Test\n\nAA: arachidonic acid, EPA: eicosapentaenoic acid, ALA: alpha-linolenic acid, DHA: docosapentaenoic Acid, LA: linoleic acid, GLA: gamma-linolenic acid, DGLA: dihomo-gamma-linolenic-acid, OA: oleic acid, MA: myristic acid, PA: palmitic acid, SA: stearic acid, MUFA: monounsaturated fatty acids, PUFA: polyunsaturated fatty acids.\n\nThe laboratory parameters are shown in Table 5. According to preoperative laboratory assessment, no statistical difference was found in hemoglobin, leukocyte, CRP and albumin levels between both groups. There was a statistically significant difference in leukocytes between both groups 3 days after surgery (p=0.025). CRP level was significantly lower in MCT/LCT/OO/FO group 3 days after surgery (p=0.01) and in changes within 3 days (p=0.016) compared to the standard MCT/LCT IVLE.\n\n*Independent Sample Test\n\n**Mann-Whitney Test\n\naDifferences between the MCT/LCT group and MCT/LCT/OO/FO before surgery\n\nb Differences between the MCT/LCT group and MCT/LCT/OO/FO 3 days post surgery\n\nc Differences between the MCT/LCT group and MCT/LCT/OO/FO in changes within 3 days\n\nThere were no adverse effects reported from all subjects in the study.\n\n\nDiscussion\n\nA decrease in mean leukocyte levels in the ω-3-enriched IVLE group was observed on third day after surgery. This result is in accordance with some previous studies. Wei et al. observed significant declines in leukocyte and CRP levels on patients who received ω-3-enriched IVLE for 6 days after undergoing gastric tumor resection20. Wang et al. also observed a significant decline in mean CRP level in subjects who received ω-3-enriched IVLE for 5 days after undergoing surgical interventions for acute pancreatitis21. A systematic review stated that fish oil-enriched IVLE is associated with a reduction in CRP level in patients with malignancy after undergoing gastrointestinal surgery22.\n\nThe content of EPA and DHA in ω-3 lipids may inhibit inflammatory pathways in several ways, such as inhibiting chemotaxis of leukocytes, expression of adhesion molecules, and adhesive endothelial-leukocyte interaction16.\n\nCRP is an acute-phase protein synthesized by IL-6 induction from hepatocytes. The CRP level spikes in acute traumatic condition, e.g. after undergoing surgical intervention. CRP levels reflect rapid changes which occur in inflammatory conditions. A study showed that in the majority of patients, CRP levels rise for 3-12 hours after surgery, peaking at 24-72 hours, and return to baseline in 2 weeks after surgery22.\n\nThis study did not find any significant difference in IL-1β dan IL-8 levels between the two groups. This result is in accordance with some previous studies, including Ma et al. who did not find any difference in mean preoperative and postoperative (day-6) IL-1β levels between subjects who received intravenous MCT/LCT lipid emulsion and subjects who received intravenous MCT/LCT/OO/FO lipid emulsion for five consecutive days. Nevertheless, several studies yielded contrasting results23.\n\nA study by Wei et al. on 48 subjects who had undergone gastrointestinal tumor resection also found a significant reduction in mean IL-1β level on group receiving intravenous LCT/FO lipid emulsion compared to group receiving intravenous LCT lipid emulsion20. In addition, Han et al., on 38 postoperative subjects in surgical intensive care unit, found a significant reduction in mean IL-1 and IL-8 levels on subjects who received MCT/LCT/FO lipid emulsion for 7 days, compared to subjects who received MCT/LCT lipid emulsion after undergoing surgical interventions24.\n\nA proposed explanation to why our result is inconsistent with most of the previous studies is that in our study, intravenous lipid emulsion was only given for 3 days, while in other studies, which yielded significant reduction in mean proinflammatory cytokine levels, intravenous lipid emulsion was given for at least 6 days. In this study, levels of IL-1β and IL-8 were examined preoperatively and 72 hours after surgery. According to Lin and Lowry, systemic inflammation, which occurs after surgery, may trigger proinflammatory and anti-inflammatory cytokines. Among all types of cytokines, TNF-α is the earliest to emerge, followed by IL-6 as the cytokine with the highest level amongst all. TNF-α and IL-6 levels peak in 1-2 hours after surgery25. Our previous study showed a significant difference in mean IL-6 levels between subjects receiving MCT/LCT and subjects receiving MCT/LCT/OO/FO17. Lin and Lowry stated that the half-life of IL-1β in systemic circulation is less than 10 minutes, making it more difficult to detect during stressful periods than TNF-α. Proinflammatory cytokine mediators, such as IL-8, are released as part of the inflammatory cascade initiated by IL-125.\n\nThis study also observed that ω-6 levels in the ω-3-enriched IVLE group was lower than the standard MCT/LCT IVLE group, on third day after surgery. This result is in accordance with several previous studies. Skouroliakou et al. found a significantly lower mean ω-6 level in preterm neonates who received ω-3 enriched IVLE for 15 and 30 days compared to soybean oil on the third day after abdominal surgery26. Grim et al. showed a significant decline in ω-6 levels in 33 adult patients after major abdominal surgery who received ω-3-enriched fat emulsion for 6 days27. The composition of fatty acids in cell membrane phospholipids has a significant role on cellular responses and cell function. Membrane order and lipid raft assembly are affected by the fatty acid makeup of membrane phospholipids. The fatty acid composition of the second messengers that are obtained from membrane phospholipid influences their biological activity and potency. Fatty acids that released from membrane phospholipids upon cellular activation are forming some lipid mediators28. Nevertheless, our results on profiling of phospholipid fatty acid composition, such as ω6/ω3 ratio, EPA level, and DHA level, contradict previous studies which found a significant decline in ω-3-enriched fat emulsion group27,29–31. This discrepancy might be due to dissimilarity of subjects’ characteristics, and the duration of parenteral nutrition administration. In those studies, the standard IVLE used were 100% LCT/soybean oil-based lipid emulsion, not MCT/LCT IVLE like in our study.\n\nTo our knowledge, this is the first study in Indonesia to compare the effect of MCT/LCT/OO/FO IVLE with MCT/LCT IVLE on proinflammatory IL-1β and IL-8 levels and fatty acid composition in infants who had underwent gastrointestinal surgery. Further studies are needed to determine the difference in pathogenesis between adults and infants after undergoing gastrointestinal surgery, which may be associated with the difference in effects of MCT/LCT/OO/FO IVLE on their profiling of fatty acid compositions.\n\nThere were no adverse event or serious adverse event reported in this study. Limitation of this study include that it is a single-centre study with a small sample size. Our study did not have a long period of follow-up and is not a double-blind study.\n\n\nConclusion\n\nIn infants who underwent gastrointestinal surgery, MCT/LCT/OO/FO IVLE can significantly lower leukocyte, CRP and ω-6 levels, and is comparable with standard IVLE on IL-1β & IL-8 levels.\n\n\nData availability\n\nFigshare: Data Set Comparison of Two Lipid Emulsions on Interleukin-1β, Interleukin-8 and Fatty Acid Composition in Infants Post Gastrointestinal Surgery: A Randomized Trial. https://doi.org/10.6084/m9.figshare.12906320.v219.\n\nThis project contains the underlying data for the study in SAV and CSV formats.\n\nFigshare: CONSORT checklist for ‘Comparison of two lipid emulsions on interleukin-1β, interleukin-8 and fatty acid composition in infants post gastrointestinal surgery: a randomized trial’. https://doi.org/10.6084/m9.figshare.12906395.v132.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "References\n\nParuk F, Chausse JM: Monitoring the post surgery inflammatory host response. J Emerg Crit Care Med. 2019; 3: 47. Publisher Full Text\n\nHsing C, Wang J: Clinical implication of perioperative inflammatory cytokine alteration. Acta Anaesthesiol Taiwan. 2015; 53(1): 23–8. PubMed Abstract | Publisher Full Text\n\nde Mooij CEM, Netea MG, Van Der Velden WJF, et al.: Targeting the interleukin-1 pathway in patients with hematological disorders. Blood. 2017; 129(24): 3155–64. PubMed Abstract | Publisher Full Text\n\nQazi BS, Tang K, Qazi A: Recent advances in underlying pathologies provide insight into interleukin-8 expression-mediated inflammation and angiogenesis. Int J Inflamm. 2011; 2011: 908468. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDobreva I, Waeber G, James RW, et al.: Interleukin-8 secretion by fibroblasts induced by low density lipoproteins is p38 MAPK-dependent and leads to cell spreading and wound closure. J Biol Chem. 2006; 281(1): 199–205. PubMed Abstract | Publisher Full Text\n\nJiang W, Chen G, Zhang J, et al.: The effects of two mixed intravenous lipid emulsions on clinical outcomes in infants after gastrointestinal surgery: a prospective, randomized study. Pediatr Surg Int. 2019; 35(3): 347–55. PubMed Abstract | Publisher Full Text\n\nBurrin DG, Stool NB, De Pipaóne MS: Impact of new-generation lipid emulsions on cellular mechanisms of Parenteral Nutrition - Associated Liver Disease. Adv Nutr. 2014; 5(1): 82–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKoletzko B, Goulet O, Hunt J, et al.: Guidelines on paediatric parenteral nutrition of the European Society of Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) and the European Society for Clinical Nutrition and Metabolism (ESPEN), supported by the European Society of Paediatric Research (ESPR). J Pediatr Gastroenterol Nutr. 2005; 41 Suppl 2: S1–87. PubMed Abstract | Publisher Full Text\n\nSungurtekin H, Degirmenci S, Sungurtekin U, et al.: Comparison of the effects of different intravenous fat emulsions in patients with systemic inflammatory response syndrome and sepsis. Nutr Clin Pract. 2011; 26(6): 665–71. PubMed Abstract | Publisher Full Text\n\nWanten GJ, Calder PC: Immune modulation by parenteral lipid emulsions. Am J Clin Nutr. 2007; 85(5): 1171–84. PubMed Abstract | Publisher Full Text\n\nWaitzberg DL, Torrinhas RS: Fish oil lipid emulsions and immune response: what clinicians need to know. Nutr Clin Pract. 2009; 24(4): 487–99. PubMed Abstract | Publisher Full Text\n\nSala-Vila A, Barbosa VM, Calder PC: Olive oil in parenteral nutrition. Curr Op Clin Nutr Metabol Care. 2007; 10(2): 165–74. PubMed Abstract | Publisher Full Text\n\nPradelli L, Mayer K, Muscaritoli M, et al.: n-3 fatty acid-enriched parenteral nutrition regimens in elective surgical and ICU patients: A meta-analysis. Crit Care. 2012; 16(5): R184. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang J, Yu J, Kang W, et al.: Superiority of a fish oil-enriched emulsion to medium-chain triacylglycerols/long-chain triacylglycerols in gastrointestinal surgery patients: A randomized clinical trial. Nutrition. 2012; 28(6): 623–9. PubMed Abstract | Publisher Full Text\n\nMing-wei Z, Da-nian T, Jiang H, et al.: Impact of fish oil enriched total parenteral nutrition on elderly patients after colorectal cancer surgery. Chin Med J (Engl). 2012; 125(2): 178–81. PubMed Abstract | Publisher Full Text\n\nCalder PC: Marine omega-3 fatty acids and inflammatory processes: Effects, mechanisms and clinical relevance. Biochim Biophys Acta. 2015; 1851(4): 469–84. PubMed Abstract | Publisher Full Text\n\nHanindita MH, Widjaja NA, Irawan R, et al.: Influence of intravenous fish oil-enriched lipid emulsion on the inflammatory response in children post gastrointestinal surgery. Pak J Nutr. 2019; 18(11): 1036–41. Publisher Full Text\n\nDahlan MS: Menggunakan rumus besar sampel secara benar. In: Dahlan MS (ed.), Besar sampel dan cara pengambilan sampel dalam penelitian kedokteran dan kesehatan, 2nd edn, Salemba Medica, Jakarta, 2005.\n\nHanindita MH, Irawan R, Ugrasena IDG, et al.: Data Set Comparison of Two Lipid Emulsions on Interleukin-1β, Interleukin-8 and Fatty Acid Composition in Infants Post Gastrointestinal Surgery: A Randomized Trial. figshare. Dataset. 2020. http://www.doi.org/10.6084/m9.figshare.12906320.v2\n\nWei Z, Wang W, Chen J, et al.: A prospective randomized controlled study of n-3 fish oil fat emulsion-based parenteral nutrition for patients following surgical resection of gastric tumors. Nutr J. 2014; 13: 25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang X, Li W, Li N, et al.: Omega-3 fatty acids-supplemented parenteral nutrition decreases hyperinflammatory response and attenuates systemic disease sequelae in severe acute pancreatitis: a randomized and controlled study. JPEN J Parenter Enteral Nutr. 2008; 32(3): 236–41. PubMed Abstract | Publisher Full Text\n\nZhao Y, Wang C: Effect of n-3 polyunsaturated fatty acid-supplemented parenteral nutrition on inflammatory and immune function in postoperative patients with gastrointestinal malignancy: a meta-analysis of randomized control trials in China. Medicine (Baltimore). 2018; 97(16): e0472. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMa C, Sun L, Chen F, et al.: A double-blind randomized study comparing the efficacy and safety of a composite vs a conventional intravenous fat emulsion in postsurgical gastrointestinal tumor patients. Nutr Clin Pract. 2012; 27(3): 410–5. PubMed Abstract | Publisher Full Text\n\nHan YY, Lai SL, Ko WJ, et al.: Effects of fish oil on inflammatory modulation in surgical intensive care unit patients. Nutr Clin Pract. 2012; 27(1): 91–8. PubMed Abstract | Publisher Full Text\n\nLin E, Lowry SF: The human response to endotoxin. Sepsis. 1998; 2: 252–62. Publisher Full Text\n\nSkouroliakou M, Konstantinou D, Agakidis C, et al.: Parenteral MCT/n-3 Polyunsaturated fatty acid-enriched intravenous fat emulsion is associated with cytokine and fatty acid profiles consistent with attenuated inflammatory response in preterm neonates: A randomized, double-blind clinical trial. Nutr Clin Pract. 2016; 31(2): 235–44. PubMed Abstract | Publisher Full Text\n\nGrimm H, Mertes N, Goeters C, et al.: Improved fatty acid and leukotriene pattern with a novel lipid emulsion in surgical patients. Eur J Nutr. 2006; 45(1): 55–60. PubMed Abstract | Publisher Full Text\n\nInnes JK, Calder PC: Omega-6 fatty acids and inflammation. Prostaglandins Leukot Essent Fatty Acids. 2018; 132: 41–8. PubMed Abstract | Publisher Full Text\n\nGoulet O, Antebi H, Wolf C, et al.: A new intravenous fat emulsion containing soybean oil, medium-chain triglycerides, olive oil and fish oil: A single-center, double-blind randomized study on efficacy and safety in pediatric patients receiving home parenteral nutrition. JPEN J Parenter Enteral Nutr. 2010; 34(5): 485–95. PubMed Abstract | Publisher Full Text\n\nWichmann MW, Thui P, Czametzki HD, et al.: Evaluation of clinical safety and beneficial effects of a fish oil containing lipid emulsion (Lipoplus, MLF541): data from a prospective, randomized, multicenter trial. Crit Care Med. 2007; 35(3): 700–6. PubMed Abstract | Publisher Full Text\n\nMorlion BJ, Torwesten E, Lessire H, et al.: The effect of parenteral fish oil on leukocyte membrane fatty acid composition and leukotriene-synthesizing capacity in patients with postoperative trauma. Metabolism. 1996; 45(10): 1208–13. PubMed Abstract | Publisher Full Text\n\nHanindita MH, Irawan R, Ugrasena IDG, et al.: Consort Checklist: Comparison of Two Lipid Emulsions on Interleukin-1β, Interleukin-8 and Fatty Acid Composition in Infants Post Gastrointestinal Surgery: A Randomized Trial. figshare. Preprint. 2020. http://www.doi.org/10.6084/m9.figshare.12906395.v1"
}
|
[
{
"id": "71910",
"date": "28 Sep 2020",
"name": "Philip Calder",
"expertise": [
"Reviewer Expertise Fatty acids",
"Human nutrition",
"Inflammation and immunity",
"Artificial nutrition support",
"Clinical trials"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript describes findings from a clinical trial in neonates who underwent surgery and then were randomised to receive one of two lipid emulsions intravenously over the course of 3 days. One lipid emulsion was a mix of coconut and soybean oils (control) and the other was a mix of coconut, soybean, olive and fish oils (treatment). The main outcomes were two cytokines (IL-1b and IL-8) and serum fatty acids. In addition a number of secondary outcomes are reported including white cell count and CRP. This study is of interest. A major limitation is sample size: there were 6 infants per group. The cytokines were not different between groups, but white cells and CRP were lower and omega-6 fatty acids higher at 3 days in the treatment group. The article is well written, data are clearly presented and the discussion refers to relevant existing literature.\nComments:\nI believe the abstract should report findings for serum omega-3 fatty acids.\n\nFor the sample size calculation what outcome(s) was used? This is not stated. Usually an effect size in a specific outcome is needed to calculate sample size.\n\nTable 1. Lists docosapentaenoic acid but this should read eicosapentaenoic acid.\n\nMore detail is required about the fatty acid analysis. In the methods section you refer to “free fatty acids in serum” but in the Discussion you refer to “profiling phospholipid fatty acids”. It is necessary to a) clarify what you have actually measured and b) provide details of how the serum was processed prior to gas chromatography.\n\nStat analysis section is written in future tense (will be) but should be in the past tese (were/was).\n\nStat analysis. The key comparison is between groups either at day 3 controlling for baseline or the change to day 3. It is not clear whether this comparison has been done.\n\nData display. What are the errors shown: SD or SEM?\n\nData that does nth ae normal distribution should not be shown as mean and SD/SEM but as median and IQR.\n\nTable 4. What are the units for total fatty acids?\n\nTable 4. Huge variation is apparent here bringing the small sample size into focus. For example in the treatment group ALA changes from 0.39 pre to 18.35 post but this almost 50-fold increase is not significant.\n\nTable 4. Why does EPA increase in the MCT/LCT group?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6118",
"date": "20 Nov 2020",
"name": "Meta Herdiana Hanindita",
"role": "Author Response",
"response": "Thank you for your willingness to review this paper, it's a great honour for us. 1. Serum omega-3 fatty acid finding is already included in the abstract. 2. For the sample size calculation, the outcome used was Interleukin-1beta. Detail on this sample size calculation is already added in the paper. The effect size was 20 pg/ml, with SD 10.6 pg/ml. 3.Table 1 is already revised. 4. Detail on the fatty acid analysis is already added. It should be serum free fatty acid. 5. Stat analysis section is already revised. 6. All parameters examined were compared for differences before surgery, 3 days after surgery (controlling to baseline) and differences within those 3 days (delta). This statement is already added to the paper. 7. The errors shown in SD. 8. We changed the data that does not have normal distribution, is already shown in median and IQR. 9. The unit is umol/L (added). 10. Yes, unfortunately due to limitation of this study (small sample size). The incidence of these cases in infants in our hospital that needed parenteral nutrition was only 18-20 patients/year. 11. We really tried to find reasons for this finding but we could not find one. We need to have another research on this, to understand its mechanism better."
}
]
},
{
"id": "71909",
"date": "06 Oct 2020",
"name": "Peter Stehle",
"expertise": [
"Reviewer Expertise Clinical nutrition",
"nutritional physiology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript summarizes results gained within a clinical nutrition trial in infants after gastrointestinal surgery. Infants parenterally received either a \"traditional\", omega 6-rich lipid emulsion composed of soybean oil and MCT, in the \"verum\" group a newly composed lipid emulsion produced with a blend of MCT, olive oil, fish oil and some soybean oil was administered. The study was generally well designed and performed; data presentation and interpretation is adequate. However, some information should be added and/or sharpened.\nPower calculation: Which parameter was used to make calculations? One of the primary outcomes? And which (already published) data were considered (sensitivity of the method, expected changes in concentrations)? Very crucial: Is that power calculation also valid for fatty acid analyses?\n\nGeneral nutritional concept: Which other components (glucose, amino acids...) were administered? And in which amounts? Was the concept isonitrogenous and isoenergetic?\n\nWhat was the reason to give lipids for (only) 3 consecutive days?\n\nAnalysis of fatty acids in blood samples: It is unclear what has been measured: only \"free\" fatty acids? Fatty acid composition of (total) lipids? When free fatty acids have been measured: how long after stopping infusion the blood sample was harvested? Please, comment and complete the text.\n\nTable 4: The capture is misleading. Please, reword and include that these data are \"blood\" analyses (see 4.).\n\nDiscussion: Any comparison with previous analytical data with respect to fatty acid profiles should mention what was measured (free fatty acids, lipid composition etc.) and under which clinical conditions.\n\nDiscussion: As mentioned by the authors themselves, metabolites, e.g. cytokines, are endogenously synthesized from fatty acid precursors released from membranes. With this background: how should a 3 day-infusion influence these metabolite concentrations?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": [
{
"c_id": "6119",
"date": "20 Nov 2020",
"name": "Meta Herdiana Hanindita",
"role": "Author Response",
"response": "Thank you for your willingness to review this paper. 1. Details on the power calculation are already added in the paper. The parameter that was used for sample size calculation was Interleukin-1b, with effect size of 20 pg/ml, and SD of 10.6 pg/ml. The formula used datas from our previous published research (reference no.17). No, the power calculation is only valid for interleukin-1b. 2. Yes, the concept was isonitrogenous and isocaloric. This statement is already added to the paper. 3. Our consideration for giving lipids for 3 consecutive days was because we wanted to know the inflammatory response in the initial/early inflammatory phase in the postoperative wound healing period. The initial effect of inflammation in the postoperative wound healing period occurs within 0-3 days. 4. It should be written as serum free fatty acid (revised already). Details on this were already included in the paper. The blood samples were harvested after stopping infusion for 3 hours. 5. It is already changed. 6. It is already added. 7. This study is the first study in Indonesia to examine the effects of intravenous fat emulsions on metabolites (such as cytokines) in postoperative infants. To be honest, we are making this study the baseline for our future studies. 3 days was chosen because this is the initial inflammatory period in the wound healing period. Looking at the results of this study, we probably will make further studies using lipid emulsion in infants after surgery over a longer time such as 7 days, 14 days and more than 14 days."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1168
|
https://f1000research.com/articles/9-1452/v1
|
14 Dec 20
|
{
"type": "Case Report",
"title": "Case Report: A rare presentation and diagnosis of gingival melanoacanthoma caused by teeth whitening strips",
"authors": [
"Hamad Albagieh",
"Ashwag Aloyouny",
"Shatha Alharthi",
"Hamad Albagieh",
"Shatha Alharthi"
],
"abstract": "Background: Oral melanoacanthoma is not common. It occurs mostly on the buccal mucosa. Since it happens suddenly and progresses rapidly, it clinically resembles melanoma. Melanoacanthoma occurs in regions susceptible to trauma. The clinical presentation of the lesion is not enough to diagnose it; therefore, tissue biopsy is necessary to exclude malignancy. Case report: We report a case of oral melanoacanthoma in a rare mucosal location in a 21-year old female patient. The generalized gingival melanoacanthoma was caused as a result of using teeth whitening strips. This irritating factor increased melanocyte activity in the gingival tissues and labial mucosa. Discussion: Oral melanoacanthoma is a rarely encountered pigmented lesion in the oral cavity and is especially uncommon in the gingiva. It is a reactive lesion affecting the mucous membranes with no risk of malignant transformation. This case report shows that teeth whitening strips may trigger oral melanoacanthoma in susceptible individuals. Long-term irritation of the oral tissues may increase the number of dendritic melanocytes throughout the epithelium and accordingly increase the brown pigmentation of the oral cavity. Eliminating all possible local sources of irritation and ruling out other causative factors are the standard first steps in the oral melanoacanthoma therapy. Conclusions: This case shows the importance of including oral melanoacanthoma in the differential diagnosis of diffuse gingival pigmented lesions.",
"keywords": [
"Oral Melanoacanthoma",
"gingival hyperpigmentation",
"oral pigmented lesion",
"teeth whitening strips."
],
"content": "Introduction\n\nOral Melanoacanthoma is a rare, pigmented lesion that usually occurs on the buccal mucosa. The most common site of oral melanoacanthoma is the buccal mucosa (51.4%), followed by the palate (22.2%), and lips (15.2%). The gingiva is least affected by melanoacanthoma (5.6%)1. A review of the literature identified only a few cases with generalized, diffuse, multiple, upper and lower gingival melanoacanthoma2. Clinical differentiation between benign and malignant oral pigmented lesions is very difficult at early stages. Generalized gingival melanoacanthoma is extremely rare, therefore, tissue biopsy is highly recommended to differentiate between melanoacanthoma and melanoma. We present a rare case of melanoacanthoma in the gingivae and the labial mucosae triggered by teeth whitening strips in a 21-year-old female patient. The work is reported in line with the CARE criteria3.\n\n\nCase report\n\nThe patient of this case report was a 21-year-old female, Saudi Arabian college student. She was referred by her dentist in July 2018 to an oral medicine specialist for evaluation of a three-month-history of remarkable intraoral, diffuse hyperpigmentation of the upper and lower gingivae and labial mucosae. She reported that the hyperpigmentation appeared suddenly and had rapidly increased in size. The patient reported using teeth whitening strips for six weeks before the intraoral pigmentation happened. Her past medical history revealed hypothyroidism for three years for which she had been taking 100 mcg/day of levothyroxine sodium. She was not on any other mediation. She was not aware of any relevant family history of extensive brown oral pigmentation. She reported no history of mental health disorders and denied the use of tobacco products. The patient was referred to an endocrinologist to rule out systemic diseases such as Addison’s disease. Accordingly, a blood test was carried out and the results were all within normal limits; red blood cell count of 4.99 × 106/μL, platelet count of 259× 103 /μL, hemoglobin count of 13.3 g/dL, white blood cell count of 8.88 × 103/μL, lymphocytes 33.6%, segmented neutrophil 54.6%, ferritin 31.11 ng/mL, iron 14.6 umol/L, FT4 17.67 ng/dL, TSH 0.079 mIU/L, Vitamin D 65 ng/mL and cortisol 184.9 nmol/L. A physical examination revealed no skin pigmentation and an extraoral examination showed no significant findings. An intraoral examination showed asymptomatic, diffuse, smooth, macular blackish-brown pigmentation with irregular margins on both the upper and lower attached and free gingivae, and upper and lower labial mucosae (Figures 1a–1c).\n\nMacular brown pigmentation with irregular margins involving (a) upper and lower attached and free gingivae; (b) upper labial mucosa; (c) lower labial mucosa.\n\nThe clinical presentation and the widespread nature of the lesion were worrisome to the clinician; therefore an incisional biopsy was performed. The goals and expectations of the procedure were discussed with the patient as well as the potential risks for surgical and post-surgical complications such as bleeding, swelling, discomfort, infection and scarring. Written informed consent was obtained from the patient. After applying 1.8 mL of the anesthetic solution (lidocaine HCL 2% and epinephrine 1:100,000) via intraoral injection to the gingival area of tooth No. 33, a 3 mm section of gingival tissue was removed at the apical area of tooth No. 33, at the darkest pigmented spot of the lesion and a bit away from the esthetic zone. The incised tissue was blackish-brown in colour and measured 0.3 × 0.2 × 0.1 cm. The gross specimen was fixed in 10% neutral buffered formalin and then submitted as one piece in one cassette for histopathology examination. The microscopic analysis of the hematoxylin and eosin (H&E) stained sections showed a hyperorthokeratinized, hyperplastic, stratified squamous epithelium revealing acanthosis, and long rete ridges. Many benign dendritic melanocytes with pigment-laden dendritic processes were distributed in the epithelium. In addition, melanin pigmentation of basal cell layer was noted with evidence of melanin deposits in the lamina propria (Figure 2). A microphotograph of Melan-A stained tissue shows melanotic hyperplasia in the epithelium (Figure 3).\n\nMicrophotograph of hematoxylin and eosins stained tissue shows the parakeratinized, hyperplastic, stratified squamous epithelium with acanthosis, and long rete ridges. Many benign dendritic melanocytes with dendritic processes can be seen distributed in the epithelium, as well as melanin deposits in the lamina propria.\n\nMicrophotograph of Melan-A stained biopsy tissue shows melanotic hyperplasia in the epithelium.\n\nIn light of the patient history, clinical presentation, and the histopathology report of the incisional biopsy, the final diagnosis of oral melanoacanthoma was confirmed. The patient was reassured of the benign nature of the lesion and she was advised to stop using the teeth whitening strips. At 8-months follow up, the lesion had gradually faded (Figure 4).\n\nAn eight-month-follow up visit revealed that the brown pigmentation had faded gradually.\n\n\nDiscussion\n\nCutaneous melanoacanthoma was first reported in 1927, yet oral melanoacanthoma was not described until 19784. Melanoacanthoma was named by Mishima and Pinkus in 19605,6. It shares some of the alarming characteristic features of malignant melanoma such as sudden appearance and rapid growth rate. It is described as asymptomatic, solitary or multifocal, and diffuse, with ill-defined areas of dark brown to black pigmentation, a flat macule or slightly raised, and is usually greater than 1 cm in diameter. It usually occurs in teenage to middle-aged women with dark skin pigmentation7.\n\nThe exact pathophysiology of oral melanoacanthoma is undetermined; however, the clinical manifestations of the lesion is suggestive of a reactive origin, since it occurs mainly in regions liable to trauma8. The oral melanoacanthoma, in this reported case, occurred on the soft tissue areas that came in close contact with the teeth whitening strips. After excluding all other potential causes, the author believed that the teeth whitening strips could have participated in the oral pigmented lesion in this patient, resulting in diffused hyperpigmentation. These teeth whitening strips could have been a source of irritation to the tissues as they contain a gel that has active chemical ingredients such as hydrogen peroxide. Prolonged exposure to high concentrations of hydrogen peroxide can damage oral soft tissues9.\n\nThe differential diagnoses of oral melanoacanthoma includes: melanoma, Addison’s disease, McCune-Albright syndrome, Peutz-Jeghers syndrome, physiologic racial pigmentation, post-inflammatory lichen planus, acquired melanocytic nevus, and oral melanotic macules10. Gingival pigmentation may also be caused by amalgam material, chewing tobacco (khat), gingival tattoo, graphite implantation, and several hygiene products. Moreover, utilizing different heavy metals and some medically prescribed drugs could participate in the presence of pigmented lesions in the mouth2. Fortunately, oral melanoacanthoma has an excellent prognosis; in most of the reported cases, the pigmented lesions start to fade gradually either after removing the causative factor such as dental hygiene products or tobacco products, or following minor trauma such as tissue biopsy or sharp food injury. In the present case, pigmentation gradually disappeared after performing the incisional gingival tissue biopsy of the lesion. Eliminating all possible local sources of irritation and ruling out other causative factors are the standard first step in the oral melanoacanthoma therapy.\n\nThis case study has many strengths; the clinical examination was done thoroughly, the laboratory result was reviewed carefully, and complete data was collected. After that, a tissue biopsy from the lesion was performed to determine the condition and based on the histopathology report, the causative factor was eliminated. As a result, the lesion regressed gradually.\n\n\nConclusion\n\nOral melanoacanthomas are rarely encountered pigmented lesions in the oral cavity. They are especially rare in the gingival region. These pigmented lesions should be biopsied and carefully analyzed under the microscope to rule out a diagnosis of a malignant lesion such as melanoma. This case report highlights the important role of dentists to include oral melanoacanthoma in the differential diagnosis of diffuse gingival pigmented lesions.\n\n\nPatient perspective\n\nThe patients reported that the disappearance of the brown discoloration had a good impact on the esthetic appearance of her smile which allowed her to regain confidence.\n\n\nConsent\n\nWritten informed consent was obtained from the patient for publication of this case report and accompanying images.\n\n\nData availability statement\n\nAll data underlying the results are available as part of the article and no additional source data are required.",
"appendix": "Acknowledgements\n\nThis research was funded by the Deanship of Scientific Research at Princess Nourah bint Abdulrahman University through the Fast-track Research Funding Program.\n\n\nReferences\n\nYarom N, Hirshberg A, Buchner A: Solitary and multifocal oral melanoacanthoma. Int J Dermatol. 2007; 46(12): 1232–1236. PubMed Abstract | Publisher Full Text\n\nDatta A, Lamba AK, Tandon S, et al.: A Unique Presentation of Gingival Melanoacanthoma: Case Report and Review of Literature. Cureus. 2020; 12(3): e7315. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAgha RA, Borrelli MR, Farwana R, et al.: The SCARE 2018 statement: Updating consensus Surgical CAse REport (SCARE) guidelines. Int J Surg. 2018; 60: 132–136. PubMed Abstract | Publisher Full Text\n\nAndrews BT, Trask DK: Oral melanoacanthoma: a case report, a review of the literature, and a new treatment option. Ann Otol Rhinol Laryngol. 2005; 114(9): 677–680. PubMed Abstract | Publisher Full Text\n\nTomich C: Abstract of the 32nd Annual Meeting. Am Acad Oral Pathol Ft Lauderdale, Fla.\n\nMishima Y, Pinkus H: Benign mixed tumor of melanocytes and malpighian cells. Melanoacanthoma: Its relationship to Bloch's benign non-nevoid melanoepithelioma. Arch Dermatol. 1960; 81(4): 539–50. PubMed Abstract | Publisher Full Text\n\nBrooks JK, Sindler AJ, Scheper MA: Oral melanoacanthoma in an adolescent. Pediatr Dermatol. 2010; 27(4): 384–387. PubMed Abstract | Publisher Full Text\n\nTomich CE, Zunt SL: Melanoacanthosis (melanoacanthoma) of the oral mucosa. J Dermatol Surg Oncol. 1990; 16(3): 231–236. PubMed Abstract | Publisher Full Text\n\nWalsh LJ: Safety issues relating to the use of hydrogen peroxide in dentistry. Aust Dent J. 2000; 45(4): 257–269; quiz 289. PubMed Abstract | Publisher Full Text\n\nNeville BW, Damm DD, Allen CW, et al.: Epithelial Pathology. Oral and Maxillofacial Pathology. 3rd ed. W.B. Saunders, Philadelphia; 2009."
}
|
[
{
"id": "76096",
"date": "15 Dec 2020",
"name": "Mahnaz Fatahzadeh",
"expertise": [
"Reviewer Expertise Oral medicine"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe case is well-documented and the article is well-written. Perhaps, a reviewer in that field may offer additional insights regarding the photomicrographs.\nI have made a few minor suggestions to improve the article as follows:\nTitle: It may be more appropriate to change it to \"A rare presentation of gingival melanoacanthoma caused by teeth whitening strips: A Case Report\".\n\nAbstract, case report paragraph: Please combine 1st and 2nd sentence as follows\" ..... female patient in whom generalized gingival melanaoacanthoma was related to the use of teeth whitening strips.\"\n\nAbstract, discussion paragraph, last sentence: change to \"..... first step in treatment of oral melanoacanthoma.\"\n\nPage 3, case report, 1st paragraph: It is appropriate to include if patient reported bronzing of her skin or not.\n\nPage 3, case report, 1st paragraph: If available, please include the result of typical tests for Addison's disease (serum Na, K, cortisol, ACTH and their normal ranges).\n\nPage 3, case report, second paragraph: I agree that biopsy was warranted to determine the nature of diffuse oral pigmentation in this case. However; melanoma would be a less likely concern in this case. Oral melanoma is rare and presentation of gingival melanoma in both jaws is even more rare.\n\nPage 4, discussion, second paragraph: correct the spelling for the word \"diffuse\".\n\nPage 4, discussion, third paragraph: Did the mandibular gingival pigmentation also resolve over time? If so, it is more likely to be due to discontinuation of whitening strips rather than biopsy performed in maxilla.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes",
"responses": []
},
{
"id": "76102",
"date": "21 Dec 2020",
"name": "Lujain Homeida",
"expertise": [
"Reviewer Expertise Oral medicine/ TMDs/Medcially compromised patients."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI reviewed the manuscript entitled: \" Case Report: A rare presentation and diagnosis of gingival melanoacanthoma caused by teeth whitening strips\" and found it of interest to the readers of F1000Research.\nIn this paper, they reported a case of a rare pigmented mucosal melanoacanthoma with uncommon trigger. The importance of this case arises from its resemblance clinically to malignant melanoma such as sudden appearance and rapid growth rate. However, melanoacnthoma has no risk of malignant transformation. They concluded by emphasizing on considering oral melanoacanthoma in the differential diagnosis of diffuse gingival pigmented lesion.\nThis case report highlights the importance of investigating cases with oral pigmented lesions through detailed history taking and clinical work up to rule out any potential malignancies. The presented clinical evaluation and management of this case is excellent. I found this paper very relevant to the scope of this journal. The only point I would arise here is that they referred to some relatively old references and I would kindly advise them to use more updated references.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1452
|
https://f1000research.com/articles/9-1486/v1
|
18 Dec 20
|
{
"type": "Method Article",
"title": "A Method to adjust for measurement error in multiple exposure variables measured with correlated errors in the absence of an internal validation study",
"authors": [
"Alexander K. Muoka",
"George O. Agogo",
"Oscar O. Ngesa",
"Henry G. Mwambi",
"George O. Agogo",
"Oscar O. Ngesa",
"Henry G. Mwambi"
],
"abstract": "Difficulty in obtaining the correct measurement for an individual’s longterm exposure is a major challenge in epidemiological studies that investigate the association between exposures and health outcomes. Measurement error in an exposure biases the association between the exposure and a disease outcome. Usually, an internal validation study is required to adjust for exposure measurement error; it is challenging if such a study is not available. We propose a general method for adjusting for measurement error where multiple exposures are measured with correlated errors (a multivariate method) and illustrate the method using real data. We compare the results from the multivariate method with those obtained using a method that ignores measurement error (the naive method) and a method that ignores correlations between the errors and true exposures (the univariate method). It is found that ignoring measurement error leads to bias and underestimates the standard error. A sensitivity analysis shows that the magnitude of adjustment in the multivariate method is sensitive to the magnitude of measurement error, sign, and the correlation between the errors. We conclude that the multivariate method can be used to adjust for bias in the outcome-exposure association in a case where multiple exposures are measured with correlated errors in the absence of an internal validation study. The method is also useful in conducting a sensitivity analysis on the magnitude of measurement error and the sign of the error correlation.",
"keywords": [
"Measurement error",
"Internal validation study",
"Attenuation",
"Bias",
"Questionnaire data",
"Sensitivity analysis",
"Error correlation"
],
"content": "Abbreviations\n\nHIV: Human immunodeficiency virus; HBCT: Home-based HIV counseling and testing; HSRC: Human sciences research council; NCD: Non-communicable diseases; BMI: Body mass index; kg: kilogram; m2: metre squared; g: gram; MCMC: Markov Chain Monte Carlo; CI: Credible interval; JAGS: Just another gibbs sampler; BUGS: Bayesian inference using gibbs sampling; ACF: Autocorrelation function\n\n\nIntroduction\n\nDifficulty in obtaining correct measurements of an individual’s long-term exposure is a major challenge in an epidemiological study that investigates the association between a continuous exposure and a health outcome. For instance, several studies estimated the correlations between self-reported intake from a questionnaire and the true long-term intake values to be less than 0.82 for fruits and about 0.72 for vegetables1–5, an implication that some of the variation in the diet intake measurements is due to random errors. Due to random error, the association between the dietary intakes and health outcomes may be biased. The effect of measurement error can be quantified using either: (i) the attenuation factor, which quantifies the bias in the association or (ii) the correlation coefficient between the true and the observed exposure, which quantifies the loss of statistical power to detect a significant association (i.e. validity coefficient)6.\n\nValidation studies are used to assess the accuracy of the dietary questionnaire6–12. A validation study constitutes a small number of individuals from whom dietary intakes are measured repeatedly using an unbiased instrument13. There are two types of validation studies: the external and internal validation studies. An internal validation study is conducted on a subset of individuals from the main study, whereas an external validation study is carried on a group of subjects who are not part of the main study, but who are similar in characteristics to individuals in the main study. Validation studies are often expensive to conduct and, in some cases not feasible. Several methods have been proposed to handle measurement error in the absence of internal validation data14–18.\n\nAgogo et al.14 conducted a sensitivity analysis to investigate the effect of the magnitude of the correlation between errors in the covariates of interest and found that the magnitude of measurement error adjustment is sensitive to the assumed measurement error structure. Dellaportas and Stephens15 presented a Bayesian method for analysis of non-linear error-in-variable where prior knowledge of the unknown true covariate is incorporated. Huang et al.16 proposed a quantile regression-based non-linear mixed-effects joint models for longitudinal data that simultaneously accounts for a response with non-central location and for covariate with non-normality and measurement error under the Bayesian framework. Lin17 proposed a Bayesian semi-parametric accelerated failure time model to analyze censored survival data with covariate measurement error and evaluated their method using an intensive simulation study. Muff et al.18 introduced a Bayesian method to handle a mixture of classical and Berkson measurement errors in a single explanatory variable and illustrated their method to studying cardiovascular disease mortality.\n\nThe majority of these authors considered a case where one exposure is measured with error (hereafter, a univariate case). In a univariate method, the bias in the association between an outcome and the exposure is adjusted by dividing the unadjusted association estimate by the attenuation factor19. An attenuation factor is the ratio of the variance of the true exposure to the variance of the observed exposure. This method ignores correlations between the errors, which can lead to substantial bias. In this study, we suggest a general method for adjusting for measurement error where multiple exposures are measured with correlated errors in the absence of an internal validation study (hereafter, a multivariate method). We use real data to illustrate the method in handling a case where three exposures are measured with correlated errors (hereafter, the trivariate method) under a linear regression model and demonstrate the implementation of this method using R software20. Specifically, we use a subset of data from a home-based HIV counseling and testing study that was done in rural and peri-urban communities in KwaZulu-Natal Province, South Africa21. We compare the results obtained when using a method that ignores both the measurement error and correlation between the errors (hereafter, a naive method) with those obtained when using univariate and multiple exposures methods. Moreover, we conduct a sensitivity analysis to investigate how the coefficient estimates of parameters of interest are influenced by (1) a change in the level of uncertainty assumed for the limits of the validity coefficients and (2) varying the correlation between errors in the measured exposures.\n\nThe remaining sections of this paper are organized as follows. In section 2, we discuss materials and methods used in this study. We present the results of the study in section 3. Finally, we provide a discussion and conclusion in section 4.\n\n\nMethods\n\nIn this work, we use a subset data from a home-based HIV counseling and testing (HBCT) study that was conducted in rural and peri-urban communities in KwaZulu-Natal Province, South Africa, between November 2011 and June 201221. The data were obtained from the Human Sciences Research Council (HSRC) of South Africa21. This study was conducted to provide a better understanding of the complexity, severity, and prevalence of non-communicable disease (NCDs) in a community known to have one of the highest rates of HIV incidence and prevalence in the world21.\n\nHome-based HIV counseling and testing is a cross-sectional, single-site study in South Africa that aims to increase engagement in HIV care by integrating NCDs screening with community-based HIV testing22. A random sampling approach was used, where 587 participants over the age of 18 were selected from 50,000 people living in the Mpumuza suburb21. Anthropometric and biological measures were collected in the survey with the purpose of establishing the prevalence of a range of NCDs and associated risk factors. Eligible individuals participated in a face-to-face interview, physical, psychological and clinical examinations. Persons younger than 18 years living in Mpumuza and all household members not previously enrolled, and members unable to give written consent were excluded from the study. Mobile phones were used for data collection to increase efficiency in data capture and analysis21.\n\nIn our study, we used a subset data consisting of 76 individuals who self-reported the number of cigarettes smoked, fruit and vegetable consumption. We use the dataset to illustrate the multivariate method in modeling the amount of association between body mass index (BMI) and three exposures (smoking, fruit, and vegetable intakes). BMI was measured in kg/m2, while smoking was measured as the average number of cigarettes smoked per day. Initially, fruit and vegetable intakes were measured in terms of the number of servings consumed per day. It is often assumed that a standard portion of fruit/vegetable weighs about 80g5. Therefore, for this study, we converted the number of servings to grams per day (g/day) by multiplying the reported number of servings by 80g. The subset data has the following three properties that make it suitable for use in this work: (1) measurement error in the recorded number of cigarettes smoked due to possible misreporting, (2) measurement error in fruits and vegetable consumption due to recall bias, and conversion of the number of servings of fruits and vegetables into grams, and (3) the measurement error in the three exposures is often correlated, for instance, smokers are likely to over-report fruit and vegetable intakes due to their beneficial effects, and to under-report the number of cigarettes they smoke due to the associated harmful effects. Epidemiologically, BMI is used as a risk factor of a health outcome. However, in this study, we model BMI as an outcome as in other several studies, for instance,23–26. The subset data is only used to illustrate the method and not to draw inference.\n\nEthics approval was granted by both HSRC Research Ethics Committee (REC: 1/26/05/11) and the University of Washington Institutional Review Board (48733). Informed written consent was obtained from each participant in the study. Participants were provided with written information on the study (including the study’s background and objectives) and their rights regarding participation and withdrawal at any time.\n\nAn interest in epidemiological study could be to investigate the association between BMI and three exposures namely: fruit, vegetable and smoking using the multiple linear regression\n\n\n\nwhere Y denotes the BMI, β0 is the intercept, βX1, βX2 and βX3 are the coefficient parameters for the true long-term fruit (X1), vegetable (X2) and cigarette (X3) intake respectively and ϵ is the random error term. In this study, we use vegetable intake and cigarette smoking as confounders and assume that the main interest is in estimating βX1. In practice, the true intakes are unobservable and, therefore, the intakes recorded in self-reported questionnaires are used. Let W1, W2 and W3 denote the measured versions of X1, X2 and X3, respectively. The use of Wp’s in place of Xp’s, (p = 1, 2, 3), in Equation (1) yields biased estimates β^W1, β^W2 and β^W3 of βX1, βX2 and βX3 respectively. Let β^W=(β^W1,β^W2,β^W3)T.\n\nWe assumed that the observed exposures are related to the true exposures with additive measurement error as\n\n\n\nwhere ϵW = (ϵW1, ϵW2, ϵW3)⊤, ϵW ∼ N(0, ΣϵW); W = (W1, W2, W3)⊤; α0 = (α01, α02, α03)⊤, α1 = (α11, α12, α13)⊤; with the terms in α0 and α1 quantifying the constant bias and the proportional scaling bias respectively; ϵW is a random error term, ϵWi is assumed to be independent of the true exposure Xi and the systematic bias components, α0i and α1i.\n\nA univariate method. In a univariate case, the bias in the association between an outcome and an exposure is adjusted by dividing the unadjusted association estimate by the attenuation factor19. Attenuation factor (λi) is defined as λi = var(Xi)/var(Wi), i.e., the ratio of the variance of the true exposure to the variance of the observed exposure, also referred to as reliability ratio. This method ignores correlations between the errors and also the correlation between the true exposures.\n\nMultivariate method. We propose and describe a general approach for handling p-exposures (p≥2) measured with correlated errors. For simplicity and without loss of generality, we assume that Wi is measured without systematic bias (i.e., α0i = 0, α1i = 1 in Equation 2). For multiple exposures measured with correlated errors, the adjusted association estimates can be obtained by pre-multiplying the unadjusted association estimates by the inverse of the transpose of attenuation-contamination matrix as\n\n\n\nwhere β^X∗ and β^W∗ denotes vectors of true and biased coefficients for the p-exposures respectively and Λp denotes a p × p attenuation-contamination matrix19,27. The off-diagonal elements in Λ are known as contamination factors while the diagonal elements are called attenuation factors14. Noteworthy, the attenuation factor quantifies the bias in the association between an outcome and an exposure. In contrast, the contamination factor quantifies the effect of measurement error in one exposure variable on the other exposure variable’s estimate. β^W∗ in Equation (3) can be obtained from the observed questionnaire data.\n\nIn the multiple exposures case, the estimate of attenuation-contamination matrix Λ^p is defined as\n\n\n\nwhere Σ^X* is the estimate of covariance matrix of the true exposures, Σ^W*−1 is the inverse of the estimate of covariance matrix of the measured exposures, σ^Xi2 is the variance estimate of Xi (i = 1, 2, ..., p) ; σ^XiXj (j = 1, 2, ..., p; i ≠ j) denotes the covariance estimate between the true exposures; σWi2 is the variance estimate of Wi; σ^WiWj (i ≠ j) is the covariance estimate between the observed exposures.\n\nThe elements of the variance-covariance matrix of the observed exposures, ΣW∗, are estimated from the observed data. The variances of the true exposures, σXi2 ’s, can be estimated using validity coefficients for the questionnaire. According to Kipnis et al.6, the validity coefficient is given by:\n\n\n\nwhere Wi is assumed to be the measured with error term only and ϵWi is assumed to be independent of Xi. From Equation (5), we estimate the variance of the true exposures as\n\n\n\nby incorporating external validation information on ρWiXi. To obtain covariances between the true exposures, one of the following two approaches is used: (i) if external information about the correlation between true exposures (i.e. ρ^XiXj ) is available, we obtain covariances between true exposures as follows:\n\n\n\nwhere σ^Xi are obtained as shown in Equation (6); (ii) if we can obtain prior information about the correlation between the errors in the observed exposures, ρ^ϵWiϵWj, we can solve for σ^XiXj by decomposing the covariance of observed exposures into unknown covariance between true exposures and unknown covariance between errors as follows:\n\n\n\nwhere Xi and ϵWj, Xj and ϵWi are assumed to be uncorrelated.\n\nFrom Equation (2) and Equation (6), the estimate of the error variance σ^ϵWi2 is\n\n\n\nSee Appendix B of the extended data28 for the proof.\n\nFrom Equation (8)–Equation (9), the covariances between the true exposures are given by\n\n\n\nUsing the observed data and external information, we can determine all the terms required to estimate the attenuation-contamination matrix, Λ, as shown in Equation (4) and adjust for the bias in the association between the exposures measured with error and the outcome using Equation (3).\n\nWe illustrate a method that accounts for uncertainty in the validity measures attributable to heterogeneity in the study populations and in parameter estimation. The proposed Bayesian method applies Markov Chain Monte Carlo (MCMC) estimation approach to combine observed self-reported data and external validation data in adjusting for measurement error in three exposures measured with correlated errors. MCMC is a class of algorithms that samples from the posterior distributions by traversing the parameter space29. The posterior distribution is obtained by updating the prior distribution with observed data. The steps for implementing the trivariate method are described below.\n\nWe first obtained external information on validity coefficients and generated validity coefficients for use by interpreting the lower and upper limits obtained from the literature as the 95% credible intervals (CIs) of the distribution of possible values respectively. Due to the skewed distribution of validity coefficients, Fisher’s transformation was used to generate the validity coefficients as explained in the next section.\n\nSecond, for the observed exposures, we estimated the posterior distribution of the covariance matrix (ΣW). The exposures were assumed to follow a multivariate normal distribution with mean and covariance, i.e., W ∼ N3(µW, ΣW). We assumed a weakly informative multivariate normal prior for µW as µW prior ∼ N3(0,106 I3), where I3 is a 3 × 3 identity matrix. In a multivariate normal distribution, ΣW must satisfy two conditions: (1) be positive definite (i.e. WTΣWW > 0, for all W) and (2) be a symmetric matrix. The semi-conjugate prior distribution for ΣW, which has these two properties, is the inverse-Wishart distribution29. To minimize the influence of the prior information on the estimate of ΣW, we considered weakly informative inverse-wishart prior as ΣW prior ∼ IW(I3, v), where v = 3 is the degrees of freedom.\n\nThird, using the validity coefficients generated from the external data and the posterior distribution of covariance matrix for observed exposures, we estimated the variance of true intakes, σ^Xi2 (i = 1, 2, 3), using the relationship given in Equation (6) so that\n\n\n\nThe covariances between true intakes (σ^XiXj ; j = 1, 2, 3) were estimated as,\n\n\n\nby incorporating external validation information on correlation between the errors (ρϵWi ϵWj). We generated the correlation between errors from a plausible range guided by correlation in the observed data and prior expert information on the most likely sign of the correlation between the exposures, as described in the next section.\n\nHaving obtained the covariance matrices of the true and observed exposures, we estimated the attenuation-contamination matrix (Λ3) from their joint distribution as\n\n\n\nwhere Σ^X is the estimate of covariance matrix of the three true exposures, Σ^W−1 is the inverse of the estimate of covariance matrix of the three measured-with-error exposures, σ^Xi2 is the variance estimate of Xi (i = 1, 2, 3); σ^XiXj (i ≠ j) denotes the covariance estimate between the true exposures; σ^Wi2 is the variance estimate of Wi; σ^WiWj (i ≠ j) is the covariance estimate between the observed exposures.\n\nLastly, we fitted a Bayesian multiple linear regression model (hereafter, naive method) to obtain the posterior distributions of the unadjusted coefficient estimates β^W=(β^W1,β^W2,β^W3)T. In the naive model, we assumed weakly informative normal independent priors by choosing a very small precision (large variance) for the unadjusted coefficient estimates as βWi prior ∼ N(0, 106). The adjusted coefficient estimates β^X were then obtained from the joint posterior distribution of Λ^3 and β^W as\n\n\n\nWe implemented the trivariate method in R version 3.6.3 using rjags (version 4-10), coda (version 0.19-3), MCMCpack (version 1.4-9), and mvtnorm (version1.1-1) packages. To facilitate Bayesian estimation of the covariance matrix of the observed exposures (ΣW), rjags package was used to provide an interface from R to the JAGS library30. JAGS is a gibbs sampler that uses MCMC to draw dependent samples from the posterior distribution of the parameters31. The Bayesian estimation of ΣW proceeded in the following steps: (1) defining a model for ΣW under Bayesian inference using gibbs sampling (BUGS) algorithm in a stand alone file, (2) reading the model file using the jags.model function, (3) updating the model using the update method for jags objects and (4) extracting the posterior samples of the model using the coda.samples function from the coda package.\n\nMCMCregress function from the MCMCpack package was used to generate a posterior density sample from the naive linear regression model32. MCMC convergence diagnostics of all the model parameters was done using trace plots and autocorrelation (ACF) plots from the coda package33. See extended data: Appendix C28 for convergence diagnostics results. For each model, the burn-in iterations were set to 2,000 and 10,000 MCMC iterations were run after the burn-in iterations. Every first sample value was kept in the MCMC simulations by using a thinning interval of 1. When compiling a JAGS model, an initial sampling step may be needed during which the samplers learn their behaviour to maximize their performance34. Therefore, the number of iterations for adaptation in the the jags model was set to 500. The results were presented in terms of density plots, posterior mean and median. We compared the results obtained under naive, univariate, and trivariate methods. The R code used for analysis is presented in the extended data28.\n\nExternal information on the validity coefficient and error correlations for fruit, vegetable, and cigarette information was obtained from the literature. According to Kaaks et al.1, the validity coefficient of self-reported fruit intake ranges from 0.33 to 0.79, while that of vegetable intake ranged from 0.30 to 0.60. A meta-analysis study on the validity of questionnaires assessing fruit and vegetable consumption by Collese et al.2 reported validity coefficients of 0.26 for vegetables and 0.49 for fruits. Other similar validation studies reported validity coefficients in the aforementioned ranges for fruits and vegetables3,4,35. Therefore, based on these information we considered a range of 0.3 to 0.8 for fruits and a range of 0.25 and 0.7 for vegetables.\n\nIn the Scottish Heart Health Study of 2,849 men and 2,900 women36, the correlation between the self-reported number of cigarettes and biochemical measures was reported between 0.67 and 0.72. In a study on the validation of self-reported smoking by analysis of hair for nicotine and cotinine37, the validity coefficient between the number of cigarettes smoked per day and nicotine/cotinine levels in hair and plasma was found to be between 0.48 and 0.63, while the correlation between the average number of cigarettes smoked and carboxyhemoglobin was 0.70. In a follow-up study to examine the relationships among self-reported cigarette consumption, exhaled carbon monoxide, and urinary cotinine/creatinine ratio in pregnant women38, a validity coefficient in the range of 0.61 to 0.70 was reported. A study by Stram et al.39 found the correlation between the self-reported number of cigarettes smoked and the true lung dose to be between 0.40 and 0.70, and this range was consistent with the findings from the previously discussed related validation studies. Based on this information, we considered a validity coefficient range of 0.40 and 0.70.\n\nWe generated the correlation between errors from plausible ranges that were determined based on the correlation in the observed data and the most probable sign of the correlation among fruits, vegetables, and cigarettes as explained below:\n\na. Since the correlation coefficient between fruit and vegetable intake in the observed data was positive, we also assumed the error correlation between fruit and vegetables to be mostly positive;\n\nb. An investigation on the correlation coefficient between cigarette smoking and fruits/vegetable intake in the observed data showed a negative correlation coefficient. Based on this and the fact that persons who tend to overstate fruit and vegetable consumption are likely to understate the number of cigarettes smoked, we assumed the error correlation to be mostly negative.\n\nWe obtained the upper limits of error correlations by assuming that the error covariance equals the covariance in the observed data and set the lower limit of the error correlation to zero, based on the assumption that the covariance in the observed data equals the covariance between the true intakes14.\n\nUsing the range of plausible values obtained from external validation information, we generated the validity coefficients using the Fisher-Z transformation method by assuming that the reported lower and upper limits are 0.05 and 0.95 quantiles of the uncertainty distribution, respectively. Fisher Z-transformation is a commonly used method to transform the sampling distribution of correlation coefficients to become approximately normally distributed40,41. The procedure is as outlined below:\n\n(i) Using the Fisher Z-transformation formula\n\n\n\ntransform the lower (rl) and upper (ru) limits of the validity coefficient ρWiXi to get the corresponding Fisher-Z transformed values FZl and FZu respectively.\n\n(ii) Compute the mean µZi and the standard deviation σZi of FZi as µZi = 0.5(FZu − FZl) and σZi=0.5(FZu−FZ1)Zα/2 where Zα/2 is the (1−α2)% quantile of a standard normal random variable.\n\n(iii) Generate FZi ’s as FZi~N(μZi,σZi2)\n\n(iv) Using the inverse of Fisher Z-transformation, back-transform the generated FZi ’s to validity coefficient as\n\n\n\nWe investigated how varying the level of uncertainty assumed for the limits of the validity coefficients reported from literature affected the estimates for fruit, vegetable, and the average number of cigarettes smoked. We also investigated how the estimates varied with the magnitude of the correlation between errors in fruit and vegetable intake, fruit and cigarette smoking, and vegetable and cigarette smoking. This helps determine the estimates’ sensitivity to various magnitudes of CI and the correlation between errors when using the multivariate method.\n\n\nResults\n\nTable 1 presents regression coefficients estimates for fruit intake (g/day), vegetable intake (g/day), and the average amount of smoked cigarettes a day obtained using the naive method and the two bias adjustment methods (i.e., univariate and trivariate methods). The regression coefficient estimate adjusted for bias using either the univariate or trivariate method was greater in absolute value than that obtained using the naive method. Specifically, for fruit intake and the average number of cigarettes smoked, the bias-adjusted coefficient estimates were three times as large as the naive coefficient estimates. For vegetable intake, the increase in the strength of the association was about four times as compared to the naive regression coefficient estimates.\n\nFor both fruit intake and the average number of cigarettes smoked, the univariate method gave slightly greater estimates while the bias-adjusted values for vegetable intake were slightly lower in the univariate method. The variability of the regression coefficient estimate of the number of cigarettes smoked was higher than that for both fruits and vegetable intake. Again, the variability in either the univariate or trivariate method was higher than in the naive method due to uncertainty involved in adjusting for measurement error.\n\nFigure 1–Figure 3 show the kernel densities representing the distributions of adjusted for measurement error (solid curves) and naive (dotted curves) estimates for fruits intake, vegetable intake, and the number of cigarettes smoked, respectively. The solid vertical lines on the density plots depict the posterior mean of the adjusted regression coefficients, while the vertical dotted lines show the posterior mean of the naive regression coefficient estimates. A careful investigation of the posterior means as represented by the vertical lines on the kernel densities reveals that the adjusted for bias regression coefficient estimates are generally higher (in absolute value) than their corresponding naive estimates.\n\nThe solid vertical lines show the posterior means of coefficient estimates adjusted for bias; the dotted vertical lines indicate the posterior means of unadjusted coefficient estimates.\n\nWith the naive method, the variance of the regression coefficient for vegetable intake is more underestimated than for fruit intake, as depicted by the smaller length between the tails of the density plots. Of the three exposures considered in this study, the regression coefficient variance for the average number of cigarettes smoked is the most underestimated (see Table 1 and Figure 1–Figure 3). In general, a comparison of the regression coefficients’ variance in the naive and the trivariate method shows that the naive method underestimates the variance of regression coefficients.\n\nPresented in Table 2 are the mean (standard deviation) and the median for the estimates of fruit, vegetable, and the average number of cigarettes smoked adjusted for measurement error using the trivariate method in exploring the effects of the magnitude of uncertainty in the reported validity coefficients. From the results, the CI assumed in the distribution of the validity coefficient does not affect the mean and the median estimates of fruit, vegetable, and smoking. With the trivariate method, the results further show that the estimates’ uncertainty is slightly affected by the level of uncertainty assumed for the validity coefficients. Figure 4 to Figure 6 presents the mean coefficient estimates of fruit, vegetable and the average number of cigarettes smoked adjusted for measurement error using the trivariate method in the sensitivity analysis by varying the magnitude of error correlation between measurements for the exposures (see Tables D1 to D3 in the extended data for more details28).\n\nThe graphs show that varying the magnitude of the correlation between errors in any two exposures affects the estimates for the three exposures. For instance, from Figure 4, increasing the magnitude of the positive correlation between errors in fruit and vegetable intakes increase the mean estimates for both fruit and vegetable intake while it causes a decrease (in absolute value) in the estimate for the average number of cigarettes smoked; decreasing the negative correlation between errors in the measurements for fruit and cigarette smoking decreases (in absolute value) the mean estimates for both fruit and the average number of cigarettes smoked while it leads to an increase in the estimate for vegetable intake (Figure 5). Similarly, a decrease in the magnitude of the negative correlation between errors in vegetable and number of cigarettes smoked causes a decrease (in absolute value) in the estimates for both vegetables and the average number of cigarettes smoked and an increase in the estimates for fruit intake (Figure 6).\n\n\nDiscussion and conclusion\n\nIn this study, we proposed and illustrated a method that adjusts for measurement error in multiple exposures measured with correlated errors in the absence of internal validation data. The method combines external validation data from the literature with the observed self-reported data to adjust for bias in the association between the exposures and the outcome and conduct a sensitivity analysis on the measurement error and correlation between the errors. The advantages of the multivariate method presented in this work includes: (1) the method can be used to adjust for bias in the outcome-exposure association caused by measurement error reported in multiple exposures measured with correlated errors, (2) the method is useful in the absence of the costly internal validation data, provided that external information on the correlation between the observed and the true data or the error correlations of the observed data are plausible within the study context, (3) it can be used in the sensitivity analysis on the effect of uncertainty of the reported validity coefficients, (4) can be used for sensitivity analysis on the magnitude and the direction of correlated errors, (5) the method can adjust for confounding effect in the outcome regression model and (6) This method can be easily implemented on the readily available and free software R as shown in the extended data28. Often, fruit and vegetable intakes are considered as one food group. Our study is relevant because fruit intake and vegetable intake are separately assessed as independent food groups and adjusted for correlated measurement errors.\n\nIn the HBCT study example used for illustration, the estimates for fruit intake, vegetable intake, and the average number of cigarettes smoked adjusted for bias using the trivariate method were almost similar to the estimates adjusted for bias using the univariate method. The slight differences between the bias-adjusted coefficient estimates in the univariate and trivariate methods could be attributed to the weak correlations between errors assumed in this study. Sensitivity analysis on the magnitude of error correlation showed that the estimates obtained using the two methods would be different when stronger error correlations are assumed. Further, from the sensitivity analysis, we found that in a case where multiple exposures are measured with correlated errors, an increase in the magnitude of error correlation between two exposures can increase their estimates and decrease the estimate of the other exposure. From the sensitivity analysis of the level of uncertainty using CI assumed for the validity coefficients, we found that the assumed CI minimally influenced the exposures’ estimates. However, the CIs for the validity coefficients should be reasonably chosen as studies have shown that uncertainty in the estimates may be affected by the level of uncertainty assigned to the validity coefficients14. From our results, we also noted that the presence of measurement error in multiple exposures can bias the association in either direction.\n\nThis study has a few limitations: (1) for simplicity, we assumed that the exposures are measured without systematic bias, i.e., only with random errors. However, in practice, the exposures can be measured with systematic error. In such a case, the systematic error components can be incorporated in the measurement error model and also in estimating the attenuation-contamination matrix; (2) although we can have a multiplicative measurement error structure42, our study assumed an additive measurement error structure. Exposures measured with multiplicative error can be handled using our method by first converting the multiplicative structure to an additive structure through a suitable transformation that linearizes the error structure and (3) our study focused on a subset of current daily smokers, which is not a representative of the HBCT cohort and, therefore, the results are not generalizable.\n\nFrom the findings of this study, we conclude that the multivariate method can be used to adjust for bias in the outcome-exposure association in a case where two or more exposures are measured with correlated errors. This is possible even in the absence of internal validation data provided that there is prior information about the validity of the data collection instruments and the magnitude of the measurement error correlation between the exposures. The method is useful in conducting a sensitivity analysis on the magnitude of measurement error and the sign of the error correlation.\n\n\nData availability\n\nData used in this study are made available to the researcher upon registration and agreeing to the terms and conditions of use in the HSRC web site at http://curation.hsrc.ac.za/ Dataset-565-datafiles.phtml.\n\nFigshare: A Method to Adjust for Measurement Error in Multiple Exposures Measured with Correlated Error in the Absence of Internal Validation Study-Supplementary materials. https://doi.org/10.6084/m9.figshare.13147970.v228\n\nThe file shows the validity coefficient derivation, Proof for the estimate of error variance, R code for implementing the methods and convergence diagnostics results (i.e. Trace plots and ACF plots for the standard deviation and naive regression coefficient estimates of the fruits, vegetables and average number of cigarettes smoked, with explanation) and the sensitivity analysis results (supporting Tables) for varying the magnitude of error correlation between the exposures.\n\nThe extended data are available under the terms of the Creative Commons Zero (CC0) license.",
"appendix": "Acknowledgements\n\nWe thank the University of KwaZulu-Natal for providing the resources needed to conduct our research. Finally, we are grateful to HSRC for allowing us to make use of their data for illustration purpose.\n\n\nReferences\n\nKaaks R, Slimani N, Riboli E: Pilot phase studies on the accuracy of dietary intake measurements in the epic project: overall evaluation of results. european prospective investigation into cancer and nutrition. Int J Epidemiol. 1997; 26 Suppl 1: S26–36. PubMed Abstract | Publisher Full Text\n\nCollese T, Vatavuk-Serrati G, Nascimento-Ferreira M, et al.: What is the validity of questionnaires assessing fruit and vegetable consumption in children when compared with blood biomarkers? a meta-analysis. Nutrients. 2018; 10(10): 1396. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoldbohm RA, van den Brandt PA, Brants HAM, et al.: Validation of a dietary questionnaire used in a large-scale prospective cohort study on diet and cancer. Eur J Clin Nutr. 1994; 48(4): 253–265. PubMed Abstract\n\nPlaete J, De Bourdeaudhuij I, Crombez G, et al.: The reliability and validity of short online questionnaires to measure fruit and vegetable intake in adults: the fruit test and vegetable test. PLoS One. 2016; 11(7): e0159834. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAgudo A: Measuring intake of fruit and vegetables. 2005; Accessed Mar.10,2020. Reference Source\n\nKipnis V, Subar AF, Midthune D, et al.: Structure of dietary measurement error: results of the open biomarker study. Am J Epidemiol. 2003; 158(1): 14–21; discussion 22-6. PubMed Abstract | Publisher Full Text\n\nGleser LJ: Improvements of the naive approach to estimation in nonlinear errors-in-variables regression models. Contemp Math. 1990; 112: 99–114. Reference Source\n\nDay NE, McKeown N, Wong MY, et al.: Epidemiological assessment of diet: a comparison of a 7-day diary with a food frequency questionnaire using urinary markers of nitrogen, potassium and sodium. Int J Epidemiol. 2001; 30(2): 309–317. PubMed Abstract | Publisher Full Text\n\nSubar AF, Kipnis V, Troiano RP, et al.: Using intake biomarkers to evaluate the extent of dietary misreporting in a large sample of adults: the open study. Am J Epidemiol. 2003; 158(1): 1–13. PubMed Abstract | Publisher Full Text\n\nNatarajan L, Pu M, Fan J, et al.: Measurement error of dietary self-report in intervention trials. Am J Epidemiol. 2010; 172(7): 819–827. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKipnis V, Freedman LS, Carroll RJ, et al.: A bivariate measurement error model for semicontinuous and continuous variables: Application to nutritional epidemiology. Biometrics. 2016; 72(1): 106–115. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCarroll RJ, Ruppert D, Stefanski LA, et al.: Measurement error in nonlinear models: a modern perspective. CRC press, 2006. Reference Source\n\nKaaks R, Riboli E, van Staveren W: Calibration of dietary intake measurements in prospective cohort studies. Am J Epidemiol. 1995; 142(5): 548–556. PubMed Abstract | Publisher Full Text\n\nAgogo GO, van der Voet H, van’t Veer P, et al.: A method for sensitivity analysis to assess the effects of measurement error in multiple exposure variables using external validation data. BMC Med Res Methodol. 2016; 16(1): 139. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDellaportas P, Stephens DA: Bayesian analysis of errors-in-variables regression models. Biometrics. 1995; 51(3): 1085–1095. Publisher Full Text\n\nHuang Y, Chen J, Qiu H: Bayesian quantile regression for nonlinear mixed-effects joint models for longitudinal data in the presence of mismeasured covariate errors. J Biopharm Stat. 2017; 27(5): 741–755. PubMed Abstract | Publisher Full Text\n\nLin X: A bayesian semiparametric accelerated failure time model for arbitrarily censored data with covariates subject to measurement error. Commun Stat Simul Comput. 2017; 46(1): 747–756. Publisher Full Text\n\nMuff S, Ott M, Braun J, et al.: Bayesian two-component measurement error modelling for survival analysis using inla—a case study on cardiovascular disease mortality in switzerland. Comput Stat Data An. 2017; 113: 177–193. Publisher Full Text\n\nRosner B, Willett WC, Spiegelman D: Correction of logistic regression relative risk estimates and confidence intervals for systematic within-person measurement error. Stat Med. 1989; 8(9): 1051–1069; discussion 1071–3. PubMed Abstract | Publisher Full Text\n\n> R Core Team: R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. 2018; 2. Reference Source\n\nVan Heerden AC: Non-communicable disease screening and HIV testing and counselling in rural KwaZulu-Natal, South Africa (NCD) 2015. [Data set]. NCD 2015. Version 1.0.: Human Sciences Research Council [distributor], 2016. Publisher Full Text\n\nBarnabas RV, van Rooyen H, Tumwesigye E, et al.: Initiation of antiretroviral therapy and viral suppression after home hiv testing and counselling in kwazulu-natal, south africa, and mbarara district, uganda: a prospective, observational intervention study. Lancet HIV. 2014; 1(2): e68–e76. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNewby PK, Muller D, Hallfrisch J, et al.: Dietary patterns and changes in body mass index and waist circumference in adults. Am J Clin Nutr. 2003; 77(6): 1417–1425. PubMed Abstract | Publisher Full Text\n\nField AE, Gillman MW, Rosner B, Rockett HR: Association between fruit and vegetable intake and change in body mass index among a large sample of children and adolescents in the united states. Int J Obes Relat Metab Disord. 2003; 27(7): 821–826. PubMed Abstract | Publisher Full Text\n\nAzagba S, Sharaf MF: Fruit and vegetable consumption and body mass index: a quantile regression approach. J Prim Care community Health. 2012; 3(3): 210–220. PubMed Abstract | Publisher Full Text\n\nYirga AA, Ayele DG, Melesse SF: Application of quantile regression: Modeling body mass index in ethiopia. The Open Public Health Journal. 2018; 11(1): 221–233. Publisher Full Text\n\nFreedman LS, Schatzkin A, Midthune D, et al.: Dealing with dietary measurement error in nutritional cohort studies. J Natl Cancer inst. 2011; 103(14): 1086–1092. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMuoka AK, Agogo GO, Ngesa OO, et al.: A method to adjust for measurement error in multiple exposures measured with correlated error in the absence of internal validation study-supplementary materials. 2020. http://www.doi.org/10.6084/m9.figshare.13147970.v2\n\nHoff PD: A first course in Bayesian statistical methods. Springer, 2009; 580. Publisher Full Text\n\nPlummer M: rjags: Bayesian Graphical Models using MCMC. R package version 4-8. 2018. Reference Source\n\nLunn D, Spiegelhalter D, Thomas A, et al.: The bugs project: Evolution, critique and future directions. Stat Med. 2009; 28(25): 3049–3067. PubMed Abstract | Publisher Full Text\n\nMartin AD, Quinn KM, Park JH: MCMCpack: Markov chain monte carlo in R. J Stat Softw. 2011; 42(9): 22. Publisher Full Text\n\nPlummer M, Best N, Cowles K, et al.: Coda: convergence diagnosis and output analysis for mcmc. R news. 2006; 6(1): 7–11. Reference Source\n\nPlummer M, Stukalov A, Denwood M, et al.: Package ‘rjags’. update, 16: 1, 2018.\n\nFeskanich D, Rimm EB, Giovannucci EL, et al.: Reproducibility and validity of food intake measurements from a semiquantitative food frequency questionnaire. J Am Diet Assoc. 1993; 93(7): 790–796. PubMed Abstract | Publisher Full Text\n\nWoodward M, Moohan M, Tunstall-Pedoe H: Selfreported smoking, cigarette yields and inhalation biochemistry related to the incidence of coronary heart disease: results from the scottish heart health study. J Epidemiol Biostat. 1999; 4(4): 285–295. PubMed Abstract\n\nEliopoulos C, Klein J, Koren G: Validation of self-reported smoking by analysis of hair for nicotine and cotinine. Ther Drug Monit. 1996; 18(5): 532–536. PubMed Abstract | Publisher Full Text\n\nSecker-Walker RH, Vacek PM, Flynn BS, et al.: Exhaled carbon monoxide and urinary cotinine as measures of smoking in pregnancy. Addict Behav. 1997; 22(5): 671–684. PubMed Abstract | Publisher Full Text\n\nStram DO, Huberman M, Wu AH: Is residual confounding a reasonable explanation for the apparent protective effects of beta-carotene found in epidemiologic studies of lung cancer in smokers? Am J Epidemiol. 2002; 155(7): 622–628. PubMed Abstract | Publisher Full Text\n\nFisher RA: Frequency distribution of the values of the correlation coefficient in samples from an indefinitely large population. Biometrika. 1915; 10(4): 507–521. Publisher Full Text\n\nFisher RA: On the ’probable error’ of a coefficient of correlation deduced from a small sample. Metron. 1921; 1: 1–32. Reference Source\n\nHeid IM, Küchenhoff H, Miles J, et al.: Two dimensions of measurement error: classical and berkson error in residential radon exposure assessment. J Expo Anal Environ Epidemiol. 2004; 14(5): 365–77. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "84307",
"date": "06 May 2021",
"name": "Erica Ponzi",
"expertise": [
"Reviewer Expertise Biostatistics",
"measurement error",
"randomized clinical trials"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper presents an application of measurement error modeling to a dataset from a home-based HIV counseling and testing (HBCT) study. It focuses on the case where multiple variables are measured with error and such errors are correlated. The proposed model is not novel in itself, as Bayesian measurement error models have been used before and extended to the case of multivariate cases, but the application is interesting and the use of the Fisher transformation for the correlation among errors hasn't been employed in these specific cases before.\nI believe this can be an interesting contribution to the field, and the implementation of the presented model can be used in similar studies, which are becoming very common in epidemiology. Nevertheless, since the paper is presented as a method paper, I think more methodological aspects should be examined:\n\n1. Is each error-prone variable assumed to have a classical measurement error structure? This is not explicitly said in the paper, but I think this aspect deserves more attention. Is it reasonable to assume a classical measurement error for all the three variables? Wouldn't a Berkson error, or a mixture of the two also make sense? If we think about some kind of \"rounding\" error, which can be plausible in these cases, a Berkson structure would seem appropriate. It is known that in the presence of a single variable measured with additive Berkson error, and uncorrelated to other variables and to the response, the attenuation problem does not occur but only an increase of uncertainty is observed. In the case of multiple, correlated, errors this is not obvious so I believe such situation should also be explored, and similar models with a Berkson or a mixture error structure should be investigated (or at least the attenuation phenomenon in such cases).\n2. Not all measurement error techniques correct for attenuation simply by dividing by the attenuation factor, see for example the simulation extrapolation technique or the hierarchical Bayesian measurement error models. Adding a latent level for the error eg in a Bayesian framework does not require the attenuation factor to be modeled explicitly and allows for different error structures (see point 1 above) and different correlation structures. The proposed model explicitly estimates the attenuation factor but a more general latent error model can be incorporated to allow for broader applicability (see Bayesian measurement error models in the Carroll book or in some of the cited papers). The author should explain their choice in more details, and try to accommodate different error structures in their model.\n3. The response variable is BMI, ie a continuous variable. What would happen if the response was binary? Do the proposed models generalize to such case?\n3. What would happen if error is correlated with the response? Worth discussing this aspect even if models do not explicitly have to account for it.\n4. It seems that smoking has a higher effect on BMI than vegetables and fruit. Is this supported by other findings?\n5. Have you tried with a higher thinning interval? It avoids correlation between samples.\n6. It would be interesting to see some sensitivity analysis on the priors too.\n7. In Figure 4,5,6 the x-axis correlation between errors. It's difficult to see the effect because of the different scales of smoking and fruit/vegetables. Maybe it would be clearer to have one plot per beta and three lines in each plot corresponding to the error correlation.\n8. I agree with the authors that modeling measurement error brings additional uncertainty in the parameter estimation. This trade-off between bias and variance is a commonly discussed aspect of measurement error modeling. Bayesian models allow to directly incorporate the uncertainty about the error in the posterior estimates of interest. I believe this aspect should be elaborated in the discussion.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "121761",
"date": "25 Mar 2022",
"name": "Kesaobaka Molebatsi",
"expertise": [
"Reviewer Expertise Biostatistical methods",
"particularly of dealing with confounding",
"selection bias",
"measurement or misclassification error and interference."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors have proposed a method that accounts for measurement errors in multiple exposures that are correlated and called it a multivariate method. They have clarified the challenges of ignoring such a problem well in the absence of validation samples, and have compared the multivariate method with other available methods. Their findings from a real data set suggest that the multivariate method can be used to adjust for the bias in the absence of an internal validation study.\nBoth advantages and disadvantages of the multivariate method have been clearly discussed. However, I expected the authors to conduct a simulation study to validate their method further. This is because the real data set falls short to give appropriate information regarding model performance as the true population parameters are unknown. Moreover, it is just one of the many possible samples that one can get from the same population. I am not insisting that the authors should conduct these extra and possibly time-consuming analyses, but they can somehow mention the lack of it as a potential limitation.\nOtherwise, the paper is well written and contributes towards solving an important statistical problem with a sound method.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1486
|
https://f1000research.com/articles/9-1485/v1
|
18 Dec 20
|
{
"type": "Research Article",
"title": "The experience of adherence among hemodialysis patients undergoing therapeutic regimen: a qualitative study",
"authors": [
"Dwi Retno Sulistyaningsih",
"Elly Nurachmah",
"Krisna Yetti",
"Sutanto Priyo Hastono",
"Dwi Retno Sulistyaningsih",
"Krisna Yetti",
"Sutanto Priyo Hastono"
],
"abstract": "Background: Hemodialysis is one of the most widely used renal replacement therapies and the most efficient procedure for managing patients with end-stage renal disease. It can reduce the symptoms of the disease; however, it affects quality of life, resulting in major changes to all areas of life. Thus, patients must adhere to the therapeutic regimen of hemodialysis. Knowledge regarding the adherence of hemodialysis patients to their therapeutic regimens and their experience is needed as it forms the basis for developing comprehensive nursing care and broader nursing practices for renal disease patients. This study aimed to explore patients’ experience of adherence to the therapeutic regimen. Methods: The design of this study was formulated using a qualitative phenomenological approach with a purposive sampling method that involved interviewing 10 selected patients. Data analysis was performed using the Colaizzi method. Results: The results of this study revealed five themes: (1) the most difficult period is the beginning of the hemodialysis program; (2) it is important for patients to perform self-care; (3) nurses play a critical role in improving adherence; (4) there is a decrease in patients’ ability to perform physical activity; and (5) there are supporting and inhibiting factors of adherence to the therapeutic regimens in hemodialysis patients. Conclusions: Adherence to the therapeutic regimen fluctuates. Thus, there is a need to optimize the role of nurses in improving adherence.",
"keywords": [
"Patient experience",
"adherence",
"hemodialysis",
"therapeutic regimen",
"qualitative research",
"renal replacement therapy",
"end-stage renal disease",
"nurses"
],
"content": "Introduction\n\nHemodialysis (HD) is one of the most widely used methods of renal replacement therapies; it is the most efficient procedure for managing end-stage renal disease1–3. Globally, the number of HD patients is increasing every year at an average rate of 7%–9%2–8. Based on data from the United States Renal Data System (USRDS) in 2019, 86.9% of incident late-stage kidney patients began renal replacement therapy using HD in 20179. In Indonesia, based on the 2017 data from the Indonesian Renal Registry, 77,892 patients undergoing HD and about 98% of patients with kidney problems still see HD as the best kidney replacement therapy10; most patients undergo HD twice a week, with each session lasting for four to five hours. According to the Center for Data Information of the Indonesian Ministry of Health in 2017, health care costs for kidney disease were ranked second highest, next to heart disease, in the Health Social Security Institution (BPJS)11.\n\nHD reduces patients’ discomfort but cannot cure, recover, or completely replace kidney function. It also affects patients’ quality of life and may cause major changes to their physiology, which can lead to disability12–15. Patients need to self-manage their condition as well as adjust their behavior and lifestyle while adhering to the therapeutic regimen of HD, all of which rely on compliance with regard to their attendance at HD sessions, medication, fluid restriction, and diet12,16–19. In addition to these four components, physical activity is another essential factor that must be carried out by patients as recommended. The National Kidney Foundation (NKF) recommends patients to undergo moderate physical activity for 30 minutes on most days20, using a therapeutic regimen that focuses primarily on the recovery or maintenance of their quality of life21.\n\nAdherence to the therapeutic regimen is central and paramount to achieving optimal, effective, and successful HD outcomes. Adherence is a crucial factor that helps patients achieve good therapeutic results. It contributes to reducing morbidity, mortality, and the side effects of HD, such as muscle cramps, malnutrition, sepsis, and infections. Moreover, it reduces hospitalization risk and promotes the maintenance of a good quality of life and health3,5,8,21–24. However, failure to adhere to the HD regimen can lead to serious and occasionally fatal conditions, such as hypertension, muscle cramps, arteriovenous fistula (AVF) blockage, dyspnea associated with pulmonary edema, or heart attack due to hyperkalemia, and can lead to poor quality of life, decreased life expectancy, increased morbidity, mortality, and a higher cost and burden on the health care system3,16,17,25.\n\nDespite possessing sufficient knowledge of its detriments, poor adherence to the therapeutic regimen among HD patients has been consistently reported. For example, according to two previous studies, adherence to the prescribed diet was 24.0% and 27.7%, adherence to fluid intake restrictions was 24.5% and 31%, adherence to HD schedule attendance was 52.0% and 91.0%, and adherence to medication was 66.5% and 81.0%, respectively3,17. Poor physical functioning is perhaps the most pervasive and disabling disturbance in patients’ dialysis, and previous studies have suggested the benefits of physical exercise26. HD patients usually find it difficult to adhere to at least one of the therapeutic regimens. Adherence to therapy in patients with chronic disease is a multifaceted issue that requires holistic and varied approaches. Understanding of the experience of patients in regard to HD therapy adherence is needed to develop comprehensive nursing care and broad nursing practices for kidney disease patients21. Through the sharing of these experiences, other patients can become more aware of the lifestyle choices and behaviors that they may need to change, allowing them to improve their own adherence. Researchers who have a special interest in medical surgical nursing, particularly in hemodialysis, need to carry out comprehensive research to gain a deeper understanding of therapeutic adherence. Health workers, especially nurses, must pay attention to patients’ experiences and perspectives.\n\nDespite the number of studies that have been conducted on adherence among HD patients, none have explored the experience of patients in adhering to the therapeutic regimen. Thus, this study aimed to identify and explore HD patients’ experiences of adherence to the therapeutic regimen regarding factors such as scheduled attendance, medication, fluid restriction, diet, and physical activity.\n\n\nMethods\n\nEthical approval was obtained from the Faculty of Nursing of University of Indonesia (No. 285/UN2.F12.D/HKP.02.04/2018), and the lead researcher attended qualitative research method training prior to the study. Patients who met the inclusion and exclusion criteria were identified by the researchers as eligible for the study. Eligible participants were informed of the objective and procedures of this research, and those who agreed to participate signed the informed consent form. To protect the confidentiality of the patients, patient names have been removed from all data.\n\nThis qualitative study uses a phenomenological approach. Data were gathered through in-depth interviews recording the experiences of several HD patients on various issues relating to therapeutic regimen adherence. The research was conducted from October to December 2018. In addition, interdialytic weight gain (IDWG) data were analyzed to check if they support adherence to therapeutic regimen therapy.\n\nSampling was performed using a purposive technique. The study was conducted at the HD unit of two hospitals in Semarang, Central Java, Indonesia by the first author. She is a lecturer in medical surgical nursing with Master’s degree in nursing and specialist of medical surgical nursing focus on caring for patients with HD. Some training had been done in qualitative research methodology. In this study, she reported the findings to the other authors after data collection. The inclusion criteria were: (1) undergoing HD treatment twice a week for a minimum of one year, (2) aged 18 years or older, (3) able to verbally communicate in Indonesian or Javanese, (4) has no cognitive impairment, and (5) willing to participate. Before conducting the research, the researcher and the chairman of the HD unit identified patients who could participate in the study. Subsequently, the researcher explained the study’s objective, benefits, rights, risks, confidentiality, and procedure for participation. Thereafter, the researcher approached the patients and built a trusting relationship, then enquired about their willingness to participate in the study; a total of 10 patients agreed to participate and signed the informed consent form27. None of the patients who were asked to participate in this study refused. We achieved data saturation from these 10 participants and found no new information; thus, no other patients were recruited.\n\nEach participant underwent an in depth semi-structured interview for approximately 30 to 40 minutes in a room within the HD unit, where there were only the researchers and the participant. The interview was conducted using a guide27, which was developed based on the concept of adherence to the therapeutic regimen and consultations with HD experts. The guide included questions relating to the experiences of the participants regarding adherence to the therapeutic regimen, such as scheduled attendance, medication, fluid restriction, diet, and physical activity. This concept was obtained by searching and reviewing references related to compliance of HD patients who were undergoing therapeutic regimens. From this arose questions that need to be asked to explore patients’ experiences and the supporting and inhibiting factors in undergoing therapeutic regimens, including implementing HD, based on the program, medication, fluid restriction, diet, and physical activity. A list of questions was compiled. Before being used in research, this interview guide was reviewed by a clinical expert; a senior nurse who a Master’s degree in nursing and has been working in the hemodialysis unit at a hospital in Central Java. During the review, the expert was asked to provide an assessment and feedback on whether the questions were formulated appropriately. After doing a review, the expert judged that the questions were appropriate. The expert suggested adding questions, “what is the impact or body’s response when the patient could implement the therapy regimen and when they couldn’t”.\n\nThe interviews were recorded, with the participants’ consent, to make it easier to transcribe them verbatim. To maintain data security and confidentiality, the audio recordings were coded and stored in locked cabinets and computer files protected with passwords, and not disseminated to unauthorized parties. In addition, the backup copy shall only be maintained for 10 years and used exclusively for research purposes and publication of the results. Field notes were also taken during the interviews. Hence, the data collected in this study included transcripts of in-depth interviews, field notes, and data from medical records relating to IDWG.\n\nBased on the approach of Colaizzi28, manual data analysis was conducted through the following steps: (1) we documented the collected data by creating verbatim transcripts of the in-depth interviews and the field notes. The transcripts were then stored electronically, and a hardcopy was made to facilitate the analysis. (2) We repeatedly read all the transcripts. (3) Three researchers coded the data. We reviewed and analyzed the transcripts, selecting significant statements in connection with the research objectives,which we then highlighted as keywords.These keywords were referred to as initial coding, followed by axial coding. (4) Keywords with similar meanings were classified into categories. Similar categories were then grouped to form themes that fit the research objective (selective coding). (5) The results were elucidated and integrated using a deep and complete narrative description. (6) Finally, the new data that emerged were converted into a narrative description.\n\nThe robustness of the study findings was based on the following four criteria: credibility, dependability, confirmability, and transferability29. Credibility of the data obtained by the researcher conveys the results of the interviews that were analyzed, in order to verify that the data obtained is in accordance with their submission. In addition, discussions and consultations with experts regarding the categories and themes were conducted. Consultation was conducted with a Professor and lecturer from the Faculty of Nursing, University of Indonesia. She is an expert in a qualitative research methodology, author of books on qualitative research methodology and reseacher with publications in national and international journals. The consultation was about the accuracy of the categories and themes. The state of the themes must be accurate and clear. Dependability was enhanced by thoroughly analyzing data using a structured approach, striving to interpret the results of the study correctly, and involving all researchers in developing the categories and themes. Confirmability was derived by showing the results of the interviews that were made temporarily, then creating themes after consulting with the expert. The researcher also codified the information and then conducted discussions and consultations with the same expert as for credibility. The discussion and consultation was about the coding process. The consultation was conducted after the reseachers made a draft of data analysis that contained the groups of keywords from participant statements, catagories and themes. The expert gave suggestions for rearranging the keyword groups because there were still participant statements that didn’t match with the group. This process was conducted to reach agreement on the categories and themes obtained. Transferability is ensured by performing an external check where the researcher presented the results of the analysis to patients who did not store the results. The researcher asked two other hemodialysis patients who did not participate in this study to read the results of the study and asked for responses about the clarity the results study. Both of these patients stated that they understand the results of the study. They said that their experience of hemodialysis was similar to the patients in this study. The principle of transferability is a form of external validity, which shows that the research results from one study population can be applied to other settings or groups of participants30.\n\n\nResults\n\nThe characteristics of participants (Mage = 48.3 years; male-to-female ratio = 1:1) are shown in Table 1. The average HD period was 3.15 years. Secondary schooling was the highest level of education for most participants. The majority of the participants were married (60%), not employed (70%), had permanent vascular access (80%), and had an average IDWG of 4%–6% between their treatments (70%). Additionally, most of them had hypertension (60%) before HD and attended their HD treatment alone (70%)31.\n\nHD, hemodialysis; DM, diabetes mellitus.\n\nData analysis was based on the transcripts of the in-depth interviews of the participants’ experiences regarding adherence to the therapeutic regimen, such as scheduled attendance, medication, fluid restriction, diet, and physical activity31. The five themes that were identified after data analysis are listed in Table 2.\n\nThis theme was generated from the responses of seven participants consisting of two categories, namely (1) the reasons given by patients who did not attend HD as scheduled and (2) those who were often hospitalized at the beginning of HD.\n\n“I go to alternative therapy, I get a massage and do not need to undergo dialysis, I tried it, but unfortunately, I failed to keep at it” (p1).\n\n“I felt better and decided not to undergo HD” (p5).\n\n“Initially, I was not aware of the schedule, and as a result, I missed three sessions before being hospitalized” (p6).\n\nPatients were often hospitalized after starting their HD therapy due to their deteriorating physical condition.\n\n“I was hospitalized for at least a year at the beginning of undergoing HD” (p6).\n\nThis theme was generated from the responses of six participants regarding patients’ self-care to perform the HD therapeutic regimen and feel better.\n\n“I always attended scheduled HD programs despite the fact that my condition was deteriorating and I didn’t have any money” (p2).\n\n“I took the medication as prescribed to prevent my health from deteriorating, thereby making things difficult for my family” (p2).\n\nHD patients were also expected to decrease their fluid intake and embrace a good diet.\n\n“Sucking ice” (p1).\n\n“Eating candy to decrease nausea” (p1).\n\nHD patients usually felt better after performing the HD therapeutic regimen.\n\n“I feel better when I restrict my fluid intake and only gain a little weight and I don’t feel good when I drink a lot and my weight increases” (p3).\n\nThis theme was generated from the responses of all 10 participants regarding the following issues: (1) the patients do not completely understand the therapeutic regimen they need to adhere to, (2) nurses are required to advocate adherence to their patients, and (3) nurses are required to possess high nursing skills. Participants did not adhere to the required medicine intake and understood that fluid and diet restrictions usually influenced their daily fluid and dietary intake.\n\n“I did not take the medicine since I felt healthy” (p7).\n\n“I only took the hypertension drug, leaving out others, since I was afraid of the effect it will have on my womb” (p3).\n\n“A glass a day may be” (p5, p6, p9).\n\n“I never measure the water I drink, if I feel thirsty I drink a little” (p2).\n\n“Whatever is served” (p4). “I eat as much as I want to, no restrictions” (p1). “I eat it all” (p7). “I rarely eat fruit” (p8).\n\nHowever, patients who do not adhere to the HD therapeutic regimen or end their sessions before the scheduled time are usually asked by the nurses to sign a consent letter.\n\n“I was asked to sign if I requested to be treated by HD for 4 hours” (p9).\n\nIn addition, one of the patients stated that nurses should possess high nursing skills, especially for handling patients with difficult vascular accesses:\n\n“Every time I undergo HD, the nurses always find it hard injecting me. My friends and the nurses always pity me...” (p2).\n\nThis theme was generated from the responses of seven participants relating to the limitations in physical activity and the body’s responses after physical activity, as illustrated by these statements from the participants:\n\n“Cooking” (p2, p9), “washing” (p2, p3).\n\n“I only sweep” (p2).\n\n“Taking care of grandchild” (p5, p9, p10).\n\n“I no longer participate in any sporting activity” (p1).\n\nIn terms of physical activity, some participants gave the following response:\n\n“I cannot walk far at the moment” (p6), “I tend to get out of breath” (p5, p9).\n\nThis theme was generated from the responses of seven participants relating to supporting and inhibiting factors.\n\n“All my family members supported me...” (p2), “wife” (p7, p8), “family” (p8), “husband, parents, others family...” ( p3).\n\n“It is my obligation to go for HD” (p1).\n\n“From myself…”(p7), “and my children who are still small...” (p4).\n\n“My colleagues supported me...” (p3).\n\n“The nurses are friendly here; they like to joke around, and I love it. Alhamdulillah, all the nurses are nice and they always motivate me” (p2).\n\nApart from these positive factors, there are also factors that do not support adherence to the therapeutic regimen, such as seasonal weather, cost, and environment.\n\n“I cannot drink less at home because it gets very hot during summers\" (p1).\n\n“Sometimes, the cost is an issue. I often borrow money from my nephew.” (p2).\n\n“I don’t go for HD therapy whenever there is a flood” (p2).\n\n\nDiscussion\n\nHD causes major changes in the lives of patients with chronic kidney disease, affecting all areas of their life. However, non-adherence to HD therapy usually results in the deterioration of the patient’s psyche7,13,14. For most patients, the initial period of undergoing HD is usually the most difficult, since they are in a transitional period and must adapt to changes. However, not all patients are well prepared, and some are unaware of the importance of attending the therapy as scheduled. Some patients even use alternative medication and do not undergo HD despite their kidney disease because they feel healthy. This finding is in line with that of a previous study stating that patients with chronic kidney disease experience a lack of adequate emotional support from nurses during the transitional period of undergoing HD32.\n\nIn addition, many patients experience health-related conditions that could cause them to become hospitalized at the commencement of their HD therapy. This may be attributed to their chronic kidney disease, which is usually asymptomatic and detected late. Additionally, some patients with many of the risk factors related to chronic kidney disease are unaware of its complications and how to prevent them; thus, their kidney disease eventually progresses to its end-stage, which involves kidney replacement therapy. Patients who undergo HD for the first time tend to have poor general health with high comorbidities since such conditions are usually reported late33. Delayed referral from primary care usually limits the choice of dialysis, and inadequate time for preparation can increase the risk of complications associated with the central venous catheters used for dialysis34.\n\nHD has the capacity to reduce patients’ complaints, but it cannot cure or enable them to fully recover from the disease despite undergoing HD permanently. As a result, patients face many challenges, changes, and stressful situations, such as fluid restrictions, diet, associated medical conditions, and loss of the normal daily routine that they were previously used to7,34. Therefore, self-care is needed on the part of patients in order to maintain their quality of life, health, and general welfare. Patients must keep undergoing HD as scheduled, take the medications as prescribed, manage their fluid intake, engage in physical activities, and meet all their nutritional needs. Based on their experiences in this study, patients observe differences in their physical health when they undergo the HD therapeutic regimen. For example, they feel optimally healthy, experience no shortness of breath, and have better psyche when they regulate their fluid consumption. However, they complain of discomforts, such as body aches, shortness of breath, swelling, inability to sleep, and a swollen stomach whenever they fail to do so. This finding is in line with that of a previous study stating that patients have the ability to perform self-care by maintaining their lifestyles, in accordance with that recommended by the HD therapeutic regimen, such as taking medication, restricting fluids, and dieting, among others35.\n\nPatients are responsible for many aspects of their therapy. For instance, most of the patients in this study restricted their fluid intake without a specific means of measurement, did not make any special modifications to their lifestyle, and ate what was normally available to them.\n\nPatients do not always have a valid and accurate understanding of the amount of medication they should take. For example, average patients consume medication based on their personal knowledge and judgment. Moreover, they often build their understanding according to their beliefs and common sense and they further use this understanding as an alibi to justify what happens to them7. However, in practice, patients do not always follow what is recommended. Patients’ education level and ability to understand the information provided as well as the nurses’ methods of delivering the said information are vital factors. This finding is in line with that of a previous study stating that due to time constraints, patients neither receive all the necessary information, nor can they understand the behavioral modifications required since the information is too complicated for them to comprehend34. However, knowledge is a vital tool to empower patients and stabilize their health. Hence, a lack of proper understanding is an obstacle and the information provided by doctors and nurses is crucial to achieve total compliance7,34,36.\n\nMost of the participants in this study found it difficult to limit their fluid intake mainly because of thirst. This finding is in line with that of a previous study stating that the most prominent theme is the challenge in controlling fluid intake37. This difficulty in restricting fluids causes an increase in IDWG, wherein an increase of<4% is normal, 4%–6% is average, and >6% is hazardous38. In our study, we found that majority of the participants (70%) fell within the average category. An average increase of 4%, 5%, and 6% was exhibited by 4, 3, and 1 participants, respectively. The IDWG should be lower than 4% of dry body weight39. Difficulty in maintaining weight during the HD period is a very common problem among many patients7. IDWG is an indicator of excess fluid that needs to be removed during dialysis, and a greater percentage indicates problems in controlling fluid intake40,41.\n\nNurses are directly responsible for providing care to patients before, during, and after undergoing HD. AVF cannulation is an important skill that all HD nurses should have. This is essential to patients who have trouble since they have temporary vascular access or small blood vessels. Such patients can experience greater levels of pain, discomfort, and anxiety. This finding is in line with that of a previous study stating that not all nurses show a sufficiently high level of competence, which makes the patients more anxious42. Needle insertion error is the main cause of recirculation43–45. Negative experiences during cannulation contribute to patients’ fear of AVF damage and distrust in the HD staff44. The results of other studies show that nurses must have technical skills combined with appropriate experiences, such as the ability to perform venous puncture, as it plays an important role in enhancing patient comfort46.\n\nIn our study, HD was carried out in accordance with a pre-determined time limit of approximately four to five hours for each procedure. However, there are patients who end their sessions before the allotted time due to medical and non-medical reasons; for example, some patients exhibit severe intra-dialysis complications and thereby discontinue the scheduled HD, while some cite lack of time and transportation. For patients who want to end HD prematurely, nurses usually explain the consequences to them first, and then ask them to sign an informed consent form, stating that they fully understand the consequences of their action and accept all the risks that their discontinuation might ensue. Ultimately, nurses can improve HD effectiveness by adhering to the accurate duration of each session45.\n\nThe ability of HD patients to carry out physical activities changes over the course of their treatment. The most noticeable changes are decreased mobility, limitations in performing certain physical activities, shortness of breath, fatigue, and weakness, all of which hinder patients from completing the course of the treatment. There are three main factors that affect physical activity: renal failure, side effects of renal replacement therapy, and worsening of comorbidities. Low physical ability is also caused by uremic intoxication, anemia, mineral and metabolic abnormalities, increased cardiovascular risk combined with high comorbid disease, uremic sarcopenia, decreased muscle strength due to muscle catabolism, and metabolic waste47–49. Patients experience physical challenges in life, which are reflected by limitations in carrying out certain physical activities. These limitations are caused by low energy and weakness due to restrictions on food and fluid intake, excess fluid, and increase in metabolic waste in the patient's body36.\n\nThere are additional factors that influence adherence; some are supporting, while others are inhibiting. Supporting factors usually stem from patients, their families, nurses, and colleagues at work. Inhibiting factors include seasonal weather, cost, and occupation. Supporting factors can motivate individuals to continue the therapeutic regimen as recommended, thereby reducing the risk of hospitalization; therefore, the ability of patients to self-motivate is very important50. Meanwhile, their family and friends could counsel and encourage them to undergo the HD therapeutic regimen as recommended. The availability of adequate support could reduce economic costs, and increase patients’ participation, respect, cooperation, hope, trust, welfare, family health, and ultimately, adherence to the therapeutic regimen51. This finding is in line with that of a previous study stating that support has a positive influence on chronic kidney disease patients through several means that help them cope better, minimize their stress, assist with practical problems (e.g., accessing health services), and improve their psychosocial functions and adherence to therapy52.\n\nNurses also play an important role by ensuring that the patients undergo their therapy programs as recommended. They also accompany patients undergoing HD and are involved in long-term intensive interactions with patients. Additionally, nurses educate and counsel while administering the therapeutic regimen, motivate and convince patients to undergo the program, provide them with emotional support, and participate in monitoring, evaluating, and maintaining long-term communication with them. This finding is in line with that of a previous study stating that nurses play a key role in supporting patients undergoing HD, understanding the patients’ lives, and ensuring that the information provided will help them adapt to the situation, needs, and demands of HD7.\n\nMeanwhile, the inhibiting factors are environmental and economic in nature. Environmental factors such as the weather prevent patients from going to the hospital to undergo their HD therapeutic regimen. Indonesia experiences two seasons: dry and wet. In the wet season, flooding makes it difficult to reach the hospital to carry out HD. In the dry season, hot temperatures make it difficult for patients to comply with their fluid restrictions, since they are thirstier than usual and hence, drink more than the prescribed amount. This, in turn, causes weight gain during the HD treatment. Thus, fluid restriction adherence is very difficult in tropical regions53.\n\nIn addition, income, cost, and transportation are the major economic inhibiting factors. Patients generally experience a decrease in income since they cannot work. Moreover, those who sell food and beverages for a living find it difficult to undergo HD as the temptation to eat and drink the products they sell is great. Most HD patients in Indonesia receive financial assistance from the BPJS. However, they still must pay for the process in terms of transportation, their food, and for their accompanying family members. Sometimes, they may need to buy medications that are not covered by the BPJS. The results of one study showed that the cost of transportation to access HD service facilities is the highest cost factor for patients with health insurance54. This is different from the United States where transportation and transit services are provided by the health system for disabled people, including patients with end-stage kidney disease55.\n\nIn this study, the minor theme is the importance of educating patients regarding adherence to the HD therapeutic regimen. The participants are sufficiently educated, but poor adherence to the therapeutic regimen in HD patients is still consistently reported. Therefore, nurses, as one of the key health workers with the responsibility of educating patients, should promote patient adherence to the HD therapeutic regimen.\n\n\nConclusion\n\nThe experience of HD patients of adhering to the HD therapeutic regimen is fluctuating and personal in nature; it is influenced by factors that support and inhibit the process. The initial period is usually the most difficult, since patients experience multiple unfavorable conditions that they need to adapt to. Nurses play a critical role in improving adherence to therapeutic regimens. Based on the experiences of patients regarding being able to carry out therapeutic regimens in accordance with the recommendations as well as being unable to make changes to their conditions, it is important for them to practice self-care to maintain their physical condition.\n\n\nNursing implications\n\nAdherence to the therapeutic regimen is fluctuating; hence, there is a need to strengthen the supporting factors and weaken the inhibiting factors. Nurses are one of the health professionals tasked with the important role of providing clinical care and support needed by patients to achieve success during hemodialysis therapy. Thus, they must exercise their role, especially in preparing and assisting patients during the initial period of undergoing HD. They can achieve this through related education and counseling with respect to the therapeutic regimen, thereby making patients more likely to undergo the process as recommended. Nurses also need to continuously improve their knowledge and skills in carrying out nursing care. Lastly, they need to improve the quality of their interaction with patients during HD.\n\n\nLimitations of the study\n\nIn this study, the data obtained from medical records were only those of body weight. Laboratory examination data supporting the therapeutic regimen such as phosphate, sodium, and albumin levels were not obtained, because laboratory tests are rarely carried out and are not covered by BPJS. Thus, future studies should address this issue.\n\n\nData availability\n\nZenodo: Underlying data the qualitative study: The experience of adherence among hemodialysis patients undergoing therapeutic regimen: a qualitative study. http://doi.org/10.5281/zenodo.430129631.\n\nThis project contains the following underlying data:\n\n- Interview transcripts in English in DOCX format\n\n- Interview transcripts in original language in DOCX format\n\n- Demographic and clinical characteristics of each participant in XLSX format\n\nZenodo: Extended data the qualitative study: The experience of adherence among hemodialysis patients undergoing therapeutic regimen: a qualitative study. http://doi.org/10.5281/zenodo.430133827.\n\nThis project contains the following extended data:\n\n- Informed consent in English.docx\n\n- Informed consent in original language.docx\n\n- Interview guidance in English.docx\n\n- Interview guidance in original language.docx\n\nZenodo: COREQ checklist for “The experience of adherence among hemodialysis patients undergoing therapeutic regimen: a qualitative study”. http://doi.org/10.5281/zenodo.430111456.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International (CC-BY 4.0).",
"appendix": "Acknowledgements\n\nSpecial thanks to the hemodialysis patients who have been willing to be participants in this study.\n\n\nReferences\n\nAglebese MO: Implementing a Culturally Sensitive Fluid and Dietary Educational Intervention in an Outpatient Dialysis Center. Doctor of Nursing Practice (DNP) Capstone Projects. paper no. 58. 2015; Accessed May 16th 2020. Reference Source\n\nEfe D, Kocaoz S: Adherence to diet and fluid restriction of individuals on hemodialysis treatment and affecting factors in Turkey. Jpn J Nurs Sci. 2015; 12(2): 113–123. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text\n\nNaalweh KS, Barakat MA, Sweileh MW, et al.: Treatment adherence and perception in patients on maintenance hemodialysis: a cross – sectional study from Palestine. BMC Nephrol. 2017; 18(1): 178. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSuwitra K: Interna Publishing. Ilmu Penyakit Dalam. 2014; 5(2). Reference Source\n\nCicolini G, Palma E, Simmoneta C, et al.: Influence of family carers on haemodialyzed patients’ adherence to dietary and fluid restrictions: an observational study. J Adv Nurs. 2012; 68(11): 2410–2417. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text\n\nLi H, Jiang YF, Lin CC: Factors associated with self-management by people undergoing hemodialysis: A descriptive study. Int J Nurs Stud. 2014; 51(2): 208–216. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text\n\nTovazzi ME, Mazzoni V: Personal Paths of Fluid Restriction In Patients on Hemodialysis. Nephrol Nurs J. 2012; 39(3): 207–215. Accessed May 16th 2020. PubMed Abstract\n\nGerogianni GK, Babatsikou FP: Identification of stress in chronic haemodialysis. Health Science Journal. 2013; 7(2): 169–176. Accessed May 16th 2020. Reference Source\n\nUnited Stated Renal Data System: USRDS Annual Data Report ESRD In The United States. 2019; Accessed May 16th 2020.\n\nIndonesian Renal Registry (IRR): 10th Report Of Indonesian Renal Registry. 2017; Accessed May 16th 2020. Reference Source\n\nInfoDatin Pusat Data dan Informasi Kementerian KesehatanRepublik Indonesia. Situasi Penyakit Ginjal Kronis. 2017; Accessed May 16th 2020. Reference Source\n\nRyu HJ, Jeon HJ, Sun HK, et al.: Repeated education improves diet compliance in maintenance Hemodialysis Patients. Int J Urol Nephrol. 2014; 4(2): 063–068. Accessed May 16th 2020. Reference Source\n\nBahadori M, Ghavidel F, Mohammadzadeh S, et al.: The effects of an interventional program based on self-care model on health-related quality of life outcomes in hemodialysis patients. J Edu Health Promot. 2014; 3: 110. Accessed April 15th 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMulder A, Sikken-Kersten L: Spirituality during hemodialysis: A pilot study. Nephrologi Nurs J. 2016; 43(4): 323–329. Accessed May 16th 2020. PubMed Abstract\n\nShahdadi H, Rahnama M: Experience of Nurses in Hemodialysis Care: A Phenomenological Study. J Clin Med. 2018; 7(2): 30. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOh HS, Park JS, Sheo WS: Psychosocial influencers and mediators of treatment adherence in haemodialysis patients. J Adv Nurs. 2013; 69(9): 2041–53. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text\n\nChan YM, Zalillah, Hii SZ: Determinants of Compliance Behaviours among Patients Undergoing Hemodialysis in Malaysia. PLoS One. 2012; 7(8): e41362. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNabolsi MM, Wardam L, Al-Halabi JO: Quality of life, depression, adherence to treatment and illness perception of patients on haemodialysis. Int Nurs Pract. 2015; 21(1): 1–10. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text\n\nPoveda V, Amado L, Filgueiras M, et al.: End-stage renal disease adherence questionnaire: translation and validation to the Portuguese language. Ren Fail. 2016; 38(10): 1633–1638. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text\n\nK/DOQI Workgroup: K/DOQI clinical practice guidelines for cardiovascular disease in dialysis patients. Am J of Kidney Dis. 2005; 45(4 Suppl 3): S1–153. Accessed May 16th 2020. PubMed Abstract\n\nGuerra-Guerrerro V, Plazas MPC, Cameron BL, et al.: Understanding the Life Experience Of People on Hemodialysis: Adherence To Treatment and Quality of Life. Nephrol Nurs J. 2014; 41(3): 289–97. Accessed May 16th 2020. PubMed Abstract\n\nVaiciuniene R, Kuzminskis V, Ziginskiene E, et al.: Adherence to treatment and hospitalization risk in hemodialysis patients. J Nephrol. 2012; 25(5): 672–678. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text\n\nShimoda T, Matsuzawa R, Yoneki K, et al.: Changes in physical activity and risk of all-cause mortality in patients on maintence hemodialysis: a retrospective cohort study. BMC Nephrol. 2017; 18(1): 154. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFilho JCD, Matsuzawa R, Yoneki K, et al.: Physical activity level of patients on hemodialysis: a cross-sectional study. Fisioterapia e Pesquisa. 2016; 23(3): 234–240. Accessed May 16th 2020. Publisher Full Text\n\nSuganthi S, Porkodi A, Geetha P: Assess the Illness Perception and Treatment Adherence among Patient with End Stage Renal Disease. Iran J Nurs Midwifery Res. 2019; 25(1): 12–17. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nManfredini F, Mallamaci F, D'Arrigo G, et al.: Exercise in Patients on Dialysis: A Multicenter, Randomized Clinical Trial. J Am Soc Nephrol. 2017; 28(4): 1259–1268. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDwiRetno S, Elly N, YettiKrisna, et al.: Extended data the qualitative study \"The experience of adherence among hemodialysis patients undergoing therapeutic regimen: a qualitative study [Data set]. Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4301338\n\nPolit DF, Beck CT: Nursing Research: Generating and Assessing Evidence for Nursing Practice9th Edition. 2012. Reference Source\n\nStreubert HJ, Carpenter DR: Qualitative Research in Nursing: Advancing the Humanistic Imperative. 2011. Reference Source\n\nMoleong LJ: Metodologi Penelitian Kualitatif. 2010. Reference Source\n\nDwiRetno S, Elly N, YettiKrisna, et al.: Underlying data the qualitative study \"The experience of adherence among hemodialysis patients undergoing therapeutic regimen: a qualitative study [Data set]. Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4301296\n\nSturesson A, Ziegert K: Prepare the patient for future challenges when facing hemodialysis: nurses’ experiences. Int J Qual Stud Health Well-being. 2014; 9: 22952. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUmami V, Lydia A, Nainggolan G, et al.: Pengembangan Model Prediksi Mortalitas 3 Bulan Pertama pada Pasien Penyakit Ginjal Kronik yang Menjalani Hemodialisis. Jurnal Penyakit Dalam Indonesia. 2015; 2(3). Accessed May 16th 2020. Publisher Full Text\n\nClarkson KA, Robinson K: Life on Dialysis: A Lived Experience. Nephrol Nurs J. 2010; 37(1): 29–35. Accessed May 16th 2020. PubMed Abstract\n\nLaurie S: Dorthea Orem’s Self Care Theory as Related To Nursing Practice in Hemodialysis. Nephrol Nurs J. 2009; 36(4): 419–21. Accessed May 16th 2020. PubMed Abstract\n\nBayhakki, Hatthakit U: Lived Experiences of Patients On Hemodialysis: A Meta-Synthesis. Nephrol Nurs J. 2012; 39(4): 295–304. Accessed May 16th 2020. PubMed Abstract\n\nAnderson CAF: Lived Experiences: Hemodialysis and Adherence. J Nephrol Soc Work. 2011; 35. Accessed May 16th 2020. Reference Source\n\nYetti K: Pengaturan Cairan Secara Mandiri pada Pasien Yang Menjalani Hemodialisis. Jurnal Keperawatan Indonesia. 2001; V(2). Accessed May 16th 2020. Publisher Full Text\n\nBossola M, Pepe G, Vulpio C: The Frustrating Attempt to Limit the Interdialytic Weight Gain in Patients on Chronic Hemodialysis: New Insights Into an Old Problem. J RenNutr. 2018; 28(5): 293–301. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text\n\nWileman V, Chilcot J, Armitage CJ, et al.: Evidence of improved fluid management in patients receiving haemodialysis following a self-affirmation theory-based intervention: A randomised controlled trial. Psychol Health. 2016; 31(1): 100–114. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text\n\nLee MJ, Doh FM, Kim CH, et al.: Interdialytic Weight Gain and Kardiovascular Outcome in Incident Hemodialysis Patients. Am J Nephrol. 2014; 39(5): 427–435. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text\n\nSciberras A, Scerri J: Facilitators and barriers to treatment with haemodialysis among persons in mid-adulthood: an interpretative phenomenological analysis. Scand J Caring Sci. 2017; 31(4): 695–701. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text\n\nHarwood L, Wilson B, Goodman M: Cannulation Outcomes of the Arteriovenous Fistula for Hemodialysis: A Scoping Review. Nephrology Nursing Journal. 2017; 44(5): 411–425. Accessed May 16th 2020. PubMed Abstract\n\nWilson B, Harwood L: Outcomes for Successful Cannulation of the Arteriovenous Fistula: Perspectives from Patients on Hemodialysis. Nephrol Nurs J. 2017; 44(5): 381–388. Accessed May 16th 2020. PubMed Abstract\n\nRezaie O, Shahgholian N, Shahidi S, et al.: Assessment of hemodialysis adequacy and its relationship with individual and personal factors. Iran J Nurs Midwifery Res. 2016; 21(6): 577–582. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNobahar M, Tamadon MR: Barriers to and facilitators of care for hemodialysis patients; a qualitative study. J Renal Inj Prev. 2016; 5(1): 39–44. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAucelle F, Gesuete A, Battaglia Y: A “Nephrological “ Approach to Physical Activity. Kidney Blood Press Res. 2014; 39(2–3): 189–96. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text\n\nCapitanini A, Lange S, D'Alessandro C, et al.: Dialysis Exercise Team: The Way to Sustain Exercise Programs in Hemodialysis Patients. Kidney Blood Press Res. 2014; 39(2-3): 129–33. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text\n\nAnding K, Bar T, Trojniak-Henning J, et al.: A structured exercise programme during haemodialysis for patients with chronic kidney disease: clinical benefit and long-term adherence. BMJ Open. 2015; 5(8): e008709. Accessed May 16th 2020. Publisher Full Text\n\nRahimi F, Oskouie F, Naser O, et al.: The effect of self-care on patients undergoing Hemodialysis in the Sanandaj Hospitals affiliated to Kurdistan University of Medical Sciences in 2016. Bali Med J. 2017; 6(3): 684–689. Accessed May 16th 2020. Publisher Full Text\n\nZolfaghari M, Asgari P, Bahramnezhad F, et al.: Comparison of two educational methods (family-centered and patient-centered) on hemodialysis: Related complications. Iran J Nurs Midwifery Res. 2015; 20(1): 87–92. Accessed May 16th 2020. PubMed Abstract | Free Full Text\n\nTheodoritsi A, Aravantinou ME, Gravani V, et al.: Factors Associated with the Social Support of Hemodialysis Patients. Iran J Public Health. 2016; 45(10): 1261–1269. Accessed May 16th 2020. PubMed Abstract | Free Full Text\n\nSmyth W, Hartig V, Hayes M, et al.: Patients’ Adherence To Aspects of Haemodialysis Regimens in Tropical North Queensland, Australia. J Ren Care. 2015; 41(2): 110–118. Accessed May 16th 2020. PubMed Abstract | Publisher Full Text\n\nWiguna GNC, Ahmad RA, Utarini A: Biaya Pelayanan Hemodialisis Peserta Asuransi Kesehatan Menurut Perspektif Pasien Di Rumah Sakit Umum Daerah Tipe B di Propinsi Bali. Jurnal Manajemen Pelayanan Kesehatan. 2013, 16(01). Accessed May 16th 2020. Reference Source\n\nPark S, Kear TM: Current State-of-Practice: Transportation for Patients with End Stage Renal Disease. Nephrol Nurs J. 2017, 44(4): 309–315. Accessed May 16th 2020. PubMed Abstract\n\nDwiRetno S, Elly N, YettiKrisna, et al.: Consolidated criteria for reporting qualitative studies (COREQ): 32-item checklist for \"The experience of adherence among hemodialysis patients undergoing therapeutic regimen: a qualitative study\" [Data set]. Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4301114"
}
|
[
{
"id": "86135",
"date": "21 Jun 2021",
"name": "Maya Clark-Cutaia",
"expertise": [
"Reviewer Expertise Chronic kidney disease",
"patient experience",
"outcomes",
"symptom and self-management",
"disparities in chronic illness"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nPurpose: The stated purpose of this report was to “explore patients’ experience of adherence to the therapeutic [hemodialysis] regimen.” Specifically, the authors set out to identify themes regarding attendance, medication, fluid restriction, diet, and physical activity adherence in a sample of ten individuals undergoing hemodialysis therapy in two hospitals in Indonesia, using a phenomenological approach. While the study adds to the extant body of literature regarding adherence in end stage kidney disease (ESKD), there are some areas that require additional attention and are not entirely consistent with the available literature. In particular, there is literature that explores the adherence to various aspects of the ESKD therapeutic regimen as a whole, and individual components e.g., dietary and fluid restriction.\nThe authors state that “Despite the number of studies that have been conducted on adherence among HD patients, none have explored the experience of patients in adhering to the therapeutic regimen.” - This is not true, and demonstrates a lack of full exploration of the extant literature.\nMethods:\nWe would have expected there to be more delineation about specific aspects of the regimen given the background - this is no different than the body of literature that already exists. The Rigor section is extremely long and provides an unnecessary amount of detail, while detailing regarding the methods for potential reproducibility is minimal.\nResults: Example quotes generally lack context and do not always support the themes as described. The emphasis on the role of nurses does not appear to be consistent with the stated purpose of the analysis. Despite providing quotes that describe symptoms and symptom burden, that concept is not described in results analysis\nDiscussion:\nWe feel the authors could have clearly articulated the following given the data available: how what they found differed or supported the current literature, implications of their results, and recommendations for the future.\nOverall, this paper would work better if it focused on its specific context of the experiences of Indonesian HD patients who are on long term dialysis (>1 year). Then we can compare these results to the existing literature to see how the themes here may be different from data found in other HD populations.\nEnglish language translation is good overall but there are some spots where there is a barrier to understanding the authors’ intentions because of language issues. We recommend another English language edit to improve clarity.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "100752",
"date": "06 Dec 2021",
"name": "Abdolreza Shaghaghi",
"expertise": [
"Reviewer Expertise Health promotion and behavior change"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article represents the dilemmas hemodialysis patients (HDs) are facing in relation to their therapeutic regimen. This is an important aspect of the concern in the health care delivery continuum for HDs. Despite the qualitative nature of the study, applied small sample size and the purposive sampling method, its findings could make major contributions to planning of evidence informed therapeutic interventions and therefore to a better outcome in patients who might otherwise have poor adherence to the prescribed therapeutic regimen. The identified themes that correspond to the supporting and inhibiting factors of therapeutic adherence in the studied HDs are important additions to the scientific literature and could fuels development of comprehensive nursing care for HDs.\nThis manuscript is well written and all required details about the conducted qualitative study was provided by the authors. This include details about the importance of this study, the methodological procedures followed and findings with their robust interpretation. No further amendments are required and my suggestion is to accept the paper as it is.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1485
|
https://f1000research.com/articles/9-1478/v1
|
17 Dec 20
|
{
"type": "Software Tool Article",
"title": "Uncovering host-microbiome interactions in global systems with collaborative programming: a novel approach integrating social and data sciences",
"authors": [
"Jenna Oberstaller",
"Swamy Rakesh Adapa",
"Guy W. Dayhoff II",
"Justin Gibbons",
"Thomas E. Keller",
"Chang Li",
"Jean Lim",
"Minh Pham",
"Anujit Sarkar",
"Ravi Sharma",
"Agaz H. Wani",
"Andrea Vianello",
"Linh M. Duong",
"Chenggi Wang",
"Celine Grace F. Atkinson",
"Madeleine Barrow",
"Nathan W. Van Bibber",
"Jan Dahrendorff",
"David A. E. Dean",
"Omkar Dokur",
"Gloria C. Ferreira",
"Mitchell Hastings",
"Gregory S. Herbert",
"Khandaker Tasnim Huq",
"Youngchul Kim",
"Xiangyun Liao",
"XiaoMing Liu",
"Fahad Mansuri",
"Lynn B. Martin",
"Elizabeth M. Miller",
"Ojas Natarajan",
"Jinyong Pang",
"Francesca Prieto",
"Peter W. Radulovic",
"Vyoma Sheth",
"Matthew Sumpter",
"Desirae Sutherland",
"Nisha Vijayakumar",
"Rays H. Y. Jiang",
"Jenna Oberstaller",
"Swamy Rakesh Adapa",
"Guy W. Dayhoff II",
"Justin Gibbons",
"Thomas E. Keller",
"Chang Li",
"Jean Lim",
"Minh Pham",
"Anujit Sarkar",
"Ravi Sharma",
"Agaz H. Wani",
"Andrea Vianello",
"Linh M. Duong",
"Chenggi Wang",
"Celine Grace F. Atkinson",
"Madeleine Barrow",
"Nathan W. Van Bibber",
"Jan Dahrendorff",
"David A. E. Dean",
"Omkar Dokur",
"Gloria C. Ferreira",
"Mitchell Hastings",
"Gregory S. Herbert",
"Khandaker Tasnim Huq",
"Youngchul Kim",
"Xiangyun Liao",
"XiaoMing Liu",
"Fahad Mansuri",
"Lynn B. Martin",
"Elizabeth M. Miller",
"Ojas Natarajan",
"Jinyong Pang",
"Francesca Prieto",
"Peter W. Radulovic",
"Vyoma Sheth",
"Matthew Sumpter",
"Desirae Sutherland",
"Nisha Vijayakumar"
],
"abstract": "Microbiome data are undergoing exponential growth powered by rapid technological advancement. As the scope and depth of microbiome research increases, cross-disciplinary research is urgently needed for interpreting and harnessing the unprecedented data output. However, conventional research settings pose challenges to much-needed interdisciplinary research efforts due to barriers in scientific terminologies, methodology and research-culture. To breach these barriers, our University of South Florida OneHealth Codeathon was designed to be an interactive, hands-on event that solves real-world data problems. The format brought together students, postdocs, faculty, researchers, and clinicians in a uniquely cross-disciplinary, team-focused setting. Teams were formed to encourage equitable distribution of diverse domain-experts and proficient programmers, with beginners to experts on each team. To unify the intellectual framework, we set the focus on the topics of microbiome interactions at different scales from clinical to environmental sciences, leveraging local expertise in the fields of genetics, genomics, clinical data, and social and geospatial sciences. As a result, teams developed working methods and pipelines to face major challenges in current microbiome research, including data integration, experimental power calculations, geospatial mapping, and machine-learning classifiers. This broad, transdisciplinary and efficient workflow will be an example for future workshops to deliver useful data-science products.",
"keywords": [
"hackathon",
"codeathon",
"data science",
"transdisciplinary",
"gut microbiome",
"oral microbiome",
"human migration microbiome",
"Clinical Informatics",
"Bioinformatics",
"Operational Taxonomic Unit (OTU)",
"16S rRNA",
"machine learning",
"Geographic Information Systems (GIS)"
],
"content": "Introduction\n\nThe National Institutes of Health National Center for Biotechnology Information (NIH NCBI) model for codeathons—intensely collaborative, time-limited data workshops which encourage teams of participants to produce software prototypes to solve problems related to a common biomedical topic—are an effective avenue for the generation of software prototypes in the biomedical informatics space. Our previous “Iron Hack” event1, centered on rare iron-related diseases, was a transdisciplinary twist on this NCBI model designed to complement and unite local University of South Florida (USF) research programs, inspiring participation from clinicians, genetic counsellors, and researchers from a diversity of biomedical fields at all different career-stages.\n\nWe set out to further expand on the more traditional foundation of codeathons for this year’s event, working with the local research-community to select challenges that would encourage and more heavily utilize skillsets less-traditionally drawn to codeathons (e.g. social science researchers), while also supporting emerging USF research initiatives and addressing wider challenges in biomedical data science. This year’s event (dubbed the USF OneHealth Codeathon) therefore focused on the fast-evolving field of host-microbiome interactions, with concepts for our team-projects designed around data-centric problems encountered by our interdisciplinary participants in their research and practice. The event took place on USF’s Tampa campus over February 26–28, 2020.\n\nAs a result of these intense collaborative efforts, teams developed resources that are relevant not only to microbiome studies, but also general bioinformatics problems. The objective of this report is to demonstrate the utility of a codeathon model to rapidly develop tools for human and environmental health research, with the added community-building benefits of (1) providing opportunities for meaningful, long-term, cross-departmental interactions that stimulate collaborations and creative project design, and (2) offering in-depth exposure to applied data-science for members of traditionally less-computational fields.\n\nWe addressed challenges related to the host microbiome, including the great need for novel genomics tools to handle large, recently generated heterogenous microbiome datasets. We established six OneHealth Codeathon teams to develop six computational-tool prototypes broadly focused on (1) power calculation for microbiome study design, (2) geographical information systems-analysis of microbiome data and associated risk factors, (3) mining archaeological microbiome data, and (4) searching for ecological drivers of earth microbiomes (Figure 1). These team-efforts have led to the convergence of social science, ecology and medical communities with genomics data-science researchers to produce promising computational tools, strengthened through an iterative process of soliciting ideas and feedback from domain experts.\n\nTwo teams (Teams MicroPower Plus and Zero) focused on developing practical computational tools for microbiome study-design and data-analysis. Four teams (Teams Geo, Animal, Track and Yolo) focused on exploring different aspects of host-microbiome interactions from environmental consequences to clinical presentations.\n\nThe remainder of this report is organized into subsections by project, beginning with a detailed description for the six projects, the motivations behind them, and the gaps they seek to fill. We next describe the methodologies and implementations of the projects into usable software applications, how to operate the software applications, and results produced using the software applications. Finally, we discuss the pros and cons of this new highly interdisciplinary and community-driven twist on more traditional hackathons.\n\n\nTeam 1 – MicroPower Plus\n\nProject title: Microbiome power-calculation tool for biologists: towards rigorous, reproducible microbiome study-design\n\nRationale: Measured differences between sample groups can result from any number of experimental artifacts not reflective of actual biology, including differing definitions of what a clinical population signifies within different studies, how samples are prepared, and analytical decisions (e.g., bioinformatic and statistical tool-selection, parameter-settings2–4). Statistical power calculations are a key part of quality study-design, informing the sample-size required to have sufficient statistical power to detect differences between experimental groups. The size of this difference between groups—the effect size—should also be taken into account during experimental planning; smaller effects are more sensitive to being obscured by experimental noise. Sufficiently powered studies are critical for robust biological conclusions, and funding agencies increasingly require power and sample-size analyses to consider applications for support.\n\nR-based software packages enabling power analyses modeling relationships between sample-size and detectable effect-size using PERMANOVA-based methods have been developed to estimate required samples for microbiome experimental design5, given input data from pilot studies. These handy tools are not generally accessible to biologists with limited computational experience and/or a more cursory grasp of statistics. We sought to build on these methods to create a more intuitive calculator/guide for biologists, who often need only a quick point-and-click reference for experimental planning.\n\nGoal: To provide an intuitive power- and effect-size calculator-tool for biologists with limited computational experience.\n\n\nMethods\n\nPredicted effect sizes detectable at a range of sample sizes and power-levels were precomputed on OTU tables from a variety of human body-site datasets from the Human Microbiome Project (HMP) using the R package micropower (v0.4) (Jaccard distance method)5,6. We used these precomputed data as a reference for quick and interactive power calculations for commonly used sample sizes by body-site.\n\nWe added additional functionality for calculating the effect size of the experimental intervention given a control group vs. an experimental group using linear modeling. Our tool computes the Bray-Curtis distance between all samples, then uses the Adonis function from the vegan package (v2.5.6) to calculate the correlation parameter Pearson’s R7.\n\nA conceptual overview of MicroPower Plus functionality is provided in Figure 2.\n\nInput-data are OTU or ASV tables selected from curated, published microbiome studies of various human body-sites from which effect size has been pre-calculated for several common sample-sizes using complementary methodologies. The user can then use the interactive, graphical output to explore the relationships between effect-size, sample size and statistical power to use as a quick reference for their own experimental planning.\n\nThe MicroPower Plus8 workflow is implemented in a user-friendly R-Shiny web application. RStudio and the R packages shiny, plotly and tidyverse are required to operate MicroPower Plus9–12. Further documentation and a tutorial are available at the GitHub repository as listed in the code-availability section.\n\nAfter installation of required packages, all necessary tutorial files can be downloaded from GitHub onto the user’s local computer, and MicroPowerPlus can be launched by opening the “app.R” file in RStudio.\n\n\nUse cases\n\nMicroPower Plus8 is most useful as a statistical reference-guide for biologists to make quick calculations to aid in experimental design of microbiome studies. We built a user-interface around the human gut microbiome reference dataset that allows the user to visualize the relationship between sample size, effect size and statistical power as a proof of concept using R Shiny10. Resulting effect size is reported as a bar graph, with reference to effect sizes reported in the literature for comparisons. We created an additional tool that allows the user to input their own data, calculate the effect size from their experiment and report it as a bar graph. Future iterations of this tool will include interactive visualizations for the pre-computed reference data from other body-sites.\n\nThe provided tutorial walks the user through an example power calculation (Figure 3) and effect size calculation (Figure 4) using the pre-computed human gut microbiome datasets.\n\nThe user selects the sample type, the sample size for each group and a distance measure. When the user moves the power slider, the estimated effect-size graph (red) changes to the minimum effect size required to attain the given power level. The gray bars reference effect sizes calculated from the indicated sources. By comparing the estimated effect size to the reference effect sizes, the user can get a sense of how large a difference would have to be between their samples to detect significance using different experimental designs.\n\nThe user uploads a matrix of their microbiome measurements, enters the names of the groups that can be used to distinguish the sample columns by group. MicroPower Plus then calculates the effect size for the experiment (red bar). The gray bars are effect sizes calculated from the indicated literature. Comparing the red bar to the gray bars allows the user to get a sense of the magnitude of their experimental effect.\n\n\nTeam 2 – GEO\n\nProject title: Environmental Chemicals: Impact on Human Microbiomes\n\nRationale: Environmental exposures to chemicals have been a public health concern due to the ubiquitous nature of its effects on human health and the environment. Industries and manufacturing sectors contribute to chemical exposures by releasing these chemicals into the environment. Chemicals commonly found in commercial products, such as heavy metals and chlorinated hydrocarbon solvents, can persist in the environment for extended periods, increasing the latency of exposure13.\n\nA lack of information led to relatively few rules for handling and disposing of chemicals in the first part of the 20th century, which resulted in the random release of these hazardous chemicals and toxins into the environment. Knowledge of toxic waste dumps and their associated human health and environmental health consequences received national attention in the late 1970’s14. In response to public outcry, Congress created “Superfund” in the 1980’s to fund toxic waste clean-up at industrial sites14,15. Superfund sites require long-term remediation efforts, and sites are evaluated for eligibility on a point-based system requiring a preliminary assessment and site-inspection (known as the Hazard Ranking System, or HRS)16. Reporting from the public or an agency is also considered in assessing a site for the qualification. Superfund sites are prioritized by HRS score onto the National Priority List (NPL)16. Currently there are 1335 NPL sites around the U.S., each having specific chemical contaminations.\n\nHuman exposure to toxic chemicals has been shown to elicit different effects depending on the host’s immune response, with long-term exposures associating with a range of serious maladies varying from cancers acting on various bodily tissues to neurological effects17. The gut microbiome is hypothesized to have a unique role in enhancing and maintaining host health through the microbiome-gut-brain axis and can impact endocrine, immunological and nutrient signals18. Microbiome dysbiosis can occur with exposure to toxic environmental contaminants via ingestion or inhalation and can lead to several chronic conditions. Due to its diverse functions in the body, the gut microbiome acts as an indicator for health, and there is a growing body of literature exploring the interactions of environmental contaminants with the host microbiome13,17,18.\n\nEnvironmental contaminants present in Superfund sites around the U.S. can significantly affect the health of the population in the surrounding areas. To illustrate this effect, we created a tool for visualizing the impact of environmental toxicants on the gut microbiome.\n\nGoals: 1) To illustrate the trends of environmental chemical exposures from U.S. Superfund sites over time. 2) to create a tool for visualizing the impact of exposure to environmental chemicals on the gut microbiome around the U.S.\n\n\nMethods\n\nData-sources and processing: We processed and combined datasets from the American Gut Project (AGP), census data, and EPA Superfund data to search for informative patterns using the R package phyloseq 1.30.019. We identified most abundant taxa by Superfund site/geographic location. We then performed basic association analyses to assess relationships between abundant/rare taxa, various Superfund sites and contaminants. Archived code are available, see Software availability20.\n\n1) American Gut Project data: The American Gut Project (AGP) is a large-scale, crowdsourced project (n =29778) of microbial sequence data with the aim of characterizing the human gut microbiome including associated mitigating factors ranging from diet, lifestyle, overall health, and the broader environment. The metadata file obtained from AGP sample information (file 04-meta). was reduced to responses from participants within the United States only. Important variables that have been previously found to be associated with differential phenotypes mediated by air pollution in microbial communities in published studies were also selected and included in subsequent testing for associations with Superfund-site proximity.\n\n2) Superfund data: Superfund sites and associated contamination data for current NPL sites were retrieved from EPA data using the R superfundr 0.0.0.9000 package. The data were prepared and transformed using Statistical Analysis Software (SAS v 9.4, Cary, NC). We focused on 10 priority chemicals listed by the EPA.\n\n3) Census data: Select data from the American Community Survey (ACS) were downloaded from the U.S. Census Bureau: American FactFinder website via the download center (U.S. Census Bureau, 2020). This population-based data source provides descriptive socio-demographic data (e.g., sex, race, ethnicity, economic indicators, etc.) by zip code across the nation. Once all datasets were downloaded for each variable, all variables were then merged by a linking variable (i.e., zip code) that each dataset had in common. After data-cleaning, percentages were calculated for each variable. All data-cleaning was conducted using Statistical Analysis Software (SAS v 9.4).\n\nLoading and filtering OTU tables was memory-intensive, as the initial dataset is very large. Initial attempts for loading the OTU table with a 16 GB laptop were insufficient. To solve the problem, we performed this filtering on a high-performance computation cluster with 180 GB of memory.\n\nMerging data across disparate datasets: Several distinct datasets across the AGP, Superfund, and ACS provided unique information connected only by geographic location and could be merged by an appropriate linking variable (e.g., zip code). Data from all three sources were combined for a total of ~1000 samples. We further reduced the dataset to only samples that were directly related to the gut for downstream prediction using machine learning approaches.\n\nArcMap version 10.7 (2020) was used to create choropleth maps from the combined ACS and Superfund datasets to evaluate the association of chemicals found at EPA Superfund sites with select population-based socio-demographic data by zip code overtime. An open source software can be used for the same work is QGIS Geographic Information System, at Open Source Geospatial Foundation Project (http://qgis.org).\n\nMachine learning analysis on data collected from individuals near Superfund sites: We selected individuals that were self-identified to be within 5 km of Superfund sites from the final combined dataset. We next performed a classification analysis using random forests implemented via the R package ranger21. For each contaminant, we classified each individual as exposed or unexposed based on their proximity to a Superfund site with that contaminant. We then performed 10-fold cross-fold validation and reported the accuracy of the most and least informative contaminants in regard to the microbiome.\n\n\nResults\n\nGeographic distribution of select Superfund-site contaminants and abundance of Bacteriodetes OTUs are shown in Figure 5. We next explored a potential relationship between abundance of this bacterial phylum and individual contaminants, and further possible predictive efficacy of contaminants for certain OTUs, using proof-of-concept modeling. We restricted samples to those within 5 km of a Superfund site for these analyses. We constructed a random forest using each contaminant as a binary predictor-variable. We found a strong relationship between several contaminants and microbial composition. The two most predictive contaminants were polycyclic aromatic hydrocarbons and poly-chlorinated biphenyls (PAH, 94% and PCB, 81%, respectively). The contaminant with the lowest accuracy was lead (60%).\n\nGeographic distribution of select Superfund-site contaminants (circles color-coded by contaminant) and abundance of Bacteriodetes OTUs (underlying heatmap) from samples collected within 5km of a Superfund site. We found a strong relationship between several contaminants and microbial composition.\n\nIt is worth noting that PAH are known to bio-amplify as they go through food-webs. Other health outcomes linked to PAH exposure are various forms of cancer, as well as developmental impacts. PCB have been banned in the manufacturing process since 1979, yet they do not readily break down and remain a hazard over long periods of time. Because of these properties, they are commonly listed as Superfund contaminants of high concern. In conclusion, we found that for several contaminants the microbial composition varied significantly among individuals living near Superfund sites with high or low levels of PAH and PCB, respectively.\n\n\nTeam 3 - ZERO\n\nProject title: Creating a web app to study human gut microbiome variation across geographic regions of the world\n\nProject Rationales, Descriptions and Goals\n\nRationale: The human gut microbiome is one of the most densely populated sites by bacteria in the human body. It performs numerous functions, and its dysbiosis has been associated with several diseases. A major goal of microbiome researchers has been to understand the diversity of the gut microbiome across human populations. Although several studies have been undertaken for this purpose, these studies are limited in scope and comparative ability. Therefore, the rationale of the present work was to create a web tool which will be equipped with reference databases, populations and necessary scripts for the users to upload, analyze and visualize their own microbiome data at the server, with additional options to compare with the reference populations. Results can subsequently be downloaded by the user. Finally, all the reference population data is to be made available for download, along with necessary scripts to enable the user to run the program on their local computers, without the need to upload their raw data. Such a tool will be extremely useful to any interdisciplinary researchers who may have microbiome-related research questions but are not experienced in writing code, handling large microbiome datasets or who do not have access to advanced computational facilities. The codes, instructions and guidelines are available through a GitHub repository. The flowchart summarizing the approach is provided in Figure 6.\n\nUsers will be able to upload fastq files for analyses and choose reference-datasets for comparison. The in-built pipeline will then generate the Amplicon Sequence Variants (ASVs) from which the most informative for differentiating populations will be chosen using a Gaussian-Mixture EM algorithm followed by unsupervised K-means clustering. Heatmaps and PCA-plots describing the data will be generated and made available for download.\n\nGoals: 1) To download raw microbiome data (V4 region of 16S rRNA gene) from various world populations and generate amplicon sequence variant (ASV) table for comparison purposes. 2) Construct simple, but informative plots such as heatmaps and principle component analysis (PCA) plots to visualize relationships/patterns in the data through the proposed web app. 3) Provide all raw sequencing data, bash scripts and R scripts to run all steps of the analyses, as well as appropriate documentation and guidelines for an easy and error-free run of the pipeline on the user’s local computer.\n\n\nMethods\n\nWe first mined microbiome data from various world populations by geographical region. We narrowed our focus to studies on the human gut microbiome involving the V4 region of the 16S rRNA gene. A total of 1428 samples spanning populations from China, the Indian subcontinent (Himalayan region), Brazil and Europe meeting these criteria were incorporated. Raw data were downloaded from the European Nucleotide Archive (Accessions: China, PRJNA396815; Indian subcontinent, PRJEB29137; Europe, PRJNA497734; Brazil, PRJEB19103) (Table 1).\n\nDespite this initial filtration step, analysis-time was still estimated to be too high to move forward under Codeathon time-restrictions. Thus, in a second step to reduce data volume, 5000 sequences were subsampled using Seqtk 1.3-r115-dirty22 from each of the forward and reverse fastq files for each of the samples. All the downstream analyses were based on the subsampled reads. The fastq files were analyzed using the standard DADA2 1.14.1 pipeline23 to generate the distribution of ASVs observed in this dataset. The corresponding classification of each ASV was obtained using the Silva database (v132)24. The bacterial count table was further utilized for downstream analysis.\n\nThe resulting ASV table contained 1,428 samples with 2,655 bacterial taxa. Considering the very sparse data in the ASV table (only 1.231% of ASV elements exhibit reads numbers > 0), we used a Gaussian-mixture model to remove the bacteria with lower reads-coverage. A total of 1,783 taxa were removed and the remaining ASV table was normalized for each sample by the proportion of reads in each taxon using orders-of-magnitude multipliers (1-e8). The distribution of standard deviation in reads-number was calculated, and taxa at the tail-ends of the distribution were eliminated, leaving 237 taxa. Similarly, individual samples at the extreme low-end of the reads-number distribution (365 samples) were also removed using the Gaussian-mixture model. Unfortunately, all Chinese-population samples were eliminated during this step, and all downstream analyses were performed only on the populations from Europe, Brazil and the Indian subcontinent.\n\nWe used the resulting filtered dataset to perform K-means clustering to determine the optimal number of categories, finding k=18 to be most informative for the data. The Akaike information criterion (AIC) and Bayesian information criterion (BIC) were utilized to measure model robustness.\n\nWe incorporated a set of unsupervised machine learning back-end computational methods to investigate the datasets for encoded geographical information. We used python v3.6.9 along with the django web framework and conda 4.7.12 to build our workflow. The machine learning components of the workflow to identify ASVs distinguishing populations by geography are performed using TensorFlow225. Data preprocessing and data visualization are mediated through R scripts (see Implementation and Software Availability).\n\nHerein we implemented a web-based application26 for the deposition and rapid analysis of microbiome data. Importantly, users are able to (1) download a prepared database along with the server source code, or (2) construct their own database for analysis. The web-based application source code, the preprocessing and data visualization scripts, and instructions for their usage are available online as listed in the Software availability section.\n\n\nResults\n\nOur k-means parameter-exploration indicated 18 classes within the sample ASV data. The result indicates at least one or two bacterial groups are enriched for each class (Figure 7A). Classification further indicated differences in community composition by geographical location (Figure 7B). We performed a PCA to further characterize the relationship between sample categories detected via clustering. We found that the samples from classes 1, 6, 9 and 14 form clearly distinct clusters from each other (Figure 7C), further indicative of underlying geographic patterns. We identified important bacterial taxa contributing to sample classification (Figure 8) and plotted relative contribution of each ASV (classified up to genus-level) driving ordination (Figure 9). Differential relative abundance of these ASVs across all geographic populations indicated distinct geographical patterns, with several ASVs strongly associating with Indian, Brazilian, or European (to a lesser extent) populations (Figure 9). The classification of the ASVs corresponding to Figure 9 are provided in Table 2.\n\nThe color scales indicate the 18 categories used for the classification and normalized reads-number for the studied samples. (A) The heatmap indicates enrichment for at least one or two bacterial OTUs in each cluster. (B) Enrichment of K-means category by geographic location. The 18 classes showed maximum differential abundance across the three studied populations. (C) The PCA plot shows the sample affinities for the classes 1, 6, 9 and 14 which showed the greatest geographical pattern.\n\nThe X-axis shows the major ASVs, and their relative contribution to the PCA (Figure 7B) is shown on the Y-axis.\n\nThe color of the boxplot indicates geographic affiliations. The X-axis indicates the top 13 ASVs and the Y-axis shows the corresponding number of normalized reads.\n\nClassification only up to genus level were obtained since the studied region was limited to V4 of the 16S rRNA gene. When two ASVs were affiliated with the same genus, they were distinguished by adding a serial number as suffix. For example, Bacteroides_1 and Bacteroides_2 belong to the same genus.\n\n\nConclusions\n\nOur work was aimed at creating a web app to study the geographical patterns of the human microbiome and selecting features which could be useful to distinguish the populations. Using publicly available resources, we were able to include different geographical populations and select features to identify differences across groups. The resources for our study are deposited in our GitHub repository (see Software availability). Limitations of this study include that factors such as age, gender and other participant phenotypes which could be contributing to geographical patterns were not included in these analyses. However, we were able to create a web-app prototype for identifying features from the composition of the human gut microbiome related to geographical population. In the future, this work can be extended to include other variable regions of the 16S rRNA gene, as well as including other body sites such as the oral cavity, skin, etc. Similarly, batch-effect correction-tools need to be incorporated for unbiased comparison across different studies.\n\n\nTeam 4 - YOLO\n\nProject title: A web-based machine learning pipeline for disease prediction using microbial data\n\nProject Rationales, Descriptions and Goals\n\nRationale: High-throughput sequencing technologies have resulted in the generation of an increasing amount of microbial data, such as microbiome data. Using these data, machine learning methods are powerful in identifying functionally active microbes and predicting disease status. Even though machine learning algorithms are popular approaches to investigate microbiome, to adopt these methods effectively usually requires specialized training. In addition, model selection and hyper-parameter tuning can be time-consuming even for experienced practitioners. Thus, our project focused on the efficiency of AI in solving big-data problems and facilitating humans to perform other cognition-demanding tasks by developing a GUI-based pipeline for training machine learning algorithms on taxonomic microbiome data. Our pipeline expands access of computational tools to researchers in non-computational disciplines to improve cross-disciplinary study. As a proof of concept, we successfully utilized our pipeline to train a predictive algorithm for obesity rates based upon orthogonal taxonomic units which may be applied toward generating health-related features from clinical, historical, or forensic samples. Our code utilizes three methods: K-nearest neighbors (KNN), support vector machine (SVM), and adaptive boosting (AdaBoost) to achieve respective accuracies near eighty-four, ninety-one, and eighty-six percent. Both KNN and SVM utilized a 10-fold cross-validations to prevent overtraining. Under this method, training was achieved near instantaneously on a 16 GB MacBook to demonstrate feasibility. Outputs are processed into interactive graphical visualizations to improve ease-of-use. Although previous projects have utilized these computational techniques toward processing microbiome data, our pipeline removes barriers to use for researchers without coding backgrounds while streamlining efficiency for all users.\n\nStudies have revealed significant diversity in the gut microbiome composition related to various phenotypes. Obesity has been associated with changes in the microbiota at phylum-level, reduction in bacterial diversity, and different representations of bacterial genes. For example, studies of lean and obese mice suggest a strong relationship between gut microbiome and obesity. Phylogenetic marker genes uncovered by 16S rRNA gene amplicon sequencing have revolutionized the field of microbial ecology. This PCR-based method has the advantage of identifying difficult to culture bacterial organisms. Various bioinformatic pipelines can then group these sequences into clusters called OTUs. OTUs are based on their sequence similarity to each other rather than a reference taxonomic dataset which may be biased towards existing taxonomic classification27.\n\nGoals: We were interested in finding out if there is an association between gut microbiome OTUs and obesity. Additionally, we wanted to be able to use this data to distinguish between lean, overweight, and obese phenotypes in humans. We were able to successfully develop a machine-learning based pipeline that shows the association between gut microbiome OTUs and obesity with high accuracy. Furthermore, this pipeline can predict whether sample OTU data comes from a lean, overweight, or obese human phenotypes. Our work is significant because a heavy coding background is not required for use of high-accuracy machine learning tools.\n\n\nMethods\n\nTo develop our pipeline28, sample microbiome data was retrieved from 29. First, we cleaned the data by removing duplicate entries which leaves us with 151 samples. Second, to deal with the sparsity of OTU count data, we added a random small positive number to all 0 entries. Third, data was normalized using the centered log-ratio (CLR) transformation30. Then, the dimensionality reduction was performed. We chose to use the Max-min Markov Blanket method to recursively select a small subset of features that are important to the outcome of interest (Obesity or lean in this case). A total of 10 highly informative OTUs were selected during this process and various machine learning methods were explored based on a recent review article31.\n\nPrincipal component analysis (PCA) is an unsupervised dimensionality reduction technique that finds relationships in the dataset, then transforms and reduces them into principal components (i.e. uncorrelated features that embody the information contained within the dataset) that do not have redundant information.\n\nRandom forest describes a supervised machine learning strategy that splits samples into successively smaller groups based on specific features and associated threshold values. This method is in the planning phase for future versions.\n\nSVM is a method of supervised machine learning that is useful for classification, regression, and detection of outliers. SVMs are effective in higher dimensions where the dimensions are greater than the numbers of samples. Linear Support Vector Machine (SVM) classifier was used to project samples into a higher dimensional space so that they are linearly separable. Linear SVM was performed using 10-fold cross-validation with 3 repeats.\n\nKNN is a machine learning algorithm that can be used for classification and regression. In our pipeline, KNN classifier was used for the classification of disease-status, with classification determined by majority-vote of close-by data points (n = K).\n\nAdaBoost is a machine learning meta-algorithm that can be used to improve performance of other machine learning algorithms. AdaBoost classifier was used to train multiple tree classifiers (where each tree has a subset of available features) to iteratively add more weight to those misclassified samples in the next training loop. GitHub readme and description are available in the software accessibility section.\n\nWe implemented various machine learning models, namely k Nearest Neighbor, AdaBoost, and Support Vector Machines, to predict disease from the microbiome pre-processed data. It includes three main steps. 1) Users can prepare the biome OTU table to perform downstream analysis, such as PCA and machine learning. 2) In the next step, the processed data can be used to perform PCA for exploratory analysis. 3) The data is fed into machine learning models to select the highly predictive features and for the final prediction of disease-status.\n\nFeature selection and machine learning were implemented using MXM 1.4.532 and caret 6.0-85 R packages33, respectively, in R version 3.6. To make it easy for others to use this implementation, we designed a shiny application with an intuitive graphical user interface (GUI). Users can plot, visualize, and download their results generated through the app.\n\n\nResults\n\nWe show that machine learning can be used to differentiate disease from the normal states using OTU information. We used pre-processed data from a twin study with 281 samples and 5462 OTUs29. For exploratory analysis we performed PCA (Figure 10 and Figure 11; analyses and plots generated using our Shiny app) as shown in Figure 10 and Figure 11. This analysis and plots are generated using the Shiny app. We performed feature selection to select the highly significant features, shown in Table 3. Abundance of significant OTUs is shown in Figure 12. By using a set of predefined hyperparameters for each model, we achieved 10-fold cross validated accuracy of 0.936 using a linear support vector machine (Figure 13). Additionally, 10 OTUs we identified as important to obesity-status are provided in Table 3. While we do not have assigned significant functional annotations for them in the current development, future studies could use them as candidate functional groups to aid experimental design for validating clinical and public health microbiome findings.\n\nPCA plot tries to identify linear combinations of different OTUs (features) corresponding to microbiome composition discriminating by disease class. PC1 and PC2 explain only a small amount of the variance in OTUs observed across different disease classes.\n\nThe same number of OTUs and individuals are used as in Figure 10 for different classes. This PCA plot shows more separation in the OTU clusters based on ancestry than by different disease classes (shown in Figure 10).\n\nThese 10 OTUs are all bacteria which come from 2 distinct phyla. Most of the OTUs identified are at genus-level.\n\nThese OTUs are highly predictive for the classification of disease vs. normal class.\n\nThe model showed the best cross-validation score with cost=5 (accuracy = 0.936).\n\n\nTeam 5 - TRACK\n\nProject title: Tracking ancient global epidemics\n\nProject Rationales, Descriptions and Goals\n\nRationale: As the collection of human microbiome data grows, developing user-friendly tools to search proteomics databases has become critical. Bridging the gap between computer science and biological science expertise will facilitate microbiome analysis for both explanatory and predictive purposes, making significant additions to general knowledge in this field. Such effective and convenient methods of sifting through vast datasets would be well-suited to the investigation of not only modern-day microbiome samples, but also preserved historical microbial and proteomic data retrieved from ancient populations at archaeological sites worldwide. Proteomic determination of the microbes of deceased individuals would provide another dimension to forensic analysis by uncovering the pathogens that might have been responsible for their death. The significance of this determination goes beyond simply detecting the presence of bacterial peptides, also extending to tracking the migration and virulence of diseases over time in human populations.\n\nExploring ancient or paleolithic host-microbiome interactions is an emerging approach to explore widespread microbial infectious diseases, and even pandemics, by identifying pathogen-expressed proteins in human dental calculus. This approach is supplemented by data from metabolomic analyses, anthropological and paleopathological data from the skeletal material, archaeological contexts, and archival data. Examining protein content of dental calculus has typically given insight into diet and oral health of communities of past generations34–36.\n\nSince dental calculus is formed as a result of bacterial plaque accumulation around the gingiva, dental calculus consists primarily of bacteria. Thus, dental calculus lends itself well to oral microbiome analysis. For example, it was found in a medieval sample that 85–95% of the calculus was composed of bacterial proteins36. This indicates a novel method of examining the constituents of the oral microbiome and its variation across cultures, geographies, and various historical periods.\n\nThe availability of a unique set of data from the first quarantine in the world will enable substantial focus on infectious diseases and the modeling of ancient epidemics (Figure 14). All of the approximately 1500 individuals for this project died of an infectious disease, we know this from archival records. The addition of body responses to the environment and diseases (metabolites), as well as dietary data (stable isotopes to detect malnutrition), will be trialed, providing the best chance to recognize the pathogen responsible and its overall effects. In genetics and medicine, the combination of code, workflow, logic and available data will provide over 300 years of data on epidemics (especially bubonic plague) including the first influenza pandemic, dated 1580, and outbreaks of typhus and measles. It will be possible to reach ca. 600 years of data at one location using historical and medical records. The plague and other similar illnesses provoking fever are replaced by smallpox, measles and flu in later times, as medicine provides therapies, mobility increases and diet changes with many plants cultivated in different continents from where they originated. Our TRACK prototypes will enable investigations related to pathogen evolution, microbiome adaptations and human immunity responses changes.\n\nLeft: Masque porté vers 1630 par les médecins visitant les pestiférés from R. Blanchard, in Archives de parasitologie, 1900. Pl. V. Right: drawing of a doctor wearing the mask. From Thomas Bartholin, Plague doctor, Thomæ Bartholini Historiarum anatomicarum rariorum, Hafniae: Sumptibus P. Hauboldt, 1654, p. 143\n\nGoals: To achieve the transdisciplinary goals inherent to the nature of this paleo-omics project, a central database able to contain different data types is required. Towards this objective, we created and implemented a paleo-omics workflow consisting of: 1) a search engine to query the multi-data database, 2) a retrieving pipeline for paleo proteins, and 3) a query gateway for microbiome-human host interactions (Figure 15).\n\nData derived from laboratory-based analyses of biopaleological samples are processed and analyzed by established analytical software. Results from these analyses are then compared to existing databases, such as RefSeq, and both the known and unknown information are stored in a centralized Paleo-pathology database. A search engine and a web user interface (UI) then provides users access to this centralized Paleo-pathology database. The dedicated proteomics database can be expanded and rebuilt by data scientists with new data sets and novel data structures. Abbreviations: BLAST (Basic Local Alignment Tool): a popular algorithm for comparing biological sequence information, such as the amino-acid sequences of proteins or the nucleotides of DNA RNA sequences42. CIGAR (Concise Idiosyncratic Gapped Alignment Report): a string format used to represent information such as which bases align (either a match/mismatch) with the reference, are deleted from the reference, and are insertions that are not in the reference. MaxQuant: a quantitative proteomics software package designed for analyzing large mass-spectrometric data sets.\n\nWhile mass spectrometry (MS), shotgun sequencing, and 16S rRNA sequencing data can be employed in paleo-omics, we focused on an MS-based meta-proteomics approach for proof-of-concept demonstration of our prototype within the time constraints of the Codeathon, which we applied to data derived from human dental calculus protein-samples taken from archeological sites.\n\n\nMethods\n\nMS data and shotgun sequencing data obtained from ancient human dental calculus samples were used for these analyses36,37.\n\n(1) MS data: peptides were identified from raw data files by comparing spectra from the second spectrometer of a tandem-MS (MS2) to reference spectra available in protein databases. Many existing proteomics software packages, such as MaxQuant, have been designed for analyzing large MS data sets, such as the MaxQB database, and thus can perform this task38.\n\n(2) Shotgun sequencing data: the resulting short reads in FASTQ data format have been initially verified if they correspond to human DNA sequences, sequence reads were aligned to a human reference genome (Genome Reference Consortium Human Build 38) to verify human sequences using the Bowtie version 1.3.0 and BWA programs version 0.7.17 39,40. Reads not aligning to the human reference genome were characterized as non-human.\n\nAll processed data were stored in a high-performance database for future analysis. A web user interface and a search/analysis engine41 were developed to access these data.\n\nWe performed targeted pathogen searches for sequences of oral pathogenic microbes and other human pathogens, including the major human malaria parasite Plasmodium falciparum. We identified pathogenic oral microbes similar to previously published results, but no significant hits to P. falciparum from these two test-sets were identified. We additionally searched for marker oral microbiome species for other human infectious diseases as reported in detail in the results section.\n\nSource-code for our prototype is available through our GitHub repository (see Software availability section). This implementation requires the following software packages to reproduce: Python version 3.6.0; Flask version 1.1; R version 3.4.4; Perl version 5.26.1; BLAST version 2.10.0.\n\n\nResults\n\nTo test our prototype41, we searched for pathogen sequences against the two archaeological samples in the database, one from Denmark 1100-1450 AD36 and one from the United Kingdom 1770-1855 AD34. The medieval Danish samples were used with a complete set of dental pathology characterization and individual data. Consistent with the reported results36, there are oral disease pathology and bacteria normally found in the oral microbiome that can be recovered (Figure 16). For instance, the species Porphyromonas gingivalis is frequently present in individuals with orthodontic diseases, while Streptococcus sanguinis is present in both medieval and contemporary individuals with satisfactory oral health.\n\nThis approach can also be used to discover other bacteria linked to health and possibly reveal other correlations between microbiome bacteria and health status as well as recent evolutionary changes. In archaeology, the current focus is on revealing specific pathogens and there is no established reference material to investigate the past microbiome or its effects on health. Even in recent studies, any conclusions on medieval or older individuals is based on direct comparison with the contemporary microbiome. By using archaeological methods (chronological seriation) together with software developed from our code, it will be possible to investigate any correlation between microbiome and health searching individuals dating to older periods. Such work could provide a reference standard for archaeologists, and evolutionary data to health professionals. For example, using the existing data, we found the opportunistic respiratory pathogen Haemophilus parainfluenzae43,44 present less frequently in this set of medieval samples (Wilcoxen test, p < 0.05), raising interesting questions about human society transition and infectious diseases. This group appears in Neolithic agrarian human oral microbiomes (7440 BCE)45, but is at low levels in human groups practicing hunting and gathering (2000 BCE, modern day South Africa). Questions of interest to both health professionals and archaeologists that could be answered by employing our code may be when this pathogen became more frequent and why.\n\nA. A total of over 200 bacterial species have been recovered from a metaproteomics study using medieval dental calculus36. The Label- free protein quantitation (LFQ) was used to quantify all samples and conduct comparative analysis. The taxa abundance levels were normalized on a scale from 0 to 10; and the circle sizes indicate the frequency of taxa occurrences in the study B. Representative species of oral diseases (e.g. Porphyromonas gingivalis and Filifactor alocis ), oral health ( Cardiobacterium hominis and Streptococcus sanguinis), and potential respiratory disease markers (Haemophilus spp.)43,44. Modern day oral microbiomes from dental plagues and calculus are from the HMP database.\n\nUnderstanding the origins and evolution of pathogens is very important to prepare for future pandemics. The only successful work attempted on combining archaeology with genetics and health studies to investigate past pathogens, the reconstruction of the 1918 flu pathogen46 proved to be both technically challenging and costly even though fewer than a hundred years had passed since the pandemic because that work tried to reproduce an active virus now extinct. It was also very useful to demonstrate that the strong virulence reported in historical sources, but unconfirmed in medicine, was real. Since 1919, only COVID-19 has demonstrated a similar virulence, proving that data from historical record can be critical in addressing new types of known viruses and pathogens, which can regain traits unseen for a century or more within that category of pathogens (respiratory viruses with flu-like symptoms in this case). That work has shown also how the choice of suitable burial grounds is essential for such work. Our work uses new -omics analyses that are providing new sources of data and could prove equally valuable, revealing the history of recent pathogens, characteristics that may have been present only occasionally, and their successes and failures. Future pathogens might reuse and re-combine successful traits (symptoms, virulence) from past epidemics and therefore our preparedness depends on knowing what to expect, on learning from the past.\n\nThe results of our work are therefore limited to making possible future interdisciplinary research and open up a path to answer new questions. Sequencing proteomic and metabolomic data from pre-modern individuals is still rare and there is no existing database, besides data from a few academic papers, that our software code could search. Yet, making possible new studies through a working proof-of-concept will accelerate the production of databases for ancient individuals. Existing archaeological studies have borne out of early full sequencing of genomes and have been severely limited by such approaches. The benefits deriving from new -omics analyses combined with our approach can provide valuable information on older pathogens. Future work may focus on epidemics initially, but with a potential also for revealing and understanding more subtle and complex relationships between human microbiome and health.\n\n\nTeam 6 - Animal\n\nProject title: Capturing ecological and host drivers of microbiomes\n\nProject Rationales, Descriptions and Goals\n\nRationale: One primary goal of host-microbiome studies is to capture and understand ecological and host drivers of microbial diversity. Research on host-microbiome associations across host species has been facilitated by the increasing accessibility of high-throughput sequencing techniques and the availability of integrated microbiome datasets, such as the Earth Microbiome Project dataset47. These have yielded useful insights on how host-microbiome associations are impacted by host diet48, host taxonomy or phylogeny49, host immune system50, and environmental factors51. However, host species traits vary immensely across species and such diversity has been under sampled in microbiome studies. As a result, the effects of other host factors, including body mass and life history, in relation to previously characterized host and environmental effects, on host-microbiome associations have been understudied.\n\nGoal: In this project, we aim to investigate the effects of various host traits, including diet, host taxonomy, body mass, and longevity, in relation to environmental factors, on the intestine, fecal, foregut, and stomach microbiomes of Metazoan (animal) species. We first mined available microbiome and metadata datasets, then applied unsupervised learning directly on rarefied OTU abundance data to uncover clusters of microbial community similarity among animals.\n\n\nMethods\n\nRarefied OTU table (1000 reads per sample) and metadata of internal animal microbiomes from the Earth Microbiome Project47 was obtained from Woodhams et al.52. The OTU table was filtered to remove plant samples (Kingdom Plantae), OTUs with <10 total counts across samples, and OTUs occurring in <2 samples.\n\nFor each sequenced species in our dataset, we added metadata for body mass and maximum longevity, if available. Body mass data was collected from the Pantheria archives53, the Caviede Vidal dataset54, and the Encyclopedia of Life. Body mass data was categorized to create three equally sized groups (excluding Homo Sapiens): big (> 58.7 kg), medium (>19.57 kg, ≤ 58.7 kg), and small (≤ 19.57 kg). Maximum longevity data was obtained from AnAge55.\n\nTo explore distinct microbial composition structures across samples, an unsupervised cluster analysis was performed on the processed OTU table. OTUs present in less than 5% samples were discarded to obtain robust clusters. Sample-wise distance matrix was then computed using Jensen-Shannon distance on the OTU table of relative abundance. The PAM (partition around medoids) clustering analysis was completed using the cluster version 2.1.0 package in R software version 3.6.156. The optimal number of clusters was determined to maximize the Silhouette coefficient57. To visualize results of the cluster analysis, principal component analysis was completed using ade4 version 1.7-13 package in R software. Individual samples were depicted on the space of top two principal components.\n\nFor feature selection, ANOVA F-tests were implemented in python to identify quantitative metadata variables with significant means variance differences between clusters. Pearson correlation analysis was also performed in python to evaluate linear relationships between metadata variables.\n\nThe analyses can be performed on a local computer or server with R and Python installed. A step-by-step tutorial of the unsupervised clustering approach is available at https://enterotype.embl.de/enterotypes.html. R markdown and Python codes used for analyses are also available as listed in the Software availability section58.\n\n\nResults\n\nWe analyzed 726 samples spanning 199 terrestrial and freshwater Metazoan species within seven classes (Figure 17). Our unsupervised learning approach generated three sample clusters (Figure 18). The largest and most diverse cluster (cluster 1) comprised ~92% of all samples (n=667) from 21 Metazoan orders. These included lepidoptera (butterflies and moths; n=165), primates (n=85), anura (n=79), chiroptera (bats; n=44), carnivora (n=41), passeriformes (perching birds; n=37), hymenoptera (n=27), artiodactyla (n=26), diprotodontia (n=24), rodentia (n=23), lagomorpha (n=19), columbiformes (n=18), cypriniformes (n=18), squamata (n=17), anseriformes (n=9), gasterosteiformes (n=9), coleoptera (n=7), pilosa (n=7), cingulata (n=6), casuariiformes (n=5), and hemiptera (n=1). Cluster 2 comprised 34 samples from bats (n=16), butterflies and moths (n=10), perching birds (n=6), the dung beetle Teuchestes fossor (n=1), and the giant anteater Myrmecophaga tridactyla (n=1). Cluster 3 was the smallest (n=25) and exclusively comprised butterfly and moth samples belonging to seven species. These included Maculinea alcon (n=9), Durbania amakosa (n=5), Spalgis epeus (n=5), Lycaena clarki (n=2), Surendra vivarna (n=2), Anthene usamba (n=1), and Rapla iarbus (n=1).\n\nThe data clusters were generated by the Partitioning Around Medoids (PAM) clustering algorithm on Jensen-Shannon divergence calculated from OTU relative abundances. Each point on the plot represents a sample, and each cluster was labelled with its general taxonomic composition and sample sizes.\n\nANOVA analysis indicated that clusters had the most significant mean differences in microbial alpha diversity, Simpson diversity, Shannon diversity, Faith’s phylogenetic diversity, and Chao 1 diversity (Table 4). Digestive habitat type, host taxonomy/phylogeny, immune complexity, and life stage, were also significantly different between clusters, along with DNA extraction methods and environmental variables. Notably, body mass and maximum longevity were also significantly different between clusters.\n\nCluster-specific correlation analyses showed that alpha diversity in clusters 1 and 2 was consistently positively correlated with host taxonomy, immune complexity, diet, maximum longevity and latitude. Body mass, vegetation index, terrain complexity, mean temperature of the driest quarter and precipitation of the warmest and coldest quarters showed positive correlations with alpha diversity in cluster 1, but not cluster 2. Latitude and country were positively correlated with alpha diversity in cluster 2, but not cluster 1. Alpha diversity in cluster 3, which comprised butterflies and moths, was positively correlated with environmental variables (terrain complexity, mean diurnal temperature range, precipitation seasonality, elevation) and host factors (digestive habitat type and diet).\n\nThe results support our premise that host traits, including but not limited to body mass and maximum longevity, are under sampled in microbial diversity studies. Understudied host traits could also shape animal internal microbiomes together with well-characterized host traits and environmental variables. Based on our results, we propose comprehensive sampling of host traits in future microbiome studies, which may yield new and unexpected patterns of microbial community organization serving as a baseline for deeper investigations.\n\nThroughout this process we identified several areas where improvements could be made for future disease-focused hackathons. A few of these are described below.\n\nCollaboration across domains requires extensive communications with minimum use of jargons, and active learning from diverse backgrounds. We aimed to further expand on the traditional foundation of codeathons, and we generated novel tools by leveraging research strengths of the local community. However, there has been some challenges in the six teams to efficiently work together, with barriers in communicating the feasibility and significance of particular problems. In-depth and succinct explanation of the technical problems are critical for the successful operations.\n\nScalability of R has been called into question during the prototype development. For large dataset computations, more efficient implementation can be developed once the prototype has proven to be useful for the community. However, the granularity of solutions available in R make it the preferred tool for designing and experimenting with different solutions.\n\nMeticulous documentation of each analysis step remains crucial for effective dissemination of our approach and results. These necessary components of any project are also excellent opportunities to apply the skillsets of non-coders, as well as to heighten engagement of trainees by reinforcing project rationale. Good documentation, including simple flowcharts, are very useful tools for keeping focus. Non-coding participants who want to gain some experience can often quickly learn markdown language and be vital contributors to repositories.\n\n\nConclusion and next steps\n\nInterdisciplinary collaborations have proven to be very productive as shown by our six working prototypes addressing broad microbiome related challenges, ranging from power calculations, AI classifiers, GIS integration and large data set visualizations. Although working across fields has been a challenging task, we found that parsing a complex question into distinct parts can help different domain-experts to work together and accomplish tasks none of the individuals could accomplish in isolation. The codeathon workflow is thus a useful research model for many urgent societal problems that suffer from knowledge-transfer and communication issues. We have made all data and code publicly available for further exploration of these tools. Importantly, we are developing impactful projects to further pursue intersectional research spurred by this event, including microbiome-related machine learning, and data mining across archaeological time and geography.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nSoftware availability\n\nSource code available from: https://github.com/USFOneHealthCodeathon2020/Team1_MicroPowerPlus.\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.40317708.\n\nLicense: GNU General Public License 3.0.\n\nSource code available from: https://github.com/USFOneHealthCodeathon2020/Team2_GEO\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.403446620.\n\nLicense: GNU General Public License 3.0.\n\nSource code available from: https://github.com/USFOneHealthCodeathon2020/projectZer0.\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.403178026\n\nLicense: GNU General Public License 3.0.\n\nSource code available from: https://github.com/USFOneHealthCodeathon2020/Team-YOLO\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.403177628.\n\nLicense: GNU General Public License 3.0.\n\nTeam 5\n\nSource code available from: https://github.com/USFOneHealthCodeathon2020/Team5_MinhRays\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.403178541.\n\nLicense: GNU General Public License 3.0.\n\nTeam 6\n\nSource code available from: https://github.com/USFOneHealthCodeathon2020/Team6_LimSharma\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.403177858.\n\nLicense: GNU General Public License 3.0.",
"appendix": "Acknowledgements\n\nWe thank Paige Hunt for coordinating and helping organizing the event. We also thank John Adams, Derek Wildman, Monica Uddin from USF genomics program for their support. We thank Ming Ji, Derek Wildman, Christopher Collazo for forming the evaluation panel. We acknowledge the support from the USF Genomics Program, USF Omics Hub, USF Library, and Institute for the Advanced Study of Culture and the Environment for extensive logistical help in organizing the OneHealth Host-Microbiome Interactions Codeathon.\n\n\nAuthor contributions\n\n\n\n\nReferences\n\nFerreira GC, Oberstaller J, Fonseca R, et al.: Iron Hack - A symposium/hackathon focused on porphyrias, Friedreich’s ataxia, and other rare iron-related diseases [version 1; peer review: 2 approved]. F1000Res. 2019; 8: 1135. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDebelius J, Song SJ, Vazquez-Baeza Y, et al.: Tiny microbes, enormous impacts: what matters in gut microbiome studies? Genome Biol. 2016; 17(1): 217. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDebelius JW, Vázquez-Baeza Y, McDonald D, et al.: Turning Participatory Microbiome Research into Usable Data: Lessons from the American Gut Project. J Microbiol Biol Educ. 2016; 17(1): 46–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNIH Human Microbiome Portfolio Analysis Team: A review of 10 years of human microbiome research activities at the US National Institutes of Health, Fiscal Years 2007-2016. Microbiome. 2019; 7(1): 31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKelly BJ, Gross R, Bittinger K, et al.: Power and sample-size estimation for microbiome studies using pairwise distances and PERMANOVA. Bioinformatics. 2015; 31(15): 2461–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuman Microbiome Project Consortium: A framework for human microbiome research. Nature. 2012; 486(7402): 215–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOksanen J, Blanchet FG, Friendly M, et al.: vegan: Community Ecology Package. R package version 2.5-6. 2019. Reference Source\n\nOberstaller J, DokurOmkar V, Gibbons J, et al.: USFOneHealthCodeathon2020/Team1_MicroPowerPlus: v1.0.0 (Version v1.0.0). Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4031770\n\nRStudio Team: RStudio: Integrated Development Environment for R. RStudio, PBC: Boston, MA. 2020.\n\nChang W, Cheng J, Allaire JJ, et al.: shiny: Web Application Framework for R. R package version 1.5.0. 2020. Reference Source\n\nInc., P.T: Collaborative data science. 2015.\n\nWickham H, Averick M, Bryan J, et al.: Welcome to the tidyverse. J Open Source Softw. 2019; 4(43): 1686. Publisher Full Text\n\nLaue HE, Brennan KJM, Gillet V, et al.: Associations of prenatal exposure to polybrominated diphenyl ethers and polychlorinated biphenyls with long-term gut microbiome structure: a pilot study. Environ Epidemiol. 2019; 3(1): e039. PubMed Abstract | Publisher Full Text | Free Full Text\n\nU.S. Environmental Protection Agency: What is Superfund? 2018. Reference Source\n\nFriis RH: Essentials of Environmental Health. Jones & Bartlett Learning. 2019.\n\nU.S. Environmental Protection Agency: Hazard Ranking System Guidance Manual. 2020. Reference Source\n\nJin Y, Wu S, Zeng Z, et al.: Effects of environmental pollutants on gut microbiota. Environ Pollut. 2017; 222: 1–9. PubMed Abstract | Publisher Full Text\n\nCapellini FM, Vencia W, Amadori M, et al.: Characterization of MDCK cells and evaluation of their ability to respond to infectious and non-infectious stressors. Cytotechnology. 2020; 72(1): 97–109. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcMurdie PJ, Holmes S: phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One. 2013; 8(4): e61217. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCancerGenetics007, Keller T, dahrendorff A, Oberstaller J: USFOneHealthCodeathon2020/Team2_GEO: v1.0.0 (Version v1.0.0). Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4034466\n\nWright MN, Ziegler A: ranger: A Fast Implementation of Random Forests for High Dimensional Data in C++ and R. J Stat Softw. 2017; 77(1). Publisher Full Text\n\nLi H: Seqtk: a fast and lightweight tool for processing FASTA or FASTQ sequences. 2013. Reference Source\n\nCallahan BJ, McMurdie PJ, Rosen MJ, et al.: DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods. 2016; 13(7): 581–3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nQuast C, Pruesse E, Yilmaz P, et al.: The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res. 2013; 41(Database issue): D590–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAbadi M, et al.: TensorFlow: Large-scale machine learning on heterogeneous systems. 2015.\n\nAnujit-sarkar W, Oberstaller J: USFOneHealthCodeathon2020/projectZer0: v1.0.0 (Version v1.0.0). Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4031780\n\nMysara M, Vandamme P, Props R, et al.: Reconciliation between operational taxonomic units and species boundaries. FEMS Microbiol Ecol. 2017; 93(4): fix029. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAgaz NW, Bibber V, Dean A, et al.: USFOneHealthCodeathon2020/Team-YOLO: v1.0.0 (Version v1.0.0). Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4031776\n\nTurnbaugh PJ, Hamady M, Yatsunenko T, et al.: A core gut microbiome in obese and lean twins. Nature. 2009; 457(7228): 480–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRandolph TW, Zhao S, Copeland W, et al.: Kernel-Penalized Regression for Analysis of Microbiome Data. Ann Appl Stat. 2018; 12(1): 540–566. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhou YH, Gallins P: A Review and Tutorial of Machine Learning Methods for Microbiome Host Trait Prediction. Front Genet. 2019; 10: 579. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTsagris M, Tsamardinos I: Feature selection with the R package MXM [version 2; peer review: 2 approved]. F1000Res. 2018; 7: 1505. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKuhn M: Building Predictive Models in R Using the caret Package. J Stat Softw. 2008; 28(5). Publisher Full Text\n\nHendy J, Warinner C, Bouwman A, et al.: Proteomic evidence of dietary sources in ancient dental calculus. Proc Biol Sci. 2018; 285(1883). Publisher Full Text\n\nHendy J, Welker F, Demarchi B, et al.: A guide to ancient protein studies. Nat Ecol Evol. 2018; 2(5): 791–799. PubMed Abstract | Publisher Full Text\n\nJersie-Christensen RR, Lanigan LT, Lyon D, et al.: Quantitative metaproteomics of medieval dental calculus reveals individual oral health status. Nat Commun. 2018; 9(1): 4744. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVelsko IM, Yates JAF, Aron F, et al.: Microbial differences between dental plaque and historic dental calculus are related to oral biofilm maturation stage. Microbiome. 2019; 7(1): 102. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTyanova S, Temu T, Cox J: The MaxQuant computational platform for mass spectrometry-based shotgun proteomics. Nat Protoc. 2016; 11(12): 2301–2319. PubMed Abstract | Publisher Full Text\n\nLangmead B, Salzberg SL: Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012; 9(4): 357–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi H, Durbin R: Fast and accurate long-read alignment with Burrows-Wheeler transform. Bioinformatics. 2010; 26(5): 589–95. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPham M, Rays Jiang A, Nguyen D, et al.: USFOneHealthCodeathon2020/Team5_MinhRays: v1.0.0 (Version v1.0.0). Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4031785\n\nAltschul SF, Gish W, Miller W, et al.: Basic local alignment search tool. J Mol Biol. 1990; 215(3): 403–10. PubMed Abstract | Publisher Full Text\n\nHofstra JJ, Matamoros S, van de Pol MA, et al.: Changes in microbiota during experimental human Rhinovirus infection. BMC Infect Dis. 2015; 15: 336. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKosikowska U, Biernasiuk A, Rybojad P, et al.: Haemophilus parainfluenzae as a marker of the upper respiratory tract microbiota changes under the influence of preoperative prophylaxis with or without postoperative treatment in patients with lung cancer. BMC Microbiol. 2016; 16: 62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAdler CJ, Dobney K, Weyrich LS, et al.: Sequencing ancient calcified dental plaque shows changes in oral microbiota with dietary shifts of the Neolithic and Industrial revolutions. Nat Genet. 2013; 45(4): 450–5, 455e1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJordan D: The Deadliest Flu: The Complete Story of the Discovery and Reconstruction of the 1918 Pandemic Virus. 2019. Reference Source\n\nThompson LR, Sanders JG, McDonald D, et al.: A communal catalogue reveals Earth's multiscale microbial diversity. Nature. 2017; 551(7681): 457–463. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLey RE, Hamady M, Lozupone C, et al.: Evolution of mammals and their gut microbes. Science. 2008; 320(5883): 1647–51. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLim SJ, Bordenstein SR: An introduction to phylosymbiosis. Proc Biol Sci. 2020; 287(1922): 20192900. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFranzenburg S, Walter J, Künzel S, et al.: Distinct antimicrobial peptide expression determines host species-specific bacterial associations. Proc Natl Acad Sci U S A. 2013; 110(39): E3730–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKohl KD, Brun A, Magallanes M, et al.: Gut microbial ecology of lizards: insights into diversity in the wild, effects of captivity, variation across gut regions and transmission. Mol Ecol. 2017; 26(4): 1175–1189. PubMed Abstract | Publisher Full Text\n\nWoodhams DC, Bletz MC, Becker CG, et al.: Host-associated microbiomes are predicted by immune system complexity and climate. Genome Biol. 2020; 21(1): 23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCollen B, McRae L, Deinet S, et al.: Predicting how populations decline to extinction. Philos Trans R Soc Lond B Biol Sci. 2011; 366(1577): 2577–86. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCaviedes-Vidal E, McWhorter TJ, Lavin SR, et al.: The digestive adaptation of flying vertebrates: high intestinal paracellular absorption compensates for smaller guts. Proc Natl Acad Sci U S A. 2007; 104(48): 19132–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTacutu R, Thornton D, Johnson E, et al.: Human Ageing Genomic Resources: new and updated databases. Nucleic Acids Res. 2018; 46(D1): D1083–D1090. PubMed Abstract | Publisher Full Text | Free Full Text\n\nArumugam M, Raes J, Pelletier E, et al.: Enterotypes of the human gut microbiome. Nature. 2011; 473(7346): 174–80. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLovmar L, Ahlford A, Jonsson M, et al.: Silhouette scores for assessment of SNP genotype clusters. BMC Genomics. 2005; 6: 35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSwadtasnim SO, Sumpter M, Kim Y, et al.: USFOneHealthCodeathon2020/Team6_LimSharma: v1.0.0 (Version v1.0.0). Zenodo. 2020. http://www.doi.org/10.5281/zenodo.4031778"
}
|
[
{
"id": "76301",
"date": "16 Feb 2021",
"name": "Sucheta Tripathy",
"expertise": [
"Reviewer Expertise Genomics",
"Computational methods and development of pipelines for data analysis. Genome engineering."
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study describes analyzing several types of metagenomics data by several groups of participants. While the concept is great and rationale is fine, the output as a open source software is far from true. The github page links to each of the projects lead to several files often R scripts with hard coded value and can't be directly used without proper installation instruction. This is lacking for most of the projects under this study. Only in case of projectZero, there are installation instructions. However, upon trying I could not install the package. I got the following error:\nPackagesNotFoundError: The following packages are not available from current channels:\n\n- python[version='3.6.10,3.6.9.*',build=h0371630_0] PackagesNotFoundError: The following packages are not available from current channels:\n\n- python[version='3.6.10,3.6.9.*',build=h0371630_0]\nIn case of project 5: also the instructions are not very clear. For example in order to run sudo python3 ./server_setup.py It is not mentioned where this file server_setup.py is located. Whether it is coming with the distribution or this file is located somewhere is not very apparent.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? No\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1478
|
https://f1000research.com/articles/9-1465/v1
|
16 Dec 20
|
{
"type": "Research Article",
"title": "The effect of an end-effector type of robot-assisted gait training on patients with Guillain-Barre syndrome: a cross-sectional study",
"authors": [
"Seung Yeon Rhee",
"Hara Jeon",
"Seong Woo Kim",
"June Sung Lee",
"Seung Yeon Rhee",
"Seong Woo Kim",
"June Sung Lee"
],
"abstract": "Background: Guillain-Barre syndrome (GBS) is a peripheral nerve injury caused by a post-infectious immune response. Although the prognosis of GBS is relatively good, some patients have severe impairments, such as walking disabilities. Robot-assisted gait training (RAGT) is used to improve gait function in various neurologic disorders; however, no studies have reported its effectiveness in GBS patients. We aimed to evaluate the effect of gait training using an end-effector type robotic device on GBS patients. Methods: This was a retrospective study of patients diagnosed with GBS who received RAGT using Morning Walk® at an inpatient department. The main outcome measures evaluated before and after RAGT were: Medical Research Council scale, Functional Ambulation Categories, Modified Barthel Index score, Rivermead Mobility Index, and 2-minute walk test. Results: In total, 15 patients underwent RAGT 24 times. The mean age was 55.7 (±15.3) years and the average time from onset was 3.9 (±3.6) months. When compared to the baseline, all outcome measures associated with gait function were improved after RAGT. Conclusions: RAGT can improve walking ability in GBS patients. RAGT can be considered as one gait training tool to recover gait function in GBS patients.",
"keywords": [
"Guillain-Barre syndrome",
"Disability",
"Robotics",
"Gait",
"Rehabilitation"
],
"content": "Introduction\n\nGuillain-Barre syndrome (GBS), also known acute inflammatory demyelinating polyneuropathy, is a rapid-onset immune-mediated polyradiculopathy involving sensory, motor, and autonomic nerves1,2. The most common cause of GBS is post-infectious aberrant immune response that results from peripheral nerve injury. The incidence of GBS is 0.89-1.89 per 100,000 person-years and men have 1.78 times the risk of this syndrome when compared to women1,3. GBS is the most common polyradiculopathy leading to the rapid development of paralysis and sensory loss3,4. The clinical manifestations of GBS can range from mild muscle weakness to complete muscle paralysis, which may lead to severe impairment in walking ability and cause functional deficits1–3. The peak muscle weakness in GBS patients appears 2–4 weeks after the first symptoms and progressively improves over the following weeks and months1,2,5. Although most GBS patients have recovered from debilitating illness, in some patients, impairments in body functionality remain. It has been estimated that after 6 months, 20% of patients are still not able to walk3,5.\n\nThe treatment of GBS is multidisciplinary. It involves supportive care, immunomodulatory therapy using plasma exchange and intravenous immune globulin, and rehabilitation3,6–8. Rehabilitation in GBS patients is focused on the prevention and reduction of impairments in body function2. Several studies have shown that physical rehabilitation in GBS could reduce disability and improve physical abilities and quality of life9–14.\n\nRestoration of one’s walking ability is an important rehabilitative treatment goal in patients with various neurological disorders, including GBS. Therefore, it is critical to strengthen muscles and increase endurance through gait training to recover walking ability. This can be achieved using various treatments to assist with gait training, including robot-assisted gait training (RAGT). Based on the findings from various studies, RAGT has many advantages over the conventional methods including early initiation of gait training in severely dependent patients, less effort required from the physiotherapists, a longer duration and higher intensity of gait training, more physiological and reproducible gait patterns, and the possibility to measure a patient’s performance15. Additionally, RAGT has potential aerobic benefits with a positive influence on cardiopulmonary fitness, as it was shown in severely disabled spinal cord injury and stroke patients16. The feasibility and safety of Morning Walk®, a RAGT device, for patients with various neurologic disorders and the effect of Morning Walk®-assisted gait training on patients with stroke was proven in previous studies17,18. Although RAGT has been shown to be effective in improving gait function in patients with stroke and spinal cord injuries, the effectiveness of RAGT in GBS is not well documented. Therefore, the purpose of this study was to evaluate the effectiveness of RAGT in GBS patients using an end effector type of robotic device.\n\n\nMethods\n\nWe retrospectively analysed patients with GBS who were hospitalized at the National Health Insurance Service Ilsan Hospital from April 2016 to January 2020. Subjects were included if this was their first diagnosis of GBS and were 19 years of age or older. All included patients received RAGT using Morning Walk®. Morning Walk® is an end effector type of robotic device with body support provided via a saddle seat; it was developed in South Korea (Figure 1). The footplates operate independently in the sagittal plane to simulate locomotor activity and guide the feet to reproduce gait trajectories. It also offers ground walking as well as ascending and descending stairs modes.\n\nSubjects received Morning Walk®-assisted gait training for a total of 24 sessions; each session lasting 30 minutes. All participants were assessed using the following tests before and after RAGT: the Medical Research Council (MRC) scale for evaluating muscle strength; the Functional Ambulation Categories (FAC) for measuring functional gait; the Modified Barthel Index Score (MBI) for measuring activities of daily living; the Rivermead Mobility Index (RMI) for testing functional abilities; and the 2-minute walk test (2MWT) for measuring endurance of walking distance18. Information was collected from patients medical records, and improvement was measured by calculating differences in the scores before and after RAGT.\n\nSPSS statistics version 25.0 (SPSS Inc., Chicago, IL, USA) was used for statistical analyses. The changes before and after Morning Walk®-assisted gait training for all of the investigated parameters were assessed using paired t-tests. P-values <0.05 were considered statistically significant.\n\nThis study was approved by the Institutional Review Board of the Ilsan Hospital (NHIMC 2020-02-008-001). Prior written consent from patients was waived by the IRB because this is a retrospective study.\n\n\nResults\n\nSixteen patients diagnosed with GBS underwent RAGT using Morning Walk®. Among them, one participant dropped out of the trial due to pain and discomfort around the saddle seat during the RAGT. Thus, 15 patients (11 males and 4 females) were included; the mean age was 55.7 (±15.3) years and the average time from onset was 3.9 (±3.6) months.\n\nCompared to the baseline measurements, all the outcome measures were improved after Morning Walk®-assisted gait training. There were significant improvements in muscle power of the hip, knee, and ankle, FAC, MBI, 2MWT, and RMI (Table 1).\n\nFlx.: Flexor, Ext.: Extensor,. *p<0.05\n\n\nDiscussion\n\nThis study demonstrated that RAGT was beneficial and effective in patients with GBS. The patients with GBS who received Morning Walk®-assisted gait training showed significant improvements in the motor power of their lower limbs, gait function, gait endurance, and activities of daily living.\n\nGBS is associated with residual physical disability such as stroke, spinal cord injury, or traumatic brain injury. The effectiveness of rehabilitation treatment in patients with brain lesions or spinal cord injuries has been discussed in several studies19–21. However, the effectiveness of rehabilitative treatment in patients with GBS is still poor22. Some studies concluded that rehabilitation treatment for GBS patients was effective and improved body functionality and quality of life2,9,12. However, these studies do not strongly support the effect of rehabilitation in patients with GBS due to their limitations, including small sample sizes or the lack of a control group.\n\nRAGT was proven to be a significant method to improve the locomotor function of patients with various neurologic disorders18,23,24. However, to our knowledge, the efficacy of RAGT in GBS patients has not been reported to date. This was the first preliminary study to investigate the effects of RAGT among patients with GBS. We found that Morning Walk®-assisted gait training improved the MRC scale. These findings suggest that RAGT might assist in strengthening the muscle power in the lower limbs. There were also improvements in the FAC and 2MWT after RAGT, suggesting that RAGT is beneficial in improving functional gait and walking endurance. We believe that the improvements in the lower limb muscle strength were related to the improvements in functional gait and walking endurance. Finally, the MBI and RMI scores also improved after RAGT, suggesting that RAGT improves activities of daily living and functional abilities. RAGT using an end-effector type device improved walking and functional abilities in GBS patients and it can be considered as one of the gait training tools to assist in the recovery of gait function in patients with GBS.\n\nThis study had several limitations. Firstly, the sample size was small; only 15 patients from a single medical center were enrolled. Secondly, there was no control group in this study; thus, we were not able to determine if RAGT is better than conventional rehabilitative therapy. Finally, this study only assessed outcomes at the beginning and end of RAGT. Thus, we were not able to determine the persistence of treatment effects over time. Future studies with larger sample sizes and a control group are needed to evaluate the persistence of treatment effects.\n\n\nData availability\n\nDryad: The Effect of an End-Effector Type of Robot-Assisted Gait Training on Patients with Guillain-Barre Syndrome, https://doi.org/10.5061/dryad.hqbzkh1df25.\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).",
"appendix": "References\n\nvan den Berg B, Walgaard C, Drenthen J, et al.: Guillain-Barré syndrome: pathogenesis, diagnosis, treatment and prognosis. Nat Rev Neurol. 2014; 10(8): 469–82. PubMed Abstract | Publisher Full Text\n\nNovak P, Šmid S, Vidmar G: Rehabilitation of Guillain-Barré syndrome patients: an observational study. Int J Rehabil Res. 2017; 40(2): 158–163. PubMed Abstract | Publisher Full Text\n\nYuki N, Hartung HP: Guillain–Barré syndrome. N Engl J Med. 2012; 366(24): 2294–2304. PubMed Abstract | Publisher Full Text\n\nGoodman CC, Fuller KS: Pathology for the Physical Therapist Assistant-E-Book. Elsevier Health Sciences; 2016. Reference Source\n\nVucic S, Kiernan MC, Cornblath DR: Guillain-Barré syndrome: an update. J Clin Neurosci. 2009; 16(6): 733–741. PubMed Abstract | Publisher Full Text\n\nChevret S, Hughes RA, Annane D: Plasma exchange for Guillain-Barré syndrome. Cochrane Database Syst Rev. 2017; 2017(2): CD001798. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHughes RA, Swan AV, Raphael JC, et al.: Immunotherapy for Guillain-Barré syndrome: a systematic review. Brain. 2007; 130(Pt 9): 2245–57. PubMed Abstract | Publisher Full Text\n\nHughes RA, Swan AV, van Doorn PA: Intravenous immunoglobulin for Guillain-Barré syndrome. Cochrane Database Syst Rev. 2014; 2014(9): CD002063. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlexandrescu R, Siegert RJ, Turner-Stokes L: Functional outcomes and efficiency of rehabilitation in a national cohort of patients with Guillain-Barré syndrome and other inflammatory polyneuropathies. PLoS One. 2014; 9(11): e110532. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPrasad R, Hellawell DJ, Pentland B: Usefulness of the Functional Independence Measure (FIM), its subscales and individual items as outcome measures in Guillain Barré syndrome. Int J Rehabil Res. 2001; 24(1): 59–64. PubMed Abstract | Publisher Full Text\n\nNovak P, Vidmar G, Kuret Z, et al.: Rehabilitation of critical illness polyneuropathy and myopathy patients: an observational study. Int J Rehabil Res. 2011; 34(4): 336–42. PubMed Abstract | Publisher Full Text\n\nKhan F, Pallant JF, Amatya B, et al.: Outcomes of high- and low-intensity rehabilitation programme for persons in chronic phase after Guillain-Barré syndrome: a randomized controlled trial. J Rehabil Med. 2011; 43(7): 638–46. PubMed Abstract | Publisher Full Text\n\nPrada V, Massa F, Salerno A, et al.: Importance of intensive and prolonged rehabilitative treatment on the Guillain-Barrè syndrome long-term outcome: a retrospective study. Neurol Sci. 2020; 41(2): 321–327. PubMed Abstract | Publisher Full Text\n\nKhan F, Amatya B: Rehabilitation interventions in patients with acute demyelinating inflammatory polyneuropathy: a systematic review. Eur J Phys Rehabil Med. 2012; 48(3): 507–22. PubMed Abstract\n\nMazzoleni S, Focacci A, Franceschini M, et al.: Robot-assisted end-effector-based gait training in chronic stroke patients: A multicentric uncontrolled observational retrospective clinical study. NeuroRehabilitation. 2017; 40(4): 483–492. PubMed Abstract | Publisher Full Text\n\nBillinger SA, Arena R, Bernhardt J, et al.: Physical activity and exercise recommendations for stroke survivors: a statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2014; 45(8): 2532–53. PubMed Abstract | Publisher Full Text\n\nChoi S, Kim SW, Jeon HR, et al.: Feasibility of Robot-Assisted Gait Training with an End-Effector Type Device for Various Neurologic Disorders. Brain Neurorehabil. 2020; 13(1): e6. Publisher Full Text\n\nKim J, Kim DY, Chun MH, et al.: Effects of robot-(Morning Walk(®)) assisted gait training for patients after stroke: a randomized controlled trial. Clin Rehabil. 2019; 33(3): 516–523. PubMed Abstract | Publisher Full Text\n\nTurner-Stokes L, Disler PB, Nair A, et al.: Multi-disciplinary rehabilitation for acquired brain injury in adults of working age. Cochrane Database Syst Rev. 2005; (3): CD004170. PubMed Abstract | Publisher Full Text\n\nDohle C, Tholen R, Wittenberg H, et al.: [Evidence-based rehabilitation of mobility after stroke]. Nervenarzt. 2016; 87(10): 1062–1067. PubMed Abstract | Publisher Full Text\n\nNielsen JB, Willerslev-Olsen M, Christiansen L, et al.: Science-based neurorehabilitation: recommendations for neurorehabilitation from basic science. J Mot Behav. 2015; 47(1): 7–17. PubMed Abstract | Publisher Full Text\n\nKhan F, Ng L, Amatya B, et al.: Multidisciplinary care for Guillain-Barré syndrome. Cochrane Database Syst Rev. 2010; 6(10): Cd008505. PubMed Abstract | Publisher Full Text\n\nCapecci M, Pournajaf S, Galafate D, et al.: Clinical effects of robot-assisted gait training and treadmill training for Parkinson's disease. A randomized controlled trial. Ann Phys Rehabil Med. 2019; 62(5): 303–312. PubMed Abstract | Publisher Full Text\n\nNam KY, Kim HJ, Kwon BS, et al.: Robot-assisted gait training (Lokomat) improves walking function and activity in people with spinal cord injury: a systematic review. J Neuroeng Rehabil. 2017; 14(1): 24. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJeon H, Rhee SY, Kim SW, et al.: The effect of an end-effector type of robot-assisted gait training on patients with Guillain-Barré syndrome. Dryad. Dataset, 2020. http://www.doi.org/10.5061/dryad.hqbzkh1df"
}
|
[
{
"id": "76202",
"date": "12 Jan 2021",
"name": "Jong Moon Kim",
"expertise": [
"Reviewer Expertise Description of a specific method using RAGT",
"Information on the onset of inclusion patients",
"Wilcoxon signed rank test analysis"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI think this paper is a meaningful study with clear meaning, despite the limitations of no control group.\nHowever, as mentioned in the paper, GBM often improves to spontaneus, and GBM's onset is important in determining this. It would be nice to provide information on the onset of inclusion patients.\nIt would be nice if a description of a specific method using RAGT was added. For example, it would be nice if information such as 1) the patient proceeds with treatment to walk as quickly as possible; 2) uses the stair function; and 3) how much body weight support is used.\nFor statistical analysis 15 patients are non-parametric, so Wilcoxon signed rank test would be more suitable than paired t-test.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "84261",
"date": "16 Jun 2021",
"name": "Carlos A Cifuentes",
"expertise": [
"Reviewer Expertise Healthcare Robotics",
"Rehabilitation Robotics",
"Human-Robot Interaction"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper presents a study of rehabilitation robotics with patients with Guillain-Barre syndrome. The topic is fascinating because of the reduced evidence of the use of robotic therapy in this population. The authors recruited 15 patients who were intervened in gait therapy with the robotic device Morning Walk during 24 sessions of 30 minutes. I have some concerns about the contribution of this work as follows:\n\nIntroduction: The authors should provide a proper state of art analysis with previous rehabilitation results with this population defining the experimental condition. Additionally, it is required to reference studies of rehabilitation outcomes with this population or other affected populations by other pathologies with Morning Walk Device.\nMethodology: There are many missing experimental details, such as the location of the hospital, the preparation and the experimental set-up, the description of the population and the clinical levels of every involved patient, the measurements, and the procedure of the clinical evaluation.\nResults: There are also many missing information details, such as the units in Table 1, details of the results. According to the patient or the level of affectation, it is also required to analyze the kinematics results.\nDiscussion: This section is also fragile. The authors should compare with other works about the results following similar intervention with the same or similar set-up even though with other population to have a reference of comparison.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1465
|
https://f1000research.com/articles/9-603/v1
|
12 Jun 20
|
{
"type": "Case Report",
"title": "Case Report: Odontogenic carcinosarcoma arising from ameloblastic fibrodentinoma with immunohistochemical profile",
"authors": [
"Hend Mohammed Waguih Mahmoud Salem",
"Sarah Ahmed Mohamed Mahmoud Badawy",
"Hatem Wael Amer",
"Hend Mohammed Waguih Mahmoud Salem",
"Hatem Wael Amer"
],
"abstract": "Odontogenic carcinosarcoma is an extremely rare mixed malignant odontogenic neoplasm in which both the epithelial and the ectomesenchymal components are cytologically malignant. Owing to the scarcity of the reported cases, the clinical behavior of odontogenic carcinosarcoma remains unexplored. The present work reports a case of odontogenic carcinosarcoma of the mandible with the detailed immuno-histochemical profile. A 28-year-old male presented with a left painless swelling, which recurred after hemimandiblectomy of the lesion. Complete excision of the lesion with a wide safety margin was performed. The histopathological examination revealed cytological malignant features in the follicles as well as in the ectomesenchyme. The immuno-histochemistry demonstrated evidence of epithelial mesenchymal transition. After nine months, the patient demonstrated recurrence in the infra-temporal fossa. Odontogenic carcinosarcoma is a rare highly aggressive malignant odontogenic tumor. Positive expression of smooth muscle actin and vimentin in the epithelial component is a useful aid in the diagnosis of odontogenic carcinosarcoma.",
"keywords": [
"Odontogenic carcinosarcoma",
"α-smooth muscle actin",
"vimentin",
"epithelial mesenchymal transition",
"Ki-67."
],
"content": "Introduction\n\nMalignant odontogenic tumors are rare group of malignant neoplasms that arise from odontogenic remnants1. One of these neoplasms is odontogenic carcinosarcoma (OCS), an extremely rare mixed malignant odontogenic neoplasm in which both the epithelial and the ectomesenchymal components are cytologically malignant2. In the 1992 World Health Organization (WHO) classification of tumors, OCS was included in the malignant odontogenic tumors after Tanaka and co-workers (1991) first reported an odontogenic tumor with a mixture of malignant epithelial and ectomesenchymal components3. In the 2005 WHO classification, the tumor was removed due to an absence of current diagnostic criteria4. OCS has been added again in the 2017 edition because of the availability of cases with adequate diagnostic immunohistochemical and molecular criteria5,6. Owing to the scarcity of reported cases - only eleven reported cases in the English literature - OCS clinical behavior remains unexplored7,8. In the current work, we report a case of OCS with the detailed clinical, radiographic, histopathological and immunohistochemical description.\n\n\nCase report\n\nA 28-year-old Egyptian male patient working as an accountant was referred to the Department of Oral and Maxillofacial Surgery, Faculty of Dentistry, Cairo University, with a complaint of painless swelling of six months duration in the left side of his face measuring 7 × 6 cm. The patient had a pathological report of an incisional biopsy, performed outside our institute, which was diagnosed as ameloblastoma. There were no palpable lymph nodes. Intra-oral examination revealed an absence of the lower left molars with a fistula opening on the alveolar crest with no oozing pus.\n\nA cone beam computed tomography examination revealed an ill-defined multilocular osteolytic lesion with fine radiopacities extending from the lower left second premolar up to the ramus (Figure 1). A hemi-mandibulectomy was performed based on the incisional biopsy diagnosis and the pathological fracture of the mandible. The ramus was totally destructed and the tumor was invading the masseter muscle (Figure 2). Differential diagnosis was made based on the clinical examination, radiographic appearance and gross examination as: ghost cell odontogenic carcinoma, ameloblastic carcinoma, ameloblastic fibrosarcoma, ameloblastic carcinosarcoma, calcifying epithelial odontogenic tumor, atypical type of ameloblastic fibro-dentinoma9 and ameloblastoma.\n\nThere was scattered radiopacities closely related to the impacted first molar. The second molar was displaced against the angle of the mandible.\n\nGross examination: (A) Hemi-mandibulectomy with safety margin including the masseter muscle. (B) Cross-section of the resected specimen showing the fleshy neoplasm and the impacted first molar. (C) Top view of the resected specimen showing the necrotic fistula on the alveolar ridge. (D) The excisional specimen of the recurred lesion.\n\nHistopathological examination revealed follicles of odontogenic epithelium that were compressed by highly cellular hyalinized primitive connective tissue resembling dental papilla. The epithelial cells showed little pleomorphism and hyperchromatism. In some fields, dentinoid matrix was detected around the epithelial strands (Figure 3). Based on the histopathological examination, a diagnosis was made of ameloblastic fibro-dentinoma (atypical type).\n\n(A) Microscopic image showing odontogenic epithelial follicles formed of ameloblast-like cells on the periphery and stellate-reticulum like cells on the center, which were surrounded by primitive ectomesenchyme resembling dental papilla. Dentinoid material was detected in contact with the epithelial follicles. (B) Microscopic image showing another field which was hypercellular in both components with mild pleomorphism and hyperchromatism.\n\nAfter nine months, the patient returned with another large painless swelling in the left temporal and infra-temporal fossa that had appeared three months previously. An oral and maxillofacial surgeon used local anesthesia and an intra-oral approach to completely excise the recurrent lesion (Figure 2D). Augmentin 1gm tablet/12 hours for five days and anti-inflammatory medication were prescribed for the patient after the surgery. Histopathological examination revealed highly cellular neoplasm, which showed cellular atypia and increased mitosis in both epithelial and ectomesenchymal components (Figure 4).\n\n(A) Microscopic image showing hypercellular epithelial follicles along with hypercellular ectomesenchyme (X100). (B) Microscopic image showing pleomorphism, hyperchromatism, increased mitosis and increased nuclear-cytoplasmic ratio (X400). (C) Microscopic image showing atypia in ectomesenchymal component with increased mitosis and abnormal mitotic figures (X400).\n\nThe epithelial nests showed strong membranous positivity with AE1/AE3, while vimentin stained the ectomesenchyme diffusely and the epithelial nests patchily. Alpha-smooth muscle actin (α-SMA) showed positivity in the endothelial cells as well as in scattered epithelial and ectomesenchymal cells. Ki-67 index was around 40% in the epithelial component, while in the ectomesenchymal component it was around 12% (Figure 5). Based on the histopathological and immunohistochemical findings, the case was diagnosed as OCS arising from ameloblastic fibrodentinoma. The patient was missed for the follow-up appointment after the last excision of the lesion and could not be reached for further follow-up.\n\n(A) Microscopic image showing strong membranous expression of AE1/AE3 in the epithelial component. (B) Microscopic image showing strong diffuse cytoplasmic expression of vimentin in ectomesenchymal component along with patchy expression in the epithelial component. (C) Microscopic image showing cytoplasmic expression of α-SMA in the endothelial cells and scattered epithelial and ectomesenchymal cells. (D) Microscopic image showing high nuclear expression of Ki-67 in both epithelial and ectomesenchymal components.\n\n\nDiscussion\n\nOCS is a rare biphasic malignant odontogenic neoplasm that has the same architecture of ameloblastic fibroma, in which the epithelial and the ectomesenchymal components are cytologically malignant. It can develop de novo from odontogenic remnants or as a transformation from a preexisting odontogenic benign or malignant neoplasm6. This transformation may be attributed to multiple surgical procedures or recurrences of the neoplasm2. The present case arose from a preexisting ameloblastic fibrodentinoma. To the best of our knowledge, this is the first reported case of OCS to arise from ameloblastic fibrodentinoma. Kunkel et al. (2004), DeLair et al. (2007) and Chikosi et al. (2011) reported cases aroused from ameloblastic fibrosarcoma, ameloblastic fibroma and ameloblastoma, respectively2,10,11.\n\nOur case was a 28-year-old male with the lesion affecting the posterior mandible. In the English literature, there was only one case that occurred in maxilla4. The most common radiographic picture of OCS is ill-defined multilocular radiolucency with cortical perforation6. The current case showed multiple radiopacities owing to the preexisting ameloblastic fibrodentinoma.\n\nOCS must be distinguished from ameloblastic fibrosarcoma, in which the ectomesenchymal component only shows cellular atypia, and spindle cell variant of ameloblastic carcinoma, which lacks the ectomesenchymal component2–8. In the present case, there was a patchy positive expression of vimentin in the epithelial component. This could be explained as OCS undergoes epithelial mesenchymal transition, a process in which the polarized immotile epithelial cell changes to gain the mesenchymal phenotype and indicates more aggressive behavior of the neoplasm7,12,13.\n\nIn accordance with Dos Santos et al. (2018), α-SMA staining in the current case showed scattered cytoplasmic staining in both the ectomesenchymal and epithelial cells4. Its expression in the ectomesenchymal cells could be attributed to the emergence of cancer associated myofibroblasts, which play a significant role in tumor progression and the epithelial mesenchymal transition process, explaining α-SMA expression in the epithelial cells14.\n\nThe proliferative index Ki-67 in our case was around 40% in the epithelial component and 12% in the ectomesenchymal component. This is in accordance with the work of Dos Santos et al. (2018) and Soares et al. (2019), who found that Ki-67 index was higher in the epithelial component than the ectomesenchymal component4–7.\n\nThe principle line of treatment, as with other malignant odontogenic neoplasms, is surgical resection with a wide safety margin along with neck dissection. However, adjunctive radiotherapy is still a matter of question, and it may be helpful in cases with soft tissue invasion6.\n\n\nConclusion\n\nOCS is an extremely rare odontogenic malignant neoplasm that shows very aggressive clinical behavior with multiple recurrences and possible metastasis. Immunohistochemical staining with vimentin, α-SMA and Ki-67 is helpful in the diagnosis of OCS.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nConsent\n\nWritten informed consent for publication of their clinical details or clinical images was obtained from the patient.",
"appendix": "Acknowledgments\n\nThis research was funded by the Deanship of Scientific Research at Princess Nourah bint Abdulrahman University through the fast track Reaseach Funding Program. Moreover, we are sincerely grateful for our colleges in Oral & Maxillofacial Surgery department, Faculty of Dentistry, Cairo University.\n\n\nReferences\n\nEl-Naggar AK, Chan JKC, Grandis JR, et al.: WHO Classification of Head and Neck Tumours. 4th edition, Lyon: IARC Press; 2017; 9: 213. Reference Source\n\nDeLair D, Bejarano PA, Peleg M, et al.: Ameloblastic carcinosarcoma of the mandible arising in ameloblastic fibroma: a case report and review of the literature. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2007; 103(4): 516–520. PubMed Abstract | Publisher Full Text\n\nTanaka T, Ohkubo T, Fujitsuka H, et al.: Malignant mixed tumor (malignant ameloblastoma and fibrosarcoma) of the maxilla. Arch Pathol Lab Med. 1991; 115(1): 84–87. PubMed Abstract\n\nDos Santos JN, Servato JPS, Cardoso SV, et al.: Odontogenic carcinosarcoma: morphologic and immunohistochemical description of a case. Oral Surg Oral Med Oral Pathol Oral Radiol. 2018; 126(5): e264–e270. PubMed Abstract | Publisher Full Text\n\nSoluk-Tekkeşin M, Wright JM: The World Health Organization Classification of Odontogenic Lesions: A Summary of the Changes of the 2017 (4th) Edition. Turk Patoloji Derg. 2018; 34(1). PubMed Abstract | Publisher Full Text\n\nSchuch LF, de Arruda JAA, Silva LVO, et al.: Odontogenic carcinosarcoma: A systematic review. Oral Oncol. 2018; 85: 52–59. PubMed Abstract | Publisher Full Text\n\nSoares CD, Delgado-Azañero W, Morais TMDL, et al.: Odontogenic Carcinosarcoma: Clinicopathologic Features of 2 Cases. Int J Surg Pathol. 2020; 28(4): 421–426. PubMed Abstract | Publisher Full Text\n\nKim IK, Pae SP, Cho HY, et al.: Odontogenic carcinosarcoma of the mandible: a case report and review. J Korean Assoc Oral Maxillofac Surg. 2015; 41(3): 139–144. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGiraddi GB, Garg V: Aggressive atypical ameloblastic fibrodentinoma: Report of a case. Contemp Clin Dent. 2012; 3(1): 97–102. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKunkel M, Ghalibafian M, Radner H, et al.: Ameloblastic fibrosarcoma or odontogenic carcinosarcoma: a matter of classification? Oral Oncol. 2004; 40(4): 444–449. PubMed Abstract | Publisher Full Text\n\nChikosi R, Segall N, Augusto P, et al.: Odontogenic carcinosarcoma: case report and literature review. J Oral Maxillofac Surg. 2011; 69(5): 1501–1507. PubMed Abstract | Publisher Full Text\n\nMcLean-Holden AC, Bishop JA, Kessler HP, et al.: Spindle-cell variant of ameloblastic carcinoma: a report of 3 cases and demonstration of epithelial-mesenchymal transition in tumor progression. Oral Surg Oral Med Oral Pathol Oral Radiol. 2019; 128(3): e113–e121. PubMed Abstract | Publisher Full Text\n\nBello IO, Alanen K, Slootweg PJ, et al.: Alpha-smooth muscle actin within epithelial islands is predictive of ameloblastic carcinoma. Oral Oncol. 2009; 45(9): 760–5. PubMed Abstract | Publisher Full Text\n\nMahmoud SAM, Amer HW, Mohamed SI: Primary ameloblastic carcinoma: literature review with case series. Pol J Pathol. 2018; 69(3): 243–253. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "65932",
"date": "18 Aug 2020",
"name": "Oslei Paes de Almeida",
"expertise": [
"Reviewer Expertise Oral Pathology."
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI read this article and below are my comments for the journal:\nBased on documentations included in the article some points should be reconsidered by the authors.\n\nDiagnosis of ameloblastic fibrodentinoma does not seem to be correct, dentinoid material is not evident.\n\nPossibly from the beginning the lesion is an ameloblastic fibrosarcoma, therefore the microscopical aspects should be carefully reviewed including a Ki-67 staining.\n\nDiagnosis of carcinosarcoma should be reconsidered and most immune reactions repeated and reanalysed, including Ki-67. Interpretation of some reactions does not seem to be appropriate.\n\nIs the background of the case’s history and progression described in sufficient detail? No\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? No\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? No\n\nIs the case presented with sufficient detail to be useful for other practitioners? No",
"responses": [
{
"c_id": "5922",
"date": "17 Sep 2020",
"name": "Sarah Ahmed Mohamed Mahmoud Badawy",
"role": "Author Response",
"response": "First of all, thank you for your review. I hope that in my response, I will be able to clarify some points: The diagnosis of the first biopsy was based on the presence of dentinoid material which is evident in figure 3.A. Moreover, both the ectomesenchymal and the epithelial compartments in the first biopsy showed little cellular activity without cellular atypia in all fields, there was no way to consider it as a malignancy back then without histopathological evidence; even if the clinical behavior was doubted. The immunohistochemical reaction of Ki67 was interpreted with 3 pathologists last, I wished you can mention clearly what the details do you need us to clarify in the sections of case history, physical examination, investigations, and discussion? Thank you for your time."
}
]
},
{
"id": "69312",
"date": "30 Sep 2020",
"name": "João Paulo Silva Servato",
"expertise": [
"Reviewer Expertise Oral and maxillofacial pathology",
"Oral medicine",
"Immunology."
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\n1) Is the background of the case’s history and progression described in sufficient detail?\nPartly - In reason to understand OCS progression and outcomes, more follow up details should be addressed.\nMoreover, the authors should attend to the new WHO classification nomenclature. As stated by Speight PM, and Takata T., 2018: \"the ameloblastic fibrodentinoma and ameloblastic fibro-odontoma have been removed as distinct entities\"1. In this way, the nomenclature should be reviewed.\n2) Are enough details provided of any physical examination and diagnostic tests, treatment given and outcomes?\nNo - Please, improve the details in histopathological and immunohistochemical profile.\n2.1) High power field images is needed to improve the recognition of the main expected features, as described by El-Mofty, S. K., in the 2017 WHO blue book \"The cells in the sarcomatous component are markedly pleomorphic, with enlarged and bizarre nuclei and occasional multinucleation and mitosis. The epithelial component is frankly malignant, with large hyperchromatic nuclei and an increased N:C ratio. The typical ameloblastic features such as peripheral nuclear palisading and inner stellate reticulum may be lost focally.\"2\nIn the same way, Ki-67 images must be improved to clearly demonstrate the high levels of Ki-67 tagging in both epithelial and mesenchymal components.\n3) Is sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment?\n\nPartly, the discussion section should be expanded to access the relevance of this case in the understand of OCS main epidemiological characteristics, histopathological features, the biological profile and its prognosis/ behavior. Moreover, as describe in CARE guidelines, this section must describe:\na) Strengths and limitations in your approach to this case. b) The rationale for your conclusions. c) The primary “take-away” lessons from this case report.\n\n4) Is the case presented with sufficient detail to be useful for other practitioners?\nNo, since it do not attended to most recent WHO classification and to CARE guidelines.\n\nIs the background of the case’s history and progression described in sufficient detail? Partly\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? No\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? No",
"responses": [
{
"c_id": "6201",
"date": "15 Dec 2020",
"name": "Sarah Ahmed Mohamed Mahmoud Badawy",
"role": "Author Response",
"response": "First, we can't thank you enough for your sincere and detailed review. We made a second version of the article with the revisions you recommended in your review; as follow: We did contact the surgeon to collect more details about the follow-up of the patient. We changed the nomenclature of '' ameloblastic fibrodentinoma'' as you recommended, and clarified the reason why we diagnosed it like that. We changed and added histopathological figures as you recommended with more detailed interpretation. We repeated the ki67 staining and analysed it with the image analyser Leica QWin software. We tried to improve the discussion part as you recommended to follow CARE guidelines. We hope that our revisions meet your expectations. Again, thank you for your precious time."
}
]
}
] | 1
|
https://f1000research.com/articles/9-603
|
https://f1000research.com/articles/9-1462/v1
|
15 Dec 20
|
{
"type": "Review",
"title": "Cardiovascular protection effect of chlorogenic acid: focus on the molecular mechanism",
"authors": [
"Mifetika Lukitasari",
"Mohammad Saifur Rohman",
"Dwi Adi Nugroho",
"Nashi Widodo",
"Nur Ida Panca Nugrahini",
"Mifetika Lukitasari",
"Dwi Adi Nugroho",
"Nashi Widodo",
"Nur Ida Panca Nugrahini"
],
"abstract": "Vascular endothelial cells have a variety of functions such as the control of blood coagulation, vascular permeability, and tone regulation, as well as quiesce of immune cells. Endothelial dysfunction is a cardiovascular events predictor, which is considered the initial stage in atherosclerosis development. It is characterized by alterations in endothelium functions due to imbalanced vasodilators and vasoconstrictors, procoagulant and anticoagulant mediators, as well as growth inhibitor and promotor substances. Chlorogenic acid (CGA) is the primary polyphenol in coffee and some fruits. It has many health-promoting properties, especially in the cardiovascular system. Many studies investigated the efficacy and mechanism of this compound in vascular health. CGA has several vascular benefits such as anti-atherosclerosis, anti-thrombosis, and anti-hypertensive. This review focuses on the molecular mechanism of CGA in vascular health.",
"keywords": [
"Chlorogenic acid",
"polyphenol",
"endothelial dysfunction",
"vascular health"
],
"content": "Introduction\n\nVascular endothelial cells have a variety of functions, such as the control of blood coagulation, vascular permeability, and tone regulation, as well as quiesce of immune cells1. Endothelial dysfunction (ED) is considered as a cardiovascular events predictor, and it is characterized by alterations in endothelium functions that tend to be vasoconstricted, procoagulant, and prothrombotic2. Chlorogenic acid (CGA) is a compound of phenol that consists of a caffeic and quinic acid moiety; therefore, it is also called 5-O-caffeoylquinic acid (5-CQA), although many authors refer to it as 3-CQA. A cup of coffee (200 ml) consists of 20–350 mg CGA, which contains 35–175 mg of caffeic acid. Therefore, an average coffee drinker consumes 0.5–1 g of CGA daily3. Moreover, this compound is found in fruits, such as pears, strawberries, eggplant, apples, blueberries, and tomatoes4. CGA is widely studied because of its health properties, such as anticancer, antineurodegenerative, antidiabetic, anti-inflammatory, antilipidemic, and antioxidant. This review discusses CGA’s effects on vascular health, focussing on its molecular mechanism.\n\n\nEndothelial dysfunction (ED)\n\nThe functions of vascular endothelial cells includes vascular permeability, and tone regulation, as well as quiesce of immune cells. Vascular tone is mainly regulated by nitric oxide (NO). Healthy endothelium are protected from adhesion and aggregation through the release of NO, prostacyclin, and platelet ADP degradation5. As long as the endothelial layer is healthy and intact, platelets in circulation remain in an inactive state. ED is a cardiovascular events predictor and considered as the initial stage of atherosclerosis development. It is characterized by alterations in endothelium functions due to imbalanced vasodilators and vasoconstrictors, procoagulant and anticoagulant mediators, as well as growth inhibitor and promotor substances6.\n\nAdiponectin is the biomarker of some cardiovascular disease risk factors such as diabetes, metabolic syndrome, atherosclerosis, or obesity. This adipokine has antioxidant, insulin-sensitizing, and anti-inflammatory properties7. Both of its receptors, AdipoR2 and AdipoR1 have anti-atherogenic activity through the improvement of PPAR and AMPK ligand activity8. In endothelial cells, this substance may downregulate adhesion molecules expression such as ICAM-1, which facilitates monocyte attachment to the endothelium by inhibiting TNF-α-mediated activation of NF-κB. The activity of endothelial nitric oxide synthase (eNOS) can also be increased by adiponectin by facilitating its phosphorylation at Ser1177 via AMPK. It also inhibits reactive oxygen species (ROS) production by oxidized low-density lipoprotein (oxLDL) in cultured endothelial cells. These effects show that high adiponectin levels may prevent atherosclerosis9–11.\n\n\nChlorogenic acid (CGA)\n\nCGA is a compound of phenol that consists of caffeic and quinic acid moiety. It is also called 5-O-caffeoylquinic acid (5-CQA), although some authors refer to it as 3-CQA. This compound is the primary polyphenol in coffee. A cup of coffee (200 ml) consists of 20–350 mg CGA, which contains 35–175 mg of caffeic acid12. Therefore, an average coffee drinker consumes 0.5–1g of CGA daily. In addition, this compound is found in some fruits, such as pears, blueberries, eggplant, strawberries, apples, and tomatoes. It is widely studied since it has several healthy properties, such as antioxidant, anti-inflammatory, anticancer, antilipidemic, antidiabetic, anti-hypertensive, and anti-neurodegenerative.\n\n\nMechanism of CGA in inhibiting atherosclerosis\n\nAtherosclerosis is a multifactorial inflammatory disease initiated by oxidative stress and foam cell formation. Foam cell formation can be inhibited by inducing cholesterol efflux to lipid poor apoplipoprotein such as ApoA1. ABCG1 and ABCA1 are cholesterol transporters that play a significant role in mediating cholesterol efflux to high density lipoprotein. These molecules are regulated by nuclear transcriptional factors LXRa and PPARc13. CGA has been shown to significantly increase mRNA levels of PPARγ, LXRα, ABCA1 and ABCG1, as well as the transcriptional activity of PPARγ. In addition, a cholesterol efflux assay showed that three major metabolites, caffeic, ferulic and gallic acids, significantly stimulated cholesterol efflux from RAW264.7 cells. These results suggest that CGA potently reduces atherosclerosis development in ApoE−/− mice and promotes cholesterol efflux from RAW264.7 macrophages14.\n\nCGA also has a dual PPAR α/γ agonist. Previous studies revealed that its administration enhanced AMPK phosphorylation, adiponectin, and its receptors15. These mechanisms indirectly have a beneficial effect on preventing ED, and AMPK activation has been shown to inhibit protein kinase C as a potent atherogenic substance16. Several studies have revealed the effect of PPARγ agonists on improving ED. PPARγ agonist reverses oxLDL-induced ED through AMPK activation, which consequently enhances eNOS activity. Also, it increased adiponectin levels as a potent anti-inflammatory agent17–19. CGA and its major metabolite, caffeic acid, have antioxidant effects in vitro that alter LDL oxidation. The antioxidant effect of this compound increases LDL resistance to ex vivo oxidation14.\n\nLysophosphatidylcholine (LPC) is the primary atherogenic compound of oxLDL. It increases intracellular calcium through store-operated channels (SOCs). Moreover, it decreases cell viability and increases ROS generation. The expression of transient receptor potential canonical (TRPC) channel is significantly increased by LPC treatment20,21. Previous studies showed that CGA inhibited ROS production by reducing TRPC1 expression, and therefore restored cell viability. Meanwhile, it inhibits LPC-induced Ca2+ influx through SOC. Thus, CGA protects endothelial cells from LPC injury and consequently inhibits atherosclerosis22.\n\nHemeoxygenase-1 is induced in response to ROS in endothelial cells, which plays a role in preventing damage. CGA reduces xanthine oxidase-1 and ROS, as well as enhances hemeoxygenase-1 and superoxide dismutase levels in endothelial cells. Its effects were described on endothelial function in an isolated aortic ring from mice. It was also shown in this study to decrease HOCl-induced oxidative damage in endothelial cells, and this mechanism is related to the induction of hemeoxygenase-1 and NO production23. Consuming coffee high in CGA repairs ED by reducing oxidative stress. Previous studies showed that oxidative stress played an essential role in ED24. However, CGA can prevent this, owing to its antioxidant activity. Also, it inhibits vascular and intercellular adhesion molecule-1, as well as the expression of monocyte chemotactic protein-125. In addition, it prevents T2DM and blocks α-glucosidase activity. It was also reported that CGA inhibits disorder of the endothelium through this activity26.\n\n\nMechanism of CGA in inhibiting platelet activation\n\nHypertension, diabetes and dyslipidaemia, well-known cardiovascular-event risk factors, augment inflammation and might induce platelet adherence to the endothelial layer even in the absence of an activator or injury. Meanwhile, damaged endothelium triggers the release of collagen and von Willebrand factor (vWF) from the extracellular matrix and some derivatives such as thrombin, ADP, and thromboxane A2 (TXA2) that finally lead to platelet activation27. Platelet biomarkers are elevated in risk factors of cardiovascular disease such as hypertension, diabetes mellitus, and obesity28. In atherosclerosis and thrombosis, an elevated level of P-selectin becomes the predictive biomarker of potential adverse cardiovascular events like stroke and myocardial infarction29. P-selectin glycoprotein ligand-1 (PSGL-1) plays a crucial role in inflammation and the initial adhesion of leukocytes to areas of injury. Furthermore, it plays an essential role in thrombosis and homeostasis through PSGL-1 signalling and GPIbα in platelets30.\n\nCGA inhibits platelet activation by preventing their secretion and aggregation in a dose-dependent manner (0.1 to 1 mmol/L) (Figure 1), by inhibiting ADP-dependent secretion and preventing their adhesion31. CGA at these concentrations increases PKA activation or cAMP levels and decreases the inflammatory mediators of platelets (sP-selectin, CCL5, sCD40L, and IL-1β). Adenosine A2A is the target of antiplatelet therapy, activation of this receptor results in an enhanced intracellular cAMP and the inhibition of platelet activation and aggregation. Molecular modelling has shown that CGA is compatible with adenosine A2A receptor active site, which forms interactions with amino acids that specifically interact with A2A ligands, such as NECA and adenosine. Interestingly, CGA has demonstrated a lower bleeding effect compared to that of aspirin31.\n\nNO, nitric oxide, vWF, von Willebrand factor; GPVI, glycoprotein VI; GPIb-IX-V, glycoprotein Ib-IX-V; PGI2, prostacyclin2, PTGIR, prostaglandin I2 receptor, PKA, protein kinase A, PAR1, protease activated receptor; TP, thromboxane A2 receptor; cAMP, cyclic adenosine monophosphate; cGMP, cyclic guanosine monophosphate, COX-1, cyclooxygenase-1; ADP, adenosine diphosphate; P2Y1, purinergic signaling receptor Y1,; P2Y12, purinergic signaling receptor Y12, GPIIb/IIIa, glycoprotein IIb/IIIa; S-CD40L, soluble cluster of differentiation 40 ligand; PSGL-1, P-selectin glycoprotein ligand-1.\n\nCGA treatment has also been shown to inhibit TXA2 secretion and suppression of platelet aggregation. It is also an autacoidal molecule, a potent cyclooxygenase (COX)-1 inhibitor with cytochrome c reductase activity. Furthermore, CGA increases cAMP, cGMP, and intracellular Ca2R-antagonists formation31,32. These results suggest that CGA has antiplatelet activity through the increase of cAMP, cGMP and reduction of thromboxane A2 levels. Meanwhile, CGA shows an antiplatelet activity in vitro at a 50 nM concentration in mice. The same result was obtained in vivo after orally administering 400 mg per 30 g body weight to mice. In humans, this dose will be achieved after consuming about three cups of coffee rich in CGA, which will result in a low nM concentrations in the bloodstream33.\n\n\nMechanism of CGA in inhibiting hypertension\n\nThe evidence of CGA as a hypotensive agent has been suggested by many studies, for example in spontaneously hypertensive rats and mild essential hypertensive patients34. CGA controls hypertension by reducing ROS through the attenuation of NAD(P)H-dependent superoxide. This effect inhibits the proliferation of smooth muscle cells in vitro, as well as in vivo by decreasing angiotensin-converting enzyme activity34. Thus, CGA modulates the renin-angiotensin-aldosterone system. Ferulic acid, the CGA metabolite, has a considerable effect on blood pressure reduction. Its administration enhances acetylcholine-induced vasodilation and increases the bioavailability of NO in the arterial vasculature35.\n\nIn addition, CGA extracted from green coffee was tested for its efficacy in lowering the blood pressure (BP) of hypertensive patients. A double-blind and randomized clinical trial on 117 subjects, where the intervention group received different quantities of the CGA extract for 28 days compared to a placebo group, showed that the extract markedly reduced BP without any adverse effects. Meanwhile, a meta-analysis showed that CGA reduced both systolic and diastolic BP36.\n\nPrevious studies showed that CGA modulates NO levels in rat vessels37, and therefore has a vasodilation effect. In addition, a study in humans investigated its acute effect on BP, NO status, and endothelial function; administration of 400 mg resulted in lower systolic and diastolic BP (−2.41 and −1.53 mmHg respectively; p < 0.05) compared to the control group. However, endothelial function and NO status were not significantly influenced. Ward et al. also investigated the acute effect of 900 mg of CGA on BP and endothelial function, and found that there was no marked effect on peak flow-mediated dilation. Meanwhile, there was continuous dilation improvement. Both 900 and 450 mg of CGA resulted in a high (p < 0.05) continuous flow-mediated dilation at 1 h, and higher at 4 h (0.44%)38.\n\nCGA has been shown to induce the production of NO and enhance antioxidant activity. For instance, caffeoylquinic consumption for eight weeks significantly enhanced NO production and reduced NADPH-dependent ROS in the aorta of hypertensive rats39. Also, CGA has been shown to block the expression of the NADPH-oxidase gene, which helps to control vascular tone35. These results indicate that CGA might induce NO production, decrease oxidative stress, and prevent some conditions, such as hypertension and vascular hypertrophy.\n\n\nMechanism of CGA in inhibiting trans-endothelial migration\n\nAtherosclerosis is a complex process that is initiated by inflammation and leukocyte migration to the inflamed area. Expression of cell adhesion molecules (CAM) on the endothelium and the attachment of monocytes to endothelium may play a major role in the early atherogenic process. During this process, adhesion molecules play a pivotal role in leukocyte cells recruitment and cellular matrix protein development. Ninjurin is a crucial molecule that increases the recruitment and activity of leukocytes during inflammation phase. A previous study showed the dose dependent manner inhibitory effect of CGA in mRNA Ninj1 gene expression that induced by LPS. Moreover, CGA significantly inhibited not only NO production but also the expression of COX-2 and iNOS, without any cytotoxicity. CGA also attenuated pro-inflammatory cytokines (including IL-1b and TNF-a) and other inflammation-related markers such as IL-6 in a dose-dependent manner40. Moreover, CGA inhibited the nuclear translocation of NF-kB and blocked LPS-induced β2 integrin expression and L-selectin shedding. Meanwhile, it inhibited LECAM-1 expression on neutrophil membranes. CGA was also shown to inhibit immunoglobulin molecules by decreasing vascular CAM-1 expressions on the endothelium of human umbilical venule. However, another study suggested that its effects on the expression of PECAM-1 does not involve genetic synthesis25,41.\n\nA study by Chang et al.21 showed that CGA treatment significantly reduced the concentration of proinflammatory cytokines that play an important role in the progression and development of atherosclerosis (Figure 2). Its anti-inflammatory properties explained its inhibitory effects on CAM expression, as it suppressed ICAM-1, VCAM-1, and E-selectin expression, which is induced by IL-1β25. However, it should be noted that consuming a high dose coffee might increase the concentration of homocysteine in human plasma that will consequently lead to ED42. A previous study by Chang et al., showed that CGA suppressed cytokine-induced CAM expression and inhibited p50 and p65 nuclear translocation in endothelial cells25. Therefore, this study also showed that it reduced IL-1β-induced ROS production in human umbilical vein endothelial cells (HUVECs). Furthermore, CGA removed RO• and ROO•, as well as DPPH radicals, which are produced from LDL oxidation43–45. Finally, a previous study also showed that CGA at 50 and 25mmol/L inhibited U937 monocyte-like adhesion, expression of adhesion molecules, NF-KB translocation, and ROS production in HUVECs25.\n\nPPAR, peroxisome proliferator activated receptor; HIF-1α, hypoxia inducible factor1-α; VEGF, vascular endothelial growth factor; COX-2, cyclooxygenase-2; PGI, prostacyclin; TXA2, thromboxane A2; MMP9, matrix metalloproteinase 9; NADPH, Nicotinamide adenine dinucleotide phosphate; ROS, reactive oxygen species; CRP, C-reactive protein; TNFα, tumor necrosis factor α; IL6, interleukine 6; MAPK, mitogen-activated protein kinase; NF-kβ, nuclear factor kappa-B, V-CAM, vascular cell adhesion molecule-1, I-CAM, intercellular adhesion molecule; MCP-1, monocyte chemoattractant protein-1, ET-1, endothelin-1; ENOS, endothelial nitric oxidase; NO, nitric oxide, ETA, endothelin A.\n\n\nAnti-angiogenic mechanism of CGA\n\nHypoxia-induced angiogenesis plays a pivotal role in the development of atherosclerotic lesions. It enhances endothelial cell and vascular smooth muscle cell proliferation through the HIF-1a–VEGF pathway, and contributes to vulnerable plaque progression leading to destabilization. During atherogenesis, the tunica intima is thickened due to cell and matrix accumulation, thus impairing oxygen diffusion. The microenvironment within the plaque is hypothesized to be an essential determinant of plaque progression. During hypoxic condition, several HIF-responsive genes are shown to be upregulated in atherosclerosis such as VEGF, endothelin-1, and matrix metalloproteinase-246. Some studies suggest that CGA ameliorates hypoxia induced atherosclerosis via modulation of HIF-1α-VEGF pathway. A study in A549 cells, as well as in DU145 cells, showed that CGA treatment significantly decreased hypoxia-induced HIF-1α protein that consequently reduced the expression of VEGF. Moreover, during hyperglicemia CGA suppressed serum VEGF and HIF-1 alpha translocation. It was also suggested that CGA blocks in vivo and in vitro angiogenesis of HUVEC cells47. In addition, CGA has been shown to phosphorylates VEGFR2, ERK 1/2 and AKT in order to inhibit VEGF-induced proliferation, migration, and invasion of HUVEC cells48.\n\n\nCGA and vascular health in human studies\n\nFrom various human studies (Table 1), it can be seen that CGA administration in various doses resulted in favourable effect in improvement of cardiovascular function, through amelioration of flow mediated dilatation (FMD) after either acute or chronic administration of CGA. A single intake of CGA with the dose of 400 mg improved FMD and lowered BP. Moreover, it has been suggested that administration of low hydroxyhydroquinone CGA results in better improvement of FMD compared to that of high hydroxyhydroquinone36.\n\nFMD, flow mediated dilatation; RHI, reactive hyperemia index; CAVI, cardio-ankle vascular index; SNA, sympathetic nervous activity; NO, nitric oxide; RXNO, S-nitrosothiols and other nitrosylated species; NOx, nitric oxides comprising nitros(yl)ated species + nitrite; BP, blood pressure; SBP, systolic blood pressure; DBP, diastolic blood pressure; CQA, caffeoylquinic acid.\n\n\nAchieving high CGA benefit from coffee manufacturing process\n\nMany procedures have been introduced in the coffee manufacturing process to achieve more benefits during coffee consumption. Previous studies have shown that roasting levels alter CGA content and antioxidant activity; lightly roasted coffee had more of this compound compared to other groups, and has a higher antioxidant activity based on 2,2-diphenyl-1-picrylhydrazyl (DPPH) assay44,49. The most abundant CGA isomer was 5-CQA with an estimate of 69–74% in the extracts, especially those from green beans50. The 5-CQA content decreased to less than 85% in brews from non-roasted green beans obtained from the same location, and the total CGA content in the extracts of dark-, medium-, and light-roasted beans decreased to 80.60%, 62.91%, and 35.60% respectively51. In addition, 4-CQA and 3-CQA were found at higher percentages in the light-roasted brew compared to green beans. Another study using ABTS and Folin-Ciocalteu assays showed that there is high antioxidant activity in medium and light-roasted brews50.\n\nHigh CGA also can be achieved through fermentation process. Some fermentation procedures had been proposed such as using Saccharomyces cereviciae and Bacillus subtilis strains. A fermentation procedure of coffee pulp using Saccharomyces cereviciae resulted in 400% richer CGA content. Interestingly, the addition of ultrasound treatment did not increase the yield from the extracted coffee pulp. Moreover, the use of Bacillus subtilis strains during fermentation process lead to 20% greater CGA content from green coffee bean extract59,60.\n\n\nConclusion\n\nCGA protects vascular health by inhibiting ED. Several mechanisms explain its effects on LPC injury and atherosclerosis, modulation of dual PPAR α/γ agonist, AMPK phosphorylation, adiponectin, and adiponectin receptors. It plays a role in reducing proinflammatory cytokine concentration that contribute to atherosclerosis development and progression. Furthermore, it suppresses the expression of E-selectin, VCAM-1, and ICAM-1, as well as decreases HOCl-induced oxidative damage in endothelial cells. In addition, CGA induces hemeoxygenase-1 and antiplatelet activity through thromboxane A2 (TXA2) reduction, and attenuates ROS by decreasing the production of NAD(P)H-dependent superoxide. Furthermore, it inhibits the activity of ACE and the proliferation of smooth muscle cells. Finally, it has been shown to block the HIF -1α/AKT signalling pathway, which plays a crucial role in the activation of VEGF and angiogenesis.\n\n\nData availability\n\nNo data are associated with this article.",
"appendix": "Acknowledgments\n\nThanks to Brawijaya University and Ministry of Research, Technology, and Higher Education of the Republic of Indonesia.\n\n\nReferences\n\nvan Hinsbergh VWM: Endothelium--role in regulation of coagulation and inflammation. Semin Immunopathol. 2012; 34(1): 93–106. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarthelmes J, Nägele MP, Ludovici V, et al.: Endothelial dysfunction in cardiovascular disease and Flammer syndrome-similarities and differences. EPMA J. 2017; 8(2): 99–109. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBagchi D, Moriyama H, Swaroop A: eds. Green Coffee Bean Extract in Human Health. CRC Press/Taylor & Francis Group; 2017. Reference Source\n\nPontes PV, Moreira RF, Trugo LC, et al.: The content of chlorogenic acids in tropical fruits. J Sci Food Agr. 2002; 82(10): 1177–1181. Publisher Full Text\n\nCahill PA, Redmond EM: Vascular endothelium - Gatekeeper of vessel health. Atherosclerosis. 2016; 248: 97–109. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFavero G, Paganelli C, Buffoli B, et al.: Endothelium and its alterations in cardiovascular diseases: life style intervention. Biomed Res Int. 2014; 2014: 801896. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEbrahimi-Mamaeghani M, Mohammadi S, Arefhosseini SR, et al.: Adiponectin as a potential biomarker of vascular disease. Vasc Health Risk Manag. 2015; 11: 55–70. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKadowaki T, Yamauchi T, Kubota N, et al.: Adiponectin and adiponectin receptors in insulin resistance, diabetes, and the metabolic syndrome. J Clin Invest. 2006; 116(7): 1784–1792. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBahceci M, Gokalp D, Bahceci S, et al.: The correlation between adiposity and adiponectin, tumor necrosis factor alpha, interleukin-6 and high sensitivity C-reactive protein levels. Is adipocyte size associated with inflammation in adults? J Endocrinol Invest. 2007; 30(3): 210–214. PubMed Abstract | Publisher Full Text\n\nCheng KKY, Lam KSL, Wang Y, et al.: Adiponectin-induced endothelial nitric oxide synthase activation and nitric oxide production are mediated by APPL1 in endothelial cells. Diabetes. 2007; 56(5): 1387–1394. PubMed Abstract | Publisher Full Text\n\nDeng G, Long Y, Yu YR, et al.: Adiponectin directly improves endothelial dysfunction in obese rats through the AMPK-eNOS Pathway. Int J Obes (Lond). 2010; 34(1): 165–171. PubMed Abstract | Publisher Full Text\n\nRenouf M, Guy PA, Marmet C, et al.: Measurement of caffeic and ferulic acid equivalents in plasma after coffee consumption: small intestine and colon are key sites for coffee metabolism. Mol Nutr Food Res. 2010; 54(6): 760–766. PubMed Abstract | Publisher Full Text\n\nBobryshev YV, Ivanova EA, Chistiakov DA, et al.: Macrophages and Their Role in Atherosclerosis: Pathophysiology and Transcriptome Analysis. Biomed Res Int. 2016; 2016: 9582430. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWu C, Luan H, Zhang X, et al.: Chlorogenic acid protects against atherosclerosis in ApoE-/- mice and promotes cholesterol efflux from RAW264.7 macrophages. Cignarella A, ed. PLoS One. 2014; 9(9): e95452. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSanchez MB, Miranda-Perez E, Verjan JCG, et al.: Potential of the chlorogenic acid as multitarget agent: Insulin-secretagogue and PPAR α/γ dual agonist. Biomed Pharmacother. 2017; 94: 169–175. PubMed Abstract | Publisher Full Text\n\nYang L, Cong HL, Wang SF, et al.: AMP-activated protein kinase mediates the effects of lipoprotein-associated phospholipase A2 on endothelial dysfunction in atherosclerosis. Exp Ther Med. 2017; 13(4): 1622–1629. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHamblin M, Chang L, Fan Y, et al.: PPARs and the cardiovascular system. Antioxid Redox Signal. 2009; 11(6): 1415–1452. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJanani C, Ranjitha Kumari BD: PPAR gamma gene--a review. Diabetes Metab Syndr. 2015; 9(1): 46–50. PubMed Abstract | Publisher Full Text\n\nLee WH, Kim SG: AMPK-Dependent Metabolic Regulation by PPAR Agonists. PPAR Res. 2010; 2010: 549101. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLaw SH, Chan ML, Marathe GK, et al.: An Updated Review of Lysophosphatidylcholine Metabolism in Human Diseases. Int J Mol Sci. 2019; 20(5): 1149. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee YK, Lee DH, Kim JK, et al.: Lysophosphatidylcholine, oxidized low-density lipoprotein and cardiovascular disease in Korean hemodialysis patients: analysis at 5 years of follow-up. J Korean Med Sci. 2013; 28(2): 268–273. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJung HJ, Im SS, Song DK, et al.: Effects of chlorogenic acid on intracellular calcium regulation in lysophosphatidylcholine-treated endothelial cells. BMB Rep. 2017; 50(6): 323–328. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJiang R, Hodgson JM, Mas E, et al.: Chlorogenic acid improves ex vivo vessel function and protects endothelial cells against HOCl-induced oxidative damage, via increased production of nitric oxide and induction of Hmox-1. J Nutr Biochem. 2016; 27: 53–60. PubMed Abstract | Publisher Full Text\n\nJokura H, Watanabe I, Umeda M, et al.: Coffee polyphenol consumption improves postprandial hyperglycemia associated with impaired vascular endothelial function in healthy male adults. Nutr Res. 2015; 35(10): 873–881. PubMed Abstract | Publisher Full Text\n\nChang WC, Chen CH, Lee MF, et al.: Chlorogenic acid attenuates adhesion molecules upregulation in IL-1beta-treated endothelial cells. Eur J Nutr. 2010; 49(5): 267–275. PubMed Abstract | Publisher Full Text\n\nOboh G, Agunloye OM, Adefegha SA, et al.: Caffeic and chlorogenic acids inhibit key enzymes linked to type 2 diabetes (in vitro): a comparative study. J Basic Clin Physiol Pharmacol. 2015; 26(2): 165–70. PubMed Abstract | Publisher Full Text\n\nBergmeier W, Hynes RO: Extracellular matrix proteins in hemostasis and thrombosis. Cold Spring Harb Perspect Biol. 2012; 4(2): a005132. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKaur R, Kaur M, Singh J: Endothelial dysfunction and platelet hyperactivity in type 2 diabetes mellitus: molecular insights and therapeutic strategies. Cardiovasc Diabetol. 2018; 17(1): 121. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBlann AD, Ridker PM, Lip GYH: Inflammation, cell adhesion molecules, and stroke: tools in pathophysiology and epidemiology? Stroke. 2002; 33(9): 2141–2143. PubMed Abstract | Publisher Full Text\n\nKappelmayer J, Nagy B Jr: The Interaction of Selectins and PSGL-1 as a Key Component in Thrombus Formation and Cancer Progression. Biomed Res Int. 2017; 2017: 6138145. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFuentes E, Caballero J, Alarcón M, et al.: Chlorogenic Acid Inhibits Human Platelet Activation and Thrombus Formation. PLoS One. 2014; 9(3): e90699. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCho HJ, Kang HJ, Kim YJ, et al.: Inhibition of platelet aggregation by chlorogenic acid via cAMP and cGMP-dependent manner. Blood Coagul Fibrinolysis. 2012; 23(7): 629–635. PubMed Abstract | Publisher Full Text\n\nLudwig IA, Clifford MN, Lean MEJ, et al.: Coffee: biochemistry and potential impact on health. Food Funct. 2014; 5(8): 1695. PubMed Abstract | Publisher Full Text\n\nSuzuki A, Yamamoto N, Jokura H, et al.: Chlorogenic acid attenuates hypertension and improves endothelial function in spontaneously hypertensive rats. J Hypertens. 2006; 24(6): 1065–1073. PubMed Abstract | Publisher Full Text\n\nZhao Y, Wang J, Ballevre O, et al.: Antihypertensive effects and mechanisms of chlorogenic acids. Hypertens Res. 2012; 35(4): 370–374. PubMed Abstract | Publisher Full Text\n\nMubarak A, Bondonno CP, Liu AH, et al.: Acute Effects of Chlorogenic Acid on Nitric Oxide Status, Endothelial Function, and Blood Pressure in Healthy Volunteers: A Randomized Trial. J Agric Food Chem. 2012; 60(36): 9130–9136. PubMed Abstract | Publisher Full Text\n\nTom ENL, Girard-Thernier C, Demougeot C: The Janus face of chlorogenic acid on vascular reactivity: A study on rat isolated vessels. Phytomedicine. 2016; 23(10): 1037–1042. PubMed Abstract | Publisher Full Text\n\nWard NC, Hodgson JM, Woodman RJ, et al.: Acute effects of chlorogenic acids on endothelial function and blood pressure in healthy men and women. Food Funct. 2016; 7(5): 2197–2203. PubMed Abstract | Publisher Full Text\n\nLoader TB, Taylor CG, Zahradka P, et al.: Chlorogenic acid from coffee beans: evaluating the evidence for a blood pressure-regulating health claim. Nutr Rev. 2017; 75(2): 114–133. PubMed Abstract | Publisher Full Text\n\nHwang SJ, Kim YW, Park Y, et al.: Anti-inflammatory effects of chlorogenic acid in lipopolysaccharide-stimulated RAW 264.7 cells. Inflamm Res. 2014; 63(1): 81–90. PubMed Abstract | Publisher Full Text\n\nMills CE, Flury A, Marmet C, et al.: Mediation of coffee-induced improvements in human vascular function by chlorogenic acids and its metabolites: Two randomized, controlled, crossover intervention trials. Clin Nutr. 2017; 36(6): 1520–1529. PubMed Abstract | Publisher Full Text\n\nOlthof MR, Hollman PC, Zock PL, et al.: Consumption of high doses of chlorogenic acid, present in coffee, or of black tea increases plasma total homocysteine concentrations in humans. Am J Clin Nutr. 2001; 73(3): 532–538. PubMed Abstract | Publisher Full Text\n\nGordon MH, Wishart K: Effects of Chlorogenic Acid and Bovine Serum Albumin on the Oxidative Stability of Low Density Lipoproteins in Vitro. J Agric Food Chem. 2010; 58(9): 5828–5833. PubMed Abstract | Publisher Full Text\n\nCha JW, Piao MJ, Kim KC, et al.: The Polyphenol Chlorogenic Acid Attenuates UVB-mediated Oxidative Stress in Human HaCaT Keratinocytes. Biomol Ther (Seoul). 2014; 22(2): 136–142. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZang LY, Cosma G, Gardner H, et al.: Effect of chlorogenic acid on hydroxyl radical. Mol Cell Biochem. 2003; 247(1-2): 205–10. PubMed Abstract | Publisher Full Text\n\nCamaré C, Pucelle M, Nègre-Salvayre A, et al.: Angiogenesis in the atherosclerotic plaque. Redox Biol. 2017; 12: 18–34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPark JJ, Hwang SJ, Park JH, Lee HJ: Chlorogenic acid inhibits hypoxia-induced angiogenesis via down-regulation of the HIF-1α/AKT pathway. Cell Oncol (Dordr). 2015; 38(2): 111–118. PubMed Abstract | Publisher Full Text\n\nLin S, Hu J, Zhou X, Cheung PCK: Inhibition of vascular endothelial growth factor-induced angiogenesis by chlorogenic acid via targeting the vascular endothelial growth factor receptor 2-mediated signaling pathway. J Funct Foods. 2017; 32: 285–295. Publisher Full Text\n\nXu JG, Hu QP, Liu Y: Antioxidant and DNA-protective activities of chlorogenic acid isomers. J Agric Food Chem. 2012; 60(46): 11625–11630. PubMed Abstract | Publisher Full Text\n\nChoi S, Jung S, Ko K: Effects of Coffee Extracts with Different Roasting Degrees on Antioxidant and Anti-Inflammatory Systems in Mice. Nutrients. 2018; 10(3): 363. PubMed Abstract | Publisher Full Text | Free Full Text\n\ndel Castillo MD, Ames JM, Gordon MH: Effect of Roasting on the Antioxidant Activity of Coffee Brews. J Agric Food Chem. 2002; 50(13): 3698–3703. PubMed Abstract | Publisher Full Text\n\nKajikawa M, Maruhashi T, Hidaka T, et al.: Coffee with a high content of chlorogenic acids and low content of hydroxyhydroquinone improves postprandial endothelial dysfunction in patients with borderline and stage 1 hypertension. Eur J Nutr. 2019; 58(3): 989–996. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAgudelo-Ochoa GM, Pulgarín-Zapata IC, Velásquez-Rodriguez CM, et al.: Coffee Consumption Increases the Antioxidant Capacity of Plasma and Has No Effect on the Lipid Profile or Vascular Function in Healthy Adults in a Randomized Controlled Trial. J Nutr. 2016; 146(3): 524–531. PubMed Abstract | Publisher Full Text\n\nOchiai R, Jokura H, Suzuki A, et al.: Green Coffee Bean Extract Improves Human Vasoreactivity. Hypertens Res. 2004; 27(10): 731–737. PubMed Abstract | Publisher Full Text\n\nSuzuki A, Nomura T, Jokura H, et al.: Chlorogenic acid-enriched green coffee bean extract affects arterial stiffness assessed by the cardio-ankle vascular index in healthy men:a pilot study. Int J Food Sci Nutr. 2019; 70(7): 901–908. PubMed Abstract | Publisher Full Text\n\nOchiai R, Sugiura Y, Otsuka K, et al.: Coffee bean polyphenols ameliorate postprandial endothelial dysfunction in healthy male adults. Int J Food Sci Nutr. 2015; 66(3): 350–354. PubMed Abstract | Publisher Full Text\n\nOchiai R, Sugiura Y, Shioya Y, et al.: Coffee polyphenols improve peripheral endothelial function after glucose loading in healthy male adults. Nutr Res. 2014; 34(2): 155–159. PubMed Abstract | Publisher Full Text\n\nBoon EAJ, Croft KD, Shinde S, et al.: The acute effect of coffee on endothelial function and glucose metabolism following a glucose load in healthy human volunteers. Food Funct. 2017; 8(9): 3366–3373. PubMed Abstract | Publisher Full Text\n\nSantos da Silveira J, Durand N, Lacour S, et al.: Solid-state fermentation as a sustainable method for coffee pulp treatment and production of an extract rich in chlorogenic acids. Food and Bioproducts Processing. 2019; 115: 175–184. Publisher Full Text\n\nKim B, Lee DS, Kim HS, et al.: Bioactivity of Fermented Green Coffee Bean Extract Containing High Chlorogenic Acid and Surfactin. J Med Food. 2019; 22(3): 305–313. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "76178",
"date": "01 Feb 2021",
"name": "Katalina Muñoz-Durango",
"expertise": [
"Reviewer Expertise Antioxidants",
"mass spectrometry",
"biomarkers",
"bioavailability",
"clinical studies",
"obesity",
"cardiovascular health",
"chronic non-communicable diseases"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI consider that the topic is of great interest. Chlorogenic acids are important bioactive compounds that have been a matter of extensive research in the last decades. Therefore, a review related to the cardiovascular protective effect of chlorogenic acid: focus on the molecular mechanism, is convenient and necessary. Nevertheless, some important aspects should take into account.\nINTRODUCTION\nInstead of \"Chlorogenic acid (CGA) is the primary polyphenol\" I suggest \"Chlorogenic acid (CGA) is the primary phenolic compound\"\n\nChange \"compound of phenol\" by \"phenolic compound\"\n\nPlease, verify the sentence “Chlorogenic acid (CGA) is a compound of phenol that consists of a caffeic and quinic acid moiety; therefore, it is also called 5-O-caffeoylquinic acid (5-CQA), although many authors refer to it as 3-CQA”\n\nIt is well known that the major CGAs isomers in coffee include chlorogenic (3-CQA), cryptochlorogenic (4-CQA) and neochlorogenic (5-CQA) acids, as well the dimers 3,4-diCQA, 3,5-diCQA and 4,5-diCQA [1,2].\nAdd the reference: A cup of coffee (200 ml) consists of 20–350 mg CGA, which contains 35–175 mg of caffeic acid.\n\nRef 3 is used to indicate the average of CGA consumed by coffee drinkers, please do not use a reference related to green coffee extract, and be sure that you are talking about chlorogenic acid and not chlorogenic acids (CGAs), that is, in my opinion, the right on in this case. Authors should clarify when they refer to chlorogenic acid as the chemical entity (the ester of caffeic acid and (−)-quinic acid) or when they want to talk about chlorogenic acids (CGAs) as the group of phenolic compounds formed by a hydroxycinnamic acid and quinic acid classified by the number, position and type of hydroxycinnamic acid. The most common groups of CGAs are ρ-coumaroylquinic acids, feruloylquinic acids, caffeoylquinic acids (CQAs) and dicaffeoylquinic acids (diCQAs). As I mentioned before, to date, the major CGAs isomers in coffee include chlorogenic (3-CQA), cryptochlorogenic (4-CQA) and neochlorogenic (5-CQA) acids, as well the dimers 3,4-diCQA, 3,5-diCQA and 4,5-diCQA.\nENDOTHELIAL DYSFUNCTION (ED)\nTaking into account that the paper is related to molecular mechanism, I suggest the author consider go further back “ED is a cardiovascular events predictor and considered as the initial stage of atherosclerosis development”. Please, consider in this section to mention the key atherogenic process related to ED. Macrophage foam cells and their role in atherosclerosis. Oxygen species (ROS) production, inflammatory responses and accumulation of lipids, which lead to fatty streak formation in the vascular wall. The massive uptake of oxLDL by macrophages via scavenger receptors (SR-A and CD36) and lectin-like oxLDL receptor-1 (LOX-1), etc. Oxylipins and prostaglandins, among others.\nCHLOROGENIC ACID\nThis section is extremely general, in my opinion, it could be added in any other section.\n\nIdem previous sections. Please be clear with the chemistry, avoid sentences like “It is also called 5-O-caffeoylquinic acid (5-CQA), although some authors refer to it as 3-CQA”, it is not important if other authors confuse the nomenclature, do it right is enough, in my opinion. But, it is a matter of discussion in this paper, please reference the wrong ones, and discuss. Please, check: “A cup of coffee (200 ml) consists of 20–350 mg CGA” You should make the difference between the chemical compound (CGA) and the group of the isomers named CGAS (I already mentioned it before).\nMECHANISM OF CGA IN INHIBITING ATHEROSCLEROSIS\nAdd reference: CGA has been shown to significantly increase mRNA levels of PPARγ, LXRα, ABCA1 and ABCG1, as well as the transcriptional activity of PPARγ.\n\nThe next section is not adequately referenced. The only bibliography added does not correspond to the effect of chlorogenic acids.\n“CGA has been shown to significantly increase mRNA levels of PPARγ, LXRα, ABCA1 and ABCG1, as well as the transcriptional activity of PPARγ. In addition, a cholesterol efflux assay showed that three major metabolites, caffeic, ferulic and gallic acids, significantly stimulated cholesterol efflux from RAW264.7 cells. These results suggest that CGA potently reduces atherosclerosis development in ApoE−/− mice and promotes cholesterol efflux from RAW264.7 macrophages (14)”\nRelated to CGAs in inhibiting atherosclerosis, authors should consider literature that mentioned their in vitro and in vivo (clinical studies) effects on oxylipins, isoprostanes, and prostaglandins.\nMECHANISM OF CGA IN INHIBITING HYPERTENSION\nI suggest the authors consider the expression “inhibiting hypertension”, for this section. In my opinion, it exceeds what the studies have found. There are conflicting results related to this topic.\nRef 38: important to mention that There was no significant effect of any of the treatments on BP.\n\nIn the last paragraph on the action add caffeoylquinic “acid”\nIn this section, the authors analyzed FMD, BP, and NO, nevertheless, in the section “CGA and vascular health in human studies”, they go back to the same topics. I would like to see more linkage among the topics, and more discussion related to table 1. In this table, you include 11 references, but in the section authors only discuses ref 36.\nACHIEVING HIGH CGA BENEFIT FROM COFFEE MANUFACTURING PROCESS\nThis section is extremely general. I don't see the point to have an independent section. This information could be added to the introduction, in a general manner. The paper is not about coffee, is about CGAs. Coffee is a very important source, but the focus of the paper is more oriented to the molecule 5-CQA, and sometimes to the group of isomers (CGAs). Authors must clarify it, as well as when use references related to coffee.\nCONCLUSION The conclusion leaves out many of the aspects mentioned and discussed.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Partly",
"responses": []
},
{
"id": "77826",
"date": "24 Feb 2021",
"name": "Arrigo F. G. Cicero",
"expertise": [
"Reviewer Expertise Cardiovascular disease prevention in clinical settings"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI’ve read with attention the narrative review by Lukitasari et al. on the pharmacological activities of chlorogenic acid on vascular health. The review is interesting, well-organized, overall well-written and updated. My only suggestion is to report quantitative results when speaking about human studies. The conclusion should also include 1-2 sentences on the perspective of research in this field.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": []
},
{
"id": "79914",
"date": "02 Mar 2021",
"name": "Suowen Xu",
"expertise": [
"Reviewer Expertise natural product pharmacology",
"atherosclerosis",
"endothelial cell biology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a very comprehensive review of CGA in cardiovascular health and diseases. By reading through the whole manuscript, the message conveyed is scattered and not focused. For example, the authors can divide the Atherosclerosis related mechanisms into sub-sections: endothelial dysfunction, macrophage inflammation and foam cell formation, VSMC dysfunction, platelet activation etc. How other CVD protective actions, the authors can talk about cardiac hypertrophy and heart failure, as CVD is very broad, it is hard to give concise review and discussion on all forms of CVD. I would suggest to focus on hypertension and atherosclerosis in particular. The detailed molecular targets of CGA can be given in a pictorial way as new figures which is very important.\n\nIt is also good to include are there any clinical trials of CGA in Clinicaltrial.gov? What are unexplored and important in CGA research? In particular, CGA functions as more than antioxidant, which failed in most clinical trials.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1462
|
https://f1000research.com/articles/8-2024/v1
|
28 Nov 19
|
{
"type": "Method Article",
"title": "Fast effect size shrinkage software for beta-binomial models of allelic imbalance",
"authors": [
"Joshua P. Zitovsky",
"Michael I. Love",
"Joshua P. Zitovsky"
],
"abstract": "Allelic imbalance occurs when the two alleles of a gene are differentially expressed within a diploid organism, and can indicate important differences in cis-regulation and epigenetic state across the two chromosomes. Because of this, the ability to accurately quantify the proportion at which each allele of a gene is expressed is of great interest to researchers. This becomes challenging in the presence of small read counts and/or sample sizes, which can cause estimates for allelic expression proportions to have high variance. Investigators have traditionally dealt with this problem by filtering out genes with small counts and samples. However, this may inadvertently remove important genes that have truly large allelic imbalances. Another option is to use Bayesian estimators to reduce the variance. To this end, we evaluated the accuracy of three different estimators, the latter two of which are Bayesian shrinkage estimators: maximum likelihood, approximate posterior estimation of GLM coefficients (apeglm) and adaptive shrinkage (ash). We also wrote C++ code to quickly calculate ML and apeglm estimates, and integrated it into the apeglm package. The three methods were evaluated on both simulated and real data. Apeglm consistently performed better than ML according to a variety of criteria, including mean absolute error and concordance at the top. While ash had lower error and greater concordance than ML on the simulations, it also had a tendency to over-shrink large effects, and performed worse on the real data according to error and concordance. Furthermore, when compared to five other packages that also fit beta-binomial models, the apeglm package was substantially faster, making our package useful for quick and reliable analyses of allelic imbalance. Apeglm is available as an R/Bioconductor package at http://bioconductor.org/packages/apeglm.",
"keywords": [
"RNA-seq",
"Allelic imbalance",
"Allele-specific expression (ASE)",
"Beta-binomial",
"Shrinkage estimation",
"Empirical Bayes",
"Bioconductor",
"Statistical software"
],
"content": "Introduction\n\nAllelic imbalance (AI) occurs when the two alleles of a gene are expressed at different levels in a diploid organism, and the measurement of AI is valuable in elucidating the factors that regulate the expression of genes. For example, for a diploid organism, the allele on one chromosome may have higher or lower expression levels compared to the allele on the other chromosome due to genetic variation in nearby non-coding regulatory sites, a process known as cis-regulation. Allelic imbalance in expression may also be associated with differential epigenetic state of the genomic region across the chromosomes. In some cases, differential allelic expression resulting from differential epigenetic state can be linked to the parent-of-origin of the alleles, a phenomenon known as genetic imprinting.\n\nOne challenge currently faced in allelic expression studies is that estimates for allelic expression proportions can be highly variable in the presence of low read counts and/or small sample sizes. Large estimates of allelic proportions in these cases often result from estimation error as opposed to true differences in allelic expression. Though small samples and low counts are a problem for RNA-seq data in general, they are especially problematic when dealing with allele-specific counts. When a subject is heterozygous for a gene at a particular SNP(s), RNA-seq reads that overlap the SNP(s) allow for quantification of the levels of expression from either chromosome1. Thus, allelic expression cannot be measured within a gene for subjects that are homozygous for that gene, and the number of samples with allele-specific counts for a gene can be much less than the number of samples in the study. Furthermore, alleles are often differentiated by a single SNP, and RNA-seq reads that do not overlap the SNP cannot be mapped to either allele. For these reasons, the proportion of RNA-seq reads that are allele-specific can be quite low, depending on both read length and heterozygosity of the subjects. For instance, one study with 2x50 base pair (bp) paired-end reads and 30 million heterozygous SNPs from breast tumors of 550 human subjects found that allele-specific counts made up only 3.4% of RNA-seq reads2. Experiments making use of model organism crosses can maximize the number of RNA-seq reads overlapping heterozygous SNPs, for example Raghupathy et al.3 found in an RNA-seq dataset of a mouse F1 cross that 22% of uniquely mapping reads were allele-specific.\n\nOne traditional remedy investigators have used to deal with the challenges of high-variance estimates is to filter out genes that have low counts or small samples. While this does cause the resulting estimates to be more stable and thus representative of true allelic expression proportions, filtering may also remove genes that have true allelic imbalances. Furthermore, the cutoff used to determine what genes to filter out (i.e. how many counts a gene must have for it to not be removed) must be chosen per dataset by the analyst. Another potential remedy to the problem of variable estimates from low-count and low-sample genes is to use Bayesian shrinkage estimators to moderate estimates.\n\nA large number of Bayes estimators have already been developed for allelic expression studies. For instance, MMSEQ4 uses a Gamma prior on allele-specific transcript abundance to provide isoform and allelic imbalance estimates that are more accurate and stable in the face of low coverage. Other methods that have used Bayesian approaches to test for AI include those by Leòn-Novelo et al. 20145 and Skelly et al.6 However, these methods can only test overall AI, and cannot test the effects of covariates, such as different groups, on AI. More recently, a method was developed that expanded on that by Leòn-Novelo et al. 2014 and was able to estimate AI within groups as well as compare AI between groups7. It uses Bayesian shrinkage estimates for its parameters to shrink allelic proportions within groups toward 0.5, overdispersion toward a pre-specified prior mean, and the total counts of both alleles toward a pooled estimate. While the method is more flexible than the other methods listed, it still cannot estimate the effects of continuous covariates on allelic imbalance, nor can it estimate differences in AI between groups while controlling for additional confounding variables. Furthermore, while the authors showed their method to be effective in reducing type I and type II error in the face of different sources of bias, the advantage of their method in estimation accuracy itself and in the face of genes with low counts need to be more thoroughly investigated.\n\nThough gene expression read counts are typically larger than allele-specific counts and can be measured for all subjects, the uncertainty of estimates in the presence of low counts and/or low sample sizes is still an issue. Thus, several shrinkage estimators for log fold changes in gene expression have also been developed which shrink estimates that are only large due to the variance of the estimator and leave unchanged estimates that are likely to be large due to true expression changes8–11. Many of these methods directly involve or can easily be applied to linear models, which provide great flexibility in the kinds of study designs that can be treated and hypotheses that can be investigated. Though these methods were originally developed for improving accuracy and stability of log fold change estimates in gene expression, several can be can be directly applied or at least easily extended to estimating the effects of covariates on allelic expression proportions.\n\nTo this end, we look at three different estimation methods and their performance on data sets with small-to-moderate numbers of samples: maximum likelihood (ML), approximate posterior estimation of GLM coefficients (apeglm)11 and adaptive shrinkage (ash)10 ML estimates are based on estimating effects by modelling allele-specific counts with a beta-binomial GLM. Apeglm and ash are Bayesian shrinkage estimators which shrink maximum likelihood-based estimates toward zero. Our results show that while apeglm is not always the best method, it always performs better than ML and never performs much worse than ash for most metrics, making it the most robust and reliable when dealing with small sample sizes in our analysis. We also introduced new source code for the apeglm package to improve computational performance for beta-binomial GLMs, and compared our improved package to other R packages that can also fit beta-binomial GLMs. As the apeglm package can calculate both ML and Bayesian shrinkage estimates, our improvements can be used even by those who wish not to use shrinkage estimators. Compared to other R packages, we show that apeglm with our improved code gives better running times and greater scalability with the number of covariates.\n\n\nMethods\n\nWe evaluated three estimation methods on their ability to estimate allelic expression proportions (or equivalently, the effects of covariates on allelic expression proportions): maximum likelihood estimation (ML estimation or MLE) with the likelihood described below, approximate posterior estimation of GLM coefficients (apeglm) and adaptive shrinkage (ash). All analyses was done using R version 3.5.112. The first two methods mentioned are implemented in the apeglm v.1.7.5 package, while the last is implemented in the ashr v.2.2.32 package. When using the ash function in the latter package, we set the method parameter equal to \"shrink\". While there are many Bayesian estimation methods that can be used to quantify allelic imbalance, these allow for arbitrary design matrices. For instance, these methods can estimate differences in AI between groups while controlling for, or allowing interactions with, multiple additional variables, and can estimate the effects of continuous variables on AI.\n\nFor the g-th gene (1 ≤ g ≤ G), a beta-binomial GLM was fit to model allele-specific counts as follows. Let Yig be the read counts of the first of the two alleles (which allele is designated as the first allele is arbitrary) for the i-th subject, 1 ≤ i ≤ I. Investigators may designate the first and second alleles of a gene as the paternal and maternal alleles or as the alternate and reference alleles. It is assumed that Yig ∼ BetaBin(nig , pig ,ϕg), where nig is the equal to the total counts of both alleles for the i-th subject, pig is the probability of counts belonging to the first allele of the i-th subject, and ϕg is the overdispersion parameter. For the remainder of this paper, we will refer to the total allele-specific counts for both alleles of a particular gene and for a particular sample as the ‘total counts’ for that gene and sample. Furthermore, we will refer to the probability that counts for a particular gene belong to a particular allele for a particular sample as the ‘allelic proportion’ for that particular allele and sample. In this case, ϕ → ∞ implies no overdispersion beyond what would be seen in a binomial distribution and ϕ → 0 implies increasing variance. n1g , ..., nIg are assumed to be fixed and known. As the beta-binomial probability density function has multiple forms and parameterizations, we specify our parametrization as:\n\n\n\nwhere B specifies the beta function. Furthermore, let xi be the i-th row of the design matrix X (matrix where columns are vectors of covariates of interest). Potential predictors include disease status for association studies, parent of origin for imprinting studies, and the presence of a SNP for eQTL linkage studies. We also assume that pig=[1+exp(−xiTβg)]−1, or equivalently logit(pig)=xiTβg, where βg = (β1g , ...,βKg)T is a vector of coefficients representing the effect sizes for the predictors in the design matrix. For ML estimation, βg is estimated via maximum likelihood. Constrained optimization is used for the nuisance parameter ϕg with a maximum of 500, so that genes with no overdispersion have finite estimated values of ϕ.\n\nApeglm additionally assumes a zero-centered Cauchy prior distribution for the effects of one of the predictors11. For estimating the effect of the j-th predictor in our model, where 1 ≤ j ≤ K is chosen by the user, and for the g-th gene, we have:\n\n\n\n\n\n\n\nApeglm shrinks the effect of one chosen predictor at a time, across all genes. The scale parameter of the Cauchy prior, γj, is estimated by pooling information across genes. The posterior distribution of βg is the product of the above Cauchy prior and beta-binomial likelihood, and apeglm provides Bayesian shrinkage estimates based on the mode of the posterior as well as standard errors. Genes with lower expression, smaller numbers of heterozygous subjects and higher dispersion in allelic proportions will have flatter likelihoods, which will lead to the prior having more influence and shrinkage being greater. Furthermore, if the ML estimates are tightly clustered about zero, the estimated scale parameter of the Cauchy prior will be smaller. This will lead to more peakedness in the prior and also cause shrinkage to be greater.\n\nThe original apeglm package estimated regression coefficients using C++ for negative binomial GLMs, while GLMs with other likelihoods, such as the beta-binomial, were fit completely in R. To improve scalability for large data sets with beta-binomial GLMs, we wrote fast C++ code for calculating maximum likelihood and apeglm shrinkage estimates of beta-binomial regression coefficients. We also changed the source code to speed up computation of the standard errors (though such computations were still done in R) and prevent convergence issues. Details can be found in the Supplementary Methods section13.\n\nAsh is a general Empirical Bayes shrinkage estimator for hypothesis testing and measuring uncertainty in a vector of effects of interest, such as a set of log fold changes in gene expression between biological conditions10. Suppose again that one is interested in the effect sizes of the j-th predictor, β.j = (βj1, ..., βjG), where 1 ≤ j ≤ K. Ash takes as input a vector of ML estimated effects β^.j=(β^j1,...,β^jG) and corresponding estimated standard errors σβ.j = (σβ j1 , ..., σβ jG). Here we take the estimated standard errors to be the true standard errors as suggested in the original methodology for ash, though the developers of ash have recently proposed an extension to their method that allows for random errors14. For all 1 ≤ g ≤ G, it is assumed that β^jg|βjg∼N(βjg,σβjg) and that βjg ~ hj , where hj is some unimodal, zero-mode prior distribution. hj is estimated from the ML estimates using mixtures of uniforms and a point-mass at zero, a choice guided by the author’s claim that any unimodal distribution can be approximated as a mixture of uniforms with arbitrary accuracy. The posterior is βjg|β^jg∼N(βjg,σβjg)×hj, and ash provides Empirical Bayes shrinkage estimates using the mean of the posterior as well as standard errors. Genes with larger standard errors for their ML estimates will have a flatter likelihood that will be less impactful on the estimation. Thus, estimates for these genes will be shrunk more. Like apeglm, ash can only shrink estimates for one covariate at a time.\n\nWe compared the three estimation methods using the data set from the allelic expression study by Crowley et al.15,16 The study took mice from three divergent inbred strains (CAST/EiJ, PWK/PhJ and WSB/EiJ) and performed a diallel cross. The data set contains ASE counts for 72 mice and 23,297 genes in the resulting cross, with 12 mice of each possible parent combination (e.g. CAST/EiJ as mother and PWK/PhJ as father is one parent combination, and PWK/PhJ as mother and CAST/EiJ as father is another), and an equal number of males and females within each parent combination. Sequencing was performed with the Illumina HiSeq 2000 platform to generate 100-bp paired-end reads and following the TruSeq RNA Sample Preparation v2 protocol. To assure that the mice all had the same alleles, we chose one genotype to focus on, namely the genotype resulting from the cross with CAST/EiJ and PWK/PhJ. The resulting data set, which we will refer to for the remainder of this paper as the ‘mouse data set’, had 24 mice, 12 of each sex and 12 of each parent of origin, and each mouse had nearly the same nuclear genetic composition as a result of the cross.\n\nTo evaluate the estimators on estimating effect sizes of predictors when the truth is known, we first fit an intercept-only beta-binomial model on each gene for the mouse data set. ϕ = [ϕg] is the vector of ML estimates of the overdispersion parameter from each model, and µ = [µg] is the vector of ML estimates of allelic proportions (which ranges between 0 and 1). 8 mice were then selected from the data set. Denote NI×G = [nig] as the matrix of total ASE counts for the 8 mice. Finally, a matrix of counts from one of the alleles YI×G = [yig] was simulated for a sample size of 4 vs. 4, where yig was simulated from BetaBin(nig, pig, ϕg), logit(pig) = µg + βgx, βg was simulated from a standard normal distribution, and x splits the mice into two groups of size four (x = 1 if a mouse is in the first group and 0 otherwise). Samples were drawn from the beta-binomial distribution using the emdbook v1.3.11 package17. We refer to this simulation throughout the paper as the ‘standard normal simulation’, reflecting the distribution of the true effect sizes.\n\nA second simulation was also performed that was similar in setup to the first, but with modifications to the distribution of βg and ϕg. In many studies, the effect sizes of a predictor will be zero for all but a handful of genes. Thus, βg was simulated from t3/10 (a Student’s t-distribution with 3 degrees of freedom scaled by 1/10), which gave effects mostly close to zero, but with moderate and large effects occasionally appearing (Supplementary Figure 113). Furthermore, the distribution of ϕg from the mouse data appeared to be a mixture of two distributions: Genes without overdispersion had an obvious point mass at 500 with 70% proportion, and the remaining 30% genes had a distribution somewhat similar to an exponential with a mean of β = 179 (Supplementary Figure 213). To get more over-dispersed allele-specific counts, ϕg was simulated from 0.5Exp(β=89) + 0.5(500), a mixture distribution where one component was exponential with a mean of 89 and had 50% proportion, and the other component was a point mass at 500 and had 50% proportion. We refer to this simulation throughout the paper as the ‘Student’s t simulation’, again reflecting the distribution of the true effect sizes. Note that these two simulations assume a data generating process, specifically the same data generating process as our assumed likelihood.\n\nThe estimators were then evaluated on real data with the focus on estimating mean, or gene-wide, allelic imbalance. From the mouse data set, random samples of size 6 were drawn, and this process was repeated 100 times. We will refer to these samples throughout the paper as the ‘random subsamples’. For each random subsample, the ML, apeglm and ash estimates of intercept-only models were calculated for the genes (where the intercept term was shrunk), and the MLE of the held-out 18 mice was taken to be the truth. Estimating the intercept in an intercept-only model for each gene is equivalent to estimating overall allelic imbalance for each gene.\n\nAdditional simulations were conducted for evaluating computational performance of our improvements to apeglm, to see how well they would scale to larger and more complicated data sets. Allele-specific counts were simulated in a similar manner as the apeglm vignette18. Briefly, we have Y100×5000 = [yig] as our simulated count matrix for one allele with associated total count matrix N100×5000 = [nig] where rows are samples and columns are genes, yig ~ BetaBin(nig, pg, θg), θg ~ U (0, 1000), pg ~ N (.5, 0.52), nig ~ NB(µg, 1/ϕg), and µg, ϕg are based on the airway data set by Himes et al.19 To see how well our improvements scaled with increasing numbers of covariates, the data were split multiple times into differing numbers of groups of approximately equal size, where the number of groups ranged from 2 to 10. With K groups, the design matrix was X100×K = [1 x1 ... xK –1], where xj is an indicator variable for the (j + 1)-th group, or a row vector whose i-th element is 1 if the i-th sample is in the (j + 1)-th group and 0 otherwise. A simulation was also conducted to see how well apeglm would work with continuous predictors. This time, Y and N was kept the same, but with the design matrix X100×4 = (1, x1, x2, x3) = [xij], where x1 = (1, 0, 1..., 1, 0)T separates the samples into two equally sized groups and xi2, xi3 ~ N (0, 1). x1 is the covariate whose effect size estimates are shrunk.\n\nGenes where at least three samples did not have at least 10 counts were removed, which we considered minimal filtering that shouldn’t decrease statistical power. Genes without at least one count for both alleles across all individuals were removed. Genes with a marginally significant sex or parent effect were removed, so that all samples could be assumed independent and identically distributed for all genes. Genes were removed from the mouse data set prior to conducting random sampling from the data set or simulations.\n\nTo determine whether sex or parent effects were significant, beta-binomial GLMs were estimated for each gene by maximum likelihood, with a design matrix that included a sex effect (an indicator that was 1 if male and 0 if female), a parent-of-origin effect (an indicator that was 1 if the mother was the CAST/EiJ strain and 0 if the father was the CAST/EiJ strain) and an interaction term. For each gene, if the p-value for the sex, parent-of-origin or interaction effect was less than 0.1, the effect was deemed marginally significant for that gene.\n\nFor each gene, we define the shrinkage score as movement from MLE to zero. We define a gene as (noticeably) shrunk if shrinkage exceeds 0.1, and substantially or most shrunk if shrinkage is greater than max (1,|β^MLE|/4). For instance, if an apeglm estimate for a gene is 0.15 closer to zero than the MLE, then the shrinkage score is 0.15 and the gene is noticeably shrunk but not substantially shrunk by apeglm.\n\nConcordance at the top (CAT) plots20 were used to determine which estimation method could best find the most important genes (the genes with the greatest allelic imbalance or largest effect size). For an estimation method, concordance at the top takes the top genes according to the true ranking and compares it to the top genes according to the estimates, where the top genes are the genes with the largest true or estimated effect sizes in absolute value. For instance, a concordance at the top 10 of 90% means that the top 10 genes according to the estimation method and the top 10 genes according to the truth agree for 9 out of 10 genes.\n\nFor evaluating the performance of the three methods in estimating intervals, we calculated normality-based 95% confidence and credible intervals (both of which we will abbreviate as CIs) of the ML and apeglm estimators using their standard errors, or intervals based on the Laplace approximation of the likelihood and posterior. Such normality-based intervals are the default and suggested method for the apeglm package. Credible intervals in the ashr package were calculated from directly estimating tail probabilities of the posterior.\n\nFor each of the design matrices posited in our computation simulation, computational performance of apeglm estimation was compared between the old and new apeglm code. From apeglm v1.7.5, we set the method parameter equal to “betabinCR\" to run the new C++ code, and set the log.lik parameter equal to a beta-binomial log-likelihood function to run the old code from before our improvements were introduced (version 1.6.0 of the package). Details can be found in the vignette18. Computational performance of ML estimation was also compared between our improved apeglm package and the following packages: aod v1.3.121, VGAM v1.122, aods3 v0.423, gamlss v5.124 and HRQoL v1.025. Computational performance was evaluated using the microbenchmark v1.4.6 package26 for estimation of a single gene and elapsed time for estimation of all 5000 genes, on a 2012 15-inch MacBook Pro with an Intel Core i7-3720QM processor.\n\nIn addition to comparing the three estimation methods described above, maximum likelihood estimation paired with optimal filtering criteria was also assessed via concordance at the top. CAT was chosen over other benchmark metrics, such as mean absolute error, as the different number of genes after filtering would make comparisons between filtered MLE and the three unfiltered methods biased. Furthermore, as we were primarily interested in whether a good filtering rule even existed, the true ranking of genes was used to determine the filtering rule. We looked at three rules: 1) removing genes where less than half the samples had a minimum total count threshold, 2) removing genes where less than all the samples had a minimum total count threshold, and 3) removing genes where the sum of total counts across samples was less than a certain threshold. For the remainder of the paper, we will refer to the sum of total counts across samples as the ‘summed counts’ of a gene. For each rule, various different thresholds were looked at: {0, 10, ..., 200} were potential thresholds for rule 1, {0, 10, ..., 100} were potential thresholds for rule 2, and {0, 50, ..., 1000} were potential thresholds for rule 3. For each rule and threshold, the MLE was calculated and concordance among the top 50, 100, 200, 300, 400 and 500 genes were averaged. We will refer to the rule and threshold that had the best concordance as the ‘optimal filtering rule’.\n\n\nResults\n\nWe began by looking at a simulation where allelic counts came from known beta-binomial distributions and effect sizes came from a standard normal distribution. In this simulation, apeglm and ash successfully shrunk erroneously large estimates and reduced estimation error, particularly for genes that were noticeably shrunk (see Table 1 and Figure 1).\n\nMLE: Maximum Likelihood Estimation, apeglm: Approximate Posterior Estimation of Generalized Linear Model Coefficients, ash: Adaptive Shrinkage.\n\na) truth vs. estimate plot for MLE. Blue points represent genes substantially shrunk by apeglm only, orange points represent genes substantially shrunk by ash only and green points represent genes substantially shrunk by both ash and apeglm. b) truth vs. estimate plots for apeglm. c) truth vs. estimate plots for ash. d) CAT plot for the three methods as well as for MLE after filtering. CAT: Concordance at the Top, MLE: Maximum Likelihood Estimation, apeglm: Approximate Posterior Estimation of Generalized Linear Model Coefficients, ash: Adaptive Shrinkage.\n\nAll three estimation methods gave similar mean absolute error (MAE), as many genes did not differ much between the methods (Table 1). In exploring the behavior of shrinkage estimators, we were most interested in genes where shrinkage was high, and thus where estimates would be much closer to or much farther from the truth for one estimation method than for another. Thus, in addition to overall MAE, we also calculated MAE among genes that were noticeably shrunk by apeglm and genes that were noticeably shrunk by ash, to determine whether there was substantial improvement on average when apeglm or ash did noticeably shrink a gene. Among genes that were shrunk by apeglm, apeglm decreased the mean absolute error by 18.1%, and among genes that were shrunk by ash, ash decreased the mean absolute error by 21.1%. Moreover, from Figure 1a–c, it can be seen that apeglm shrunk most ML estimates that were inflated, bringing them closer to the truth, and mostly left truly large effects alone. Ash also shrunk ML estimates that were inflated, including some inflated estimates missed by apeglm. However, ash also had a tendency to incorrectly and excessively shrink: some genes with estimates close to the truth were severely shrunk, and several genes with truly large effects were shrunk to zero. Because of this tendency to over-shrink, ash performed worse among genes with large effects than among genes with small effects. For instance, among genes with effect sizes greater than two in absolute value, ash estimates had a higher mean absolute error than the MLE.\n\nAsh and apeglm also performed better than the MLE in determining the most important genes, where concordance at the top was higher regardless of the number of genes being considered (Figure 1d). Apeglm performed slightly better than ash in concordance at the top 100 genes, but otherwise they performed about the same. Concordance at the top for the MLE was optimized when filtering out genes with summed counts less than 350. Using this filtering, we were able to get CAT results better than that of the shrinkage estimates, even if only by a very small amount. Thus, for this simulation, it was possible to outperform both apeglm and ash with filtering alone (provided that the true ranking of genes was known, and used to determine the optimal filtering rule).\n\nWith regard to the extent of shrinkage, both apeglm and ash mainly exhibited shrinkage for genes that had very low counts (Supplementary Figure 313). This is not too surprising for this particular simulation, as after filtering out lowly-expressed genes, the remaining ML estimates were much closer to the truth (Supplementary Figure 413). When comparing shrinkage scores between apeglm and ash, we found that there was a clear upward shift of shrinkage scores for ash (Supplementary Table 113), further showing that ash had more extreme shrinkage than apeglm for this dataset. Though all three methods gave intervals that were similar in coverage probability, average interval width was smaller for apeglm and ash compared to the MLE (Table 1).\n\nWe also investigated the performance of the estimators when most of the effect sizes were close to zero and overdispersion was large. Here the shrinkage estimates had even more marked improvement over the ML estimates (see Table 2 and Figure 2).\n\na) truth vs. estimate plot for MLE. Orange points represent genes substantially shrunk by ash only and green points represent genes substantially shrunk by both ash and apeglm. All genes substantially shrunk by apeglm were shrunk by practically the same amount or more by ash. b) truth vs. estimate plots for apeglm. c) truth vs. estimate plots for ash. d) CAT plot for the three methods as well as for ML after filtering. MLE: Maximum Likelihood Estimation, apeglm: Approximate Posterior Estimation of Generalized Linear Model Coefficients, ash: Adaptive Shrinkage.\n\nApeglm improved mean absolute error by 49.5% among all genes, and by 65.4% among noticeably shrunk genes specifically (Table 2). Ash improved mean absolute error by 52.2% among all genes and by 65.8% among noticeably shrunk genes specifically. These improvements were greater than that seen from the standard normal simulation. Figure 2a–c show that ash successfully shrunk inflated ML estimates closer to the truth while leaving truly large effects mostly unchanged. Apeglm brought many inflated ML estimates closer to the truth as well, but not as many as ash.\n\nConcordance at the top was better for the shrinkage estimates than for the ML estimates, regardless of the number of top genes in question. Furthermore, similar to mean absolute error, the improvements seen from the shrinkage estimates over the MLE was larger than those seen from the standard normal simulation. Ash performed better than apeglm in concordance at the top 50 and 100 genes, though performance was similar when looking at larger number of genes. Concordance at the top for the MLE was optimized when filtering out genes where less than half the samples had at least 110 counts. Though this improved CAT by quite a lot, performance was still much lower than apeglm and ash Thus, unlike in the standard normal simulation, the performance in CAT obtained by shrinkage could not be matched with filtering, even when using the true gene ranking to determine the optimal filtering rule.\n\nIn this simulation, due to the increased overdispersion, there were many effects that were overestimated or underestimated by ML, even among genes with large counts. Because of this, both ash and apeglm exhibited shrinkage for effects across the dynamic range of summed counts, as opposed to only shrinking effects with small counts (Supplementary Figure 513). Together with the true vs. estimate plots, this shows that both apeglm and ash can correctly shrink falsely large effects even when the summed counts are large. Ash had larger shrinkage scores than apeglm on average, indicating that ash tended to shrink estimates more than apeglm (Supplementary Table 213). All methods had coverage slightly less than nominal (95%), ranging from 92 to 94%. However, both apeglm and ash had half the average interval width compared to maximum likelihood, despite both having slightly higher coverage rates.\n\nWe also conducted simulations similar to that of the standard normal and Student’s t, but with 5 vs. 5 samples. Like the 4 vs. 4 case, both apeglm and ash had lower average estimation error and higher concordance at the top than the MLE (results not shown).\n\nTo evaluate performance on real data, we took 100 random subsamples of 6 mice from the mouse data set and averaged various performance metrics across the random subsamples. Similar to the simulations, apeglm appeared to improve estimation accuracy and shrink erroneously large genes. Ash, on the other hand, appeared to perform worse than the MLE according to mean absolute error, concordance at the top and interval coverage (see Table 3 and Figure 3).\n\na) through c) is based on one of the 100 random subsamples used in the mouse data benchmarking, and plots the ML estimates of the associated held-out set against the MLE, apeglm and ash estimates from the random subsample, respectively. Orange points represent genes substantially shrunk by ash only and green points represent genes substantially shrunk by both ash and apeglm. All genes substantially shrunk by apeglm were shrunk by practically the same amount or more by ash. d) plots concordance at the top averaged across the 100 random subsamples for each method. MLE: Maximum Likelihood Estimation, apeglm: Approximate Posterior Estimation of Generalized Linear Model Coefficients, ash: Adaptive Shrinkage.\n\nAmong genes that were shrunk by apeglm, mean absolute error was 14.2% lower on average for apeglm (Table 1). On the other hand, average MAE rose by 34.1% for ash among genes that were shrunk. Moreover, the minimum MAE obtained by ash across all 100 random subsamples was larger than the maximum MAE obtained by apeglm (results not shown). From Figure 3a–c, ash appears to be over-shrinking, and some of the genes with the largest held-out effect estimates were shrunk to zero. Though some genes also appeared to have been incorrectly or overly shrunk by apeglm, apeglm mainly was observed to shrink genes with inflated estimates and over-shrinkage was normally less severe when it occurred.\n\nBoth apeglm and MLE had universally higher concordance at the top than ash (Figure 1d). While apeglm performed slightly better than the MLE in concordance at the top 50 genes, performance was identical when looking at larger numbers of genes. We found that any filtering only decreased concordance at the top, as many top genes had low counts (i.e. the optimal filtering rule was no filtering). The most likely reason for this is that for each random subsample, we are treating the MLE of the held-out set as the truth. Thus, estimation error in the face of low-count genes would affect the held-out effect estimates and bias CAT results to some degree, even though the held-out sets have larger numbers of samples and performance metrics are averaged over many random subsamples. However, because genes with very large held-out effect estimates are more likely to have low counts, metrics that average across all genes, such as mean absolute error, would not be biased as much by estimation error.\n\nA large amount of variability in the ML estimates was discernible for genes with low counts (Supplementary Figure 613). Like in the standard normal simulation, the low-count genes were mainly the ones shrunk by apeglm and ash. As the truth vs. estimate plots suggest, ash had larger shrinkage scores than apeglm (indicating more extreme shrinkage), and with the difference in shrinkage between the two methods being larger than in the simulations (Supplementary Table 413). Though apeglm intervals appeared to have smaller coverage than the ML intervals, the difference in coverage was very small, and average interval width was also 26.8% smaller for apeglm than that of maximum likelihood. Ash intervals were slightly more narrow than apeglm, with average interval width 32.5% smaller than that of maximum likelihood, but coverage was also lower.\n\nAs the mouse data set only had 24 samples, we determined that we didn’t have the sufficient sample size to evaluate our methods on estimating effect sizes of predictors, or models with more than just an intercept term. For instance, even if we wanted to look at the performance of our methods on estimating a group effect with only four samples in each group, each held-out set would only have eight samples in each group. Thus, the ML estimates of the held-out sets would have a lot of variance and could be far from the truth.\n\nTo evaluate the computational performance of our package on larger datasets, we simulated allelic counts for 5000 genes and 100 samples, and randomly divided the samples into differing numbers of groups. apeglm with our improvements had very fast running times for both ML and apeglm estimation and scaled well with the number of covariates (see Figure 4 and Figure 5).\n\na) computational time of ML estimation (in seconds) for the apeglm and aod packages by the number of groups (covariates). b) computational time of apeglm estimation for the new and old apeglm packages by number of groups (covariates).\n\nEstimation times per gene for ML estimation was substantially faster for apeglm than all other packages (Figure 4). The next best package, aods3, took 5 to 11 times longer than apeglm and did not scale as well with the number of groups. Furthermore, the aods3. gamlss and HRQoL packages occasionally produced errors and could not fit beta-binomial models for all the simulated genes.\n\nFor estimating all genes in the simulation via maximum likelihood, apeglm took 24 seconds for two groups and added only 1–2 seconds of computational time for every group added (Figure 5a). The next fastest package that could fit beta-binomial models for all the genes, aod, took seven times longer for two groups and grew 80 times as much for every group added. Comparisons in apeglm estimation between our improved apeglm package and the original package gave similar conclusions. Furthermore, unlike the new apeglm package, which grew roughly linearly with the number of groups in the range we assessed, the order of growth from the original package was not linear: the greater the number of groups already in the model, the greater the computational time increased for adding additional groups. At 10 groups, our improvements made apeglm 27 times faster than aod for ML estimation and 33 times faster than the old package for apeglm estimation. Our improvements also performed quite favorably when fitting beta-binomial models with two groups and two numerical controls. Elapsed time was 31 seconds for ML estimation and 43 seconds for apeglm estimation with the new apeglm package. In contrast, ML estimation took over nine minutes for aod and apeglm estimation took over seven minutes for the old apeglm package. Introducing multicollinearity into the design matrix did not substantially change computational performance for any package (results not shown).\n\n\nDiscussion\n\nHere the performance of three estimators was compared across two simulations and one real dataset of allele specific expression in mice. Though apeglm was not the best estimator in all cases, it was the most robust and with consistent performance. Apeglm had smaller mean absolute error and greater concordance at the top than the MLE, and was never much worse than ash in these respects. Ash also performed better than the MLE for the simulated data for most metrics, including mean absolute error and concordance at the top. Moreover, ash had higher concordance at the top than apeglm in the Student’s t simulation. However, ash also had a tendency to over-shrink some genes, shrinking some truly large effects close to zero. Furthermore, for the real data set, ash performed worse than the MLE for most metrics, including mean absolute error and concordance at the top, most likely due to over-shrinking of many genes. As performance on the real data set was based on taking random subsamples of mice and using the MLE of the held-out set as the truth, estimation error of the held-out effect estimates may have biased results. For future research, using larger data sets to analyze apeglm performance than that of Crowley et al. would allow for held-out sets with more samples and thus reduce estimation error of held-out effect size estimates.\n\nThe shrinkage estimators compared here typically shrunk only low-count genes, as low-count genes tend to be those with the most uncertain and variable estimates. However, during a simulation where extreme overdispersion and heavy tails of the distribution of true effects were introduced, there were some large-count highly-variable genes that were shrunk as well, showing that ash and apeglm will shrink large-count genes if there is high uncertainty in the estimates. Ash consistently had more extreme shrinkage than apeglm and greater estimation error among genes with truly large effects. Thus, ash would most likely perform best in a situation where most effects were small, such as in the Student’s t simulation.\n\nNo method gave confidence or credible intervals with the highest coverage rates for all scenarios. However, across both simulations and analysis of of the mouse data, differences in coverage rates between the three methods were small, and coverage rates for apeglm credible intervals in particular were always very close to the interval that had the largest coverage. Furthermore, interval width for apeglm and ash were always smaller than that of maximum likelihood. This suggests that interval estimates from apeglm could be advantageous over those by maximum likelihood. For future research, it would be beneficial to evaluate the accuracy of hypotheses tests based on the estimates or posterior distribution of apeglm using metrics such as type I and type II error. The method of Leòn-Novelo et al. 20187 rejected hypotheses based on credible intervals of its posterior distribution, and if a similar step was taken for apeglm, its narrower intervals and robust coverage could potentially give more powerful hypothesis tests without suffering from inflated type I error.\n\nOur changes to the apeglm package greatly improved computational performance for both ML and apeglm estimation of beta-binomial GLMs, particularly when larger numbers of covariates were involved. Among the R packages that we looked at which could fit beta-binomial models, the new apeglm package was always the fastest for fitting many GLMs in sequence, e.g. across many genes or variant locations. Thus, the new apeglm package is useful for quick and reliable analyses of allelic imbalance even for researchers who wish to only use likelihood-based estimators. Moreover, only coefficient estimates are currently calculated in C++, and even better computational performance would be achieved if overdispersion and standard error calculations were integrated into C++ as well. We are not aware of any other R packages that utilize faster programming languages such as C or C++ to estimate numerous beta-binomial regression models based on large matrices of observed allelic counts. The most similar package we noted was fastglm27, which fits individual quasi-binomial models in C++. While quasi-binomial models also estimate proportions and control for overdispersion, they do so in a different manner and with different assumptions.\n\nBased on previous work, there are several ways in which the apeglm methodology could potentially be improved for allelic expression studies. For instance, while our extension of apeglm estimated overdispersion by MLE, the original methodology for apeglm as applied to negative binomial GLMs utilized Bayesian estimates for overdispersion as well as for regression coefficients. Introducing a prior for beta-binomial overdispersion that pools information across genes may lead to better estimation and inference of regression coefficients. We also assumed that the total allele-specific counts were fixed and known. Allowing such quantities to be random, as in the method by Leòn-Novelo et al. 2018, may lead to better inference as well. Adjusting for read mapping biases and ambiguities (Leòn-Novelo et al. 20145; Leòn-Novelo et al. 20187; Raghupathy et al. 20183) could also lead to better estimates when such biases and quantification uncertainty are present. Lastly, though here we focused on beta-binomial GLMs, a wide variety of statistical models can be used for ASE, from quasi-binomial28 to Poisson-lognormal models29.\n\n\nData availability\n\nZenodo: RNA-seq Dataset from Crowley et al. 2015. http://doi.org/10.5281/zenodo.340468916.\n\nThis project contains the following underlying data:\n\nfullGeccoRnaDump.csv\n\nThis file contains the Crowley et al. mouse dataset which was was obtained from http://csbio.unc.edu/gecco/data/fullGeccoRnaDump.csv.gz15,30. We uploaded the dataset to Zenodo on the authors’ behalf with their permission, due to the fact that the original dataset is not currently hosted in a stable repository.\n\nThe dataset from this repository is available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).\n\nZenodo: Supplementary Material for Zitovsky and Love 2019. http://doi.org/10.5281/zenodo.340469713.\n\nThis project contains the following extended data:\n\nSupplementary Methods.pdf (Contains the mathematical and algorithmic details of how the apeglm package estimates beta-binomial coefficient effect sizes by maximum likelihood and apeglm, including the steps taken to improve computational performance, increase numerical stability and prevent convergence issues)\n\nSupplementary Figures and Tables.pdf (Contains supplementary figures 1–6 and supplementary tables 1–3. These figures and tables were referenced and described in the main body of the article)\n\nData are available under the terms of the CC-BY 4.0 license.\n\n\nSoftware availability\n\nZenodo: Apeglm v1.7.5 Source Code. http://doi.org/10.5281/zenodo.340450431. This repository contains the source code for the version of the apeglm package used in this paper.\n\nThe software from this repository is available under the terms of the GNU General Public License v3.0 (GPL-3).\n\nZenodo: Source Code for Zitovsky and Love 2019. http://doi.org/10.5281/zenodo.340466932. This repository contains the R scripts used to run the analyses described in this article and generate all of its figures. All figures associated with this paper, including figures present in the main article and supplementary figures, were generated as separate .png and .eps files and can also be found in this repository. The R scripts can be found under the ‘Code’ folder while the figures can be found under the ‘Figures’ folder.\n\nMaterial from this repository are available under the terms of the GPL-3 license.\n\napeglm is available as part of the Bioconductor project33 at http://bioconductor.org/packages/apeglm. The vignette18 and manual provide detailed information on how to use the package.",
"appendix": "Acknowledgements\n\nWe thank Anqi Zhu and Joseph G. Ibrahim of the Department of Biostatistics at UNC Chapel Hill for their contributions to the conceptualization and development of the original apeglm methodology, and Rob Patro for useful discussions.\n\n\nReferences\n\nCastel SE, Levy-Moonshine A, Mohammadi P, et al.: Tools and best practices for data processing in allelic expression analysis. Genome Biol. 2015; 16(1): 195. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSun W, Hu Y: Mapping of Expression Quantitative Trait Loci Using RNA-seq Data. In: Somnath Datta and Dan Nettleton, editors, Statistical Analysis of Next Gen- eration Sequencing Data. Springer International Publishing, Switzerland. 2014; 145–168. Publisher Full Text\n\nRaghupathy N, Choi K, Vincent MJ, et al.: Hierarchical analysis of RNA-seq reads improves the accuracy of allele-specific expression. Bioinformatics. 2018; 34(13): 2177–84. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTurro E, Su SY, Gonçalves Â, et al.: Haplotype and isoform specific expression estimation using multi-mapping RNA-seq reads. Genome Biol. 2011; 12(2): R13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeón-Novelo LG, McIntyre LM, Fear JM, et al.: A flexible Bayesian method for detecting allelic imbalance in RNA-seq data. BMC Genomics. 2014; 15(1): 920. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSkelly DA, Johansson M, Madeoy J, et al.: A powerful and flexible statistical framework for testing hypotheses of allele-specific gene expression from RNA-seq data. Genome Res. 2011; 21(10): 1728–37. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeón-Novelo LG, Gerken AR, Graze RM, et al.: Direct Testing for Allele-Specific Expression Differences Between Conditions. G3 (Bethesda). 2018; 8(2): 447–460. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLove MI, Huber W, Anders S: Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014; 15(12): 550. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLandau W, Niemi J, Nettleton D: Fully Bayesian analysis of RNA-seq counts for the detection of gene expression heterosis. J Am Stat Assoc. 2018; 114(526): 610–621. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStephens M: False discovery rates: a new deal. Biostatistics. 2017; 18(2): 275–94. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhu A, Ibrahim JG, Love MI: Heavy-tailed prior distributions for sequence count data: removing the noise and preserving large differences. Bioinformatics. 2018; 35(12): 2084–2092. PubMed Abstract | Publisher Full Text | Free Full Text\n\nR Core Team: R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. 2018. Reference Source\n\nZitovsky JP, Love MI: Supplementary Material for Zitovsky and Love 2019. (Version v1.0). Zenodo. 2019. http://www.doi.org/10.5281/zenodo.3404697\n\nLu M, Stephens M: Empirical Bayes Estimation of Normal Means, Accounting for Uncertainty in Estimated Standard Errors. 2019; arXiv:1901.10679. Reference Source\n\nCrowley JJ, Zhabotynsky V, Sun W, et al.: Analyses of allele-specific gene expression in highly divergent mouse crosses identifies pervasive allelic imbalance. Nat Genet. 2015; 47(4): 353–360. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCrowley JJ, Zitovsky JP, Love MI: RNA-seq Dataset from Crowley et. al. 2015. (Version v1.0). Zenodo. 2019. http://www.doi.org/10.5281/zenodo.3404689\n\nBolker B: emdbook: Ecological Models and Data in R. In: R package version 1.3.11. 2019. Reference Source\n\nZhu A, Ibrahim JG, Love MI: Effect Size Estimation with Apeglm. Bioconductor. 2019. Reference Source\n\nHimes BE, Jiang X, Wagner P, et al.: RNA-Seq transcriptome profiling identifies CRISPLD2 as a glucocorticoid responsive gene that modulates cytokine function in airway smooth muscle cells. PLoS One. 2014; 9(6): e99625. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIrizarry RA, Warren D, Spencer F, et al.: Multiple-laboratory comparison of microarray platforms. Nat Methods. 2005; 2(5): 345–350. PubMed Abstract | Publisher Full Text\n\nLesnoff M, Lancelot R: aod: Analysis of Overdispersed Data. R package version 1.3.3. 2012.\n\nYee TW: Vector Generalized Linear and Additive Models: With an Implementation in R. R package version 1.1. 2019. Publisher Full Text\n\nLesnoff M, Lancelot R: aods3: Analysis of Overdispersed Data Using S3 Methods. R package version 0.4-1.1. 2018. Reference Source\n\nRigby RA, Stasinopoulos DM: Generalized Additive Models for Location, Scale and Shape. J R Stat Soc C-Appl. 2005; 54(3): 507–54. Publisher Full Text\n\nDae-Jin L, Najera-Zuloaga J, Arostegui I: HRQoL: Health Related Quality of Life Analysis. R package version 1.0. 2017. Reference Source\n\nMersmann O: microbenchmark: Accurate Timing Functions. R package version 1.4-6. 2018.\n\nHuling J: fastglm: Fast and Stable Fitting of Generalized Linear Models using RcppEigen. R package version 0.0.1. 2019. Reference Source\n\nMcVicker G, van de Geijn B, Degner JF, et al.: Identification of genetic variants that affect histone modifications in human cells. Science. 2013; 342(6159): 747–749. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlvarez-Castro I: Bayesian Analysis of High-Dimmensional Count Data. PhD dissertation, Iowa State University. 2017. Publisher Full Text\n\nCrowley JJ, et al.: Gene Expression in the Collaborative Cross. (and Others). 2015. [Data set].\n\nZhu A, Zitovsky J, Ibrahim J, et al.: Apeglm v1.7.5 Source Code (Version v1.0). Zenodo. 2019. http://www.doi.org/10.5281/zenodo.3404504\n\nZitovsky JP, Love MI: Source Code for Zi- tovsky and Love 2019 (Version v1.3). Zenodo. 2019. http://www.doi.org/10.5281/zenodo.3404669\n\nHuber W, Carey VJ, Gentleman R, et al.: Orchestrating high-throughput genomic analysis with Bioconductor. Nat Methods. 2015; 12(2): 115–121. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "57281",
"date": "17 Dec 2019",
"name": "Matthew Stephens",
"expertise": [
"Reviewer Expertise Bayesian statistics",
"statistical genetics"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSummary:\nThe paper presents new implementations of shrinkage methods for beta binomial models, implemented in the R software package apeglm. One potential application of these models is estimating allele-specific biases in various sequencing-based assays (and differences in bias between groups), and the paper focuses on this application. The performance of the shrinkage methods is assessed via simulation and real data analysis (using performance on hold-out data as a performance metric), and the shrinkage methods implemented here are found to be competitive with another shrinkage approach (adaptive shrinkage, ash), and consistently outperform the mle. The new implementations are also shown to be computationally faster than existing implementations (eg aod or the previous version of apeglm).\nThe paper is generally well written, and carefully done, with some exceptions I note later. The new implementations seem likely to be useful in a range of applications. Certainly the use of shrinkage methods in these types of applications is to be encouraged, and I congratulate the authors for leading the way on this. I hope they will find my report helpful in revising their work. I was instructed \"Please indicate clearly which points must be addressed to make the article scientifically sound.\" I believe points 2-4 below are most important to address to make the article scientifically sound.\n1. A note on differences between the shrinkage methods:\nOne thing that I felt was missing from the paper was a qualitative summary of how the two shrinkage methods used here differ from one another. Both are a form of Empirical Bayes shrinkage, but they use different prior families, different likelihoods, and different point estimate strategies: apeglm uses a Cauchy prior, with beta-binomial likelihood, and posterior mode point estimate; whereas ash uses a more flexible unimodal prior (which includes Cauchy as a special case), a normal approximation to the likelihood, and uses a posterior mean point estimate. So the trade-off here is that ash is using an approximate likelihood, but a more flexible prior and arguably a more principled point estimate (posterior mean is optimal under mean squared error).\nI think many readers might benefit from this \"high-level\" summary of the differences.\nAnother important point, which will come up later, is that when using ash the user has a choice of how to make the normal approximation. Specifically ash requires the user to provide point estimates (beta-hat) and standard errors (s-hat), with the goal that beta-hat approx \\sim N(beta, s-hat), where beta is true value that is being estimated. So there is not only one way to apply ash to a problem, but many different ways depending on the choice of point estimate beta-hat. The mle is one natural choice, but in this application there can be problems with infinite mles; see 2. below.\n2. On dealing with infinite mles:\nTo explain the issue with infinite mles, consider first a simple binomial experiment X \\sim Bin(n,p) in which we observe X=0. Then the mle for p is 0, and the mle for theta:=log(p/(1-p)) is -Infinity. Similarly, if X=n the mle for theta is Infinity. Also, in both cases. the standard error for theta is infinite. The same issue arises in the more complex beta-binomial models considered here. Essentially if all the reads in an experiment show the same allele then the mle for the allelic bias parameter (on the logit scale) is +-Infinity. This could happen due to low coverage, but it could also happen at high coverage sites if the allelic bias is very strong.\nThis issue appears to arise in the data analyses used to produce Figure 3 (I did not check whether it arises in the simulations). In Figure 3 there appear many mles (y axis) taking values near +-(5 to 6); however, my brief investigations of the data suggested that most of these likely correspond to genes where all the reads come from one allele, and so the mle is actually +-Infinity as above. (That these infinite mles are computed to be near +-6 is presumably due to an issue with the numerical maximization method used to compute the mle.)\nI suspect that the problems with ash observed in Fig 3 stem from this issue: the mle for these situations where all the reads come from one allele are very unstable, and have a very large standard error (technically infinite, although for numeric reasons finite values are used) and these large standard errors cause these mles to be shrunk excessively.\nA simple fix for this problem, and one I suggest the authors try, is to add a pseudo-count (say 1, or 0.5) to the counts for *each* allele in the data before computing \"mles\" and corresponding standard errors. Pseudo-counts are commonly used to improve stability of mles in this type of situation. Indeed, adding pseudo-counts can be viewed as a simple kind of shrinkage method, so it seems reasonable to compare the more sophisticated EB methods with the simple pseudo-count method. For most genes the point estimates and standard errors will be very little affected by the addition of a small pseudo-count; but for the problematic genes with infinite mle the pseudo-count will stabilize the point estimate and reduce the standard error. I suspect entering the stabilized estimates + standard errors into ash will greatly reduce the problems observed with use of the mles in Figure 3.\n(Incidentally, Xing, Carbonetto and Stephens arXiv:1605.077871 encounter a closely-related issue when using ash to smooth Poisson data; they solved this using a slightly different approach that is conceptually similar to adding a pseudo-count.)\n3. Subsetting results based on shrinkage amounts and \"true\" values:\nIn several places the paper reports error measures on subsets of the results. For example, in Table 1 lines 2-4 involve subsets of results chosen based on the true effect size or shrinkage amount (which depends on the true effect). Although tempting, this type of result is hard to interpret. For example, even the optimal shrinkage rule (i.e. the one that uses the correct prior, likelihood and loss function) may not perform uniformly better than the mle on subsets that are chosen in this way. Thus the sentence on p7 (\"For instance, among genes with effect sizes greater than two...\") may also be true for the optimal shrinkage rule, and so does not constitute direct evidence for \"overshrinkage\". (I agree there is overshrinkage, but this is not the right way to show it). Comparisons like p9 (\"Among genes that were shrunk...\"), which stratify by the amount of shrinkage, have the same problem because the amount of shrinkage depends on the true value and not only on the observed value.\nIt is much cleaner and easier to interpret results if they are subsetted based on the *observed* effect (mle), rather than the true effect. This is because the optimal shrinkage rule is still optimal for *any subset chosen based only on the observed data*. (For this reason you could also subset based on other features of the observed data, like total allele count.) For example, if a method is worse than the mle for the subset of results where the mle is >4 then this is indeed evidence of a problem of some kind.\n4. Computation: speed vs accuracy:\nWhen comparing with other methods/implementations there should be some assessment not only of speed, but of accuracy of the different implementations (meaning the accuracy with which they optimize the log-likelihood, rather than the accuracy of the point estimates). Fast answers are easy if you do not care about accuracy....\nE.g. I suggest boxplots of loglik(method) - loglik(apeglm-new) for each method, to show that the apeglm-new solution is consistently as high in log-likelihood as other methods (or nearly so). Are there convergence criteria decisions to be made that might affect the trade-off between speed and accuracy?\n5. Reproducibility:\nI congratulate the authors on making all their code and data available. After a few tweaks to the code I was able to run the code used to produce Figures 1-3. However, my version of Fig 3 looked different from the one in the paper - my figure had different colors and some points seemed to be missing on my figure. I do not know the reasons for this.\nReproducibility would have been made easier by avoiding the use of absolute file paths. I also suggest not defining functions that operate on global variables (e.g. subsetCalculations = function(sub){..,}) since they are more likely to lead to reproducibility problems.\nI was unable to run the code to perform the computation time comparisons (Figure 4), since it errored out. Again I do not know the reason, but it could be due to differences in the package versions I used compared with the authors. I did not have time to troubleshoot this.\n6. Miscellaneous other comments:\nFor Table 3, I think it should be noted that the coverage probability is expected to be <0.95 because you are looking at how often the interval covers the *estimate* in the larger dataset, and not the *true* value. This makes it a hard to compare the methods here because it isn't clear what the right coverage is.\np12: \"ash would most likely perform best in a situation where most effects were small\". I don't see any evidence for this here (e.g. in the normal simulation ash performs fine) and indeed no reason to expect it to be true a priori. I think this statement should be removed.\n7. Minor comments:\np3: \"When a subject is heterozygous for a gene at a particular SNP\"; this wording seemed awkward to me.\n\np3: \"... making it the most robust and reliable when dealing with small sample sizes\"; this conclusion (\"making it\") seemed not to follow directly from the first part of the sentence.\n\np4: \"Apeglm shrinks the effect of one predictor at a time\": I think this sentence might work better at the start of the paragraph, before specifying the prior used.\n\np5: \"guided by the author's claim\": this is not just a claim, it is a theorem dating back to the 1950s (see original paper for citations).\n\np5: diallel typo?\n\np5: use of beta for the mean of the exponential distribution is confusing as beta is already used elsewhere.\n\np9: \"We also conducted...\" This did not seem worth reporting to me. The difference in sample size (5 vs 5 instead of 4 vs 4) is too small to expect that the results would be very different.\n\np9: In the paragraph \"Both apeglm and MLE...\" the acknowledgement that comparing against CAT in a hold-out set is potentially problematic is a bit buried in the middle of the paragraph. It would seem better to acknowledge this up front. Given the problems with CAT acknowledged here I suggest removing that figure (Fig 3d) or moving to an Appendix.\n\nFigure 5: this should have a y axis that starts at 0.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly",
"responses": [
{
"c_id": "6082",
"date": "14 Dec 2020",
"name": "Josh Zitovsky",
"role": "Author Response",
"response": "Summary: The paper presents new implementations of shrinkage methods for beta binomial models, implemented in the R software package apeglm. One potential application of these models is estimating allele-specific biases in various sequencing-based assays (and differences in bias between groups), and the paper focuses on this application. The performance of the shrinkage methods is assessed via simulation and real data analysis (using performance on hold-out data as a performance metric), and the shrinkage methods implemented here are found to be competitive with another shrinkage approach (adaptive shrinkage, ash), and consistently outperform the mle. The new implementations are also shown to be computationally faster than existing implementations (eg aod or the previous version of apeglm). The paper is generally well written, and carefully done, with some exceptions I note later. The new implementations seem likely to be useful in a range of applications. Certainly the use of shrinkage methods in these types of applications is to be encouraged, and I congratulate the authors for leading the way on this. I hope they will find my report helpful in revising their work. I was instructed \"Please indicate clearly which points must be addressed to make the article scientifically sound.\" I believe points 2-4 below are most important to address to make the article scientifically sound. Thank you for your constructive comments and careful evaluation of our software and analysis. We found your report helpful and have tried our best to address all of your concerns. Point-by-point responses are provided below. 1. A note on differences between the shrinkage methods: One thing that I felt was missing from the paper was a qualitative summary of how the two shrinkage methods used here differ from one another. Both are a form of Empirical Bayes shrinkage, but they use different prior families, different likelihoods, and different point estimate strategies: apeglm uses a Cauchy prior, with beta-binomial likelihood, and posterior mode point estimate; whereas ash uses a more flexible unimodal prior (which includes Cauchy as a special case), a normal approximation to the likelihood, and uses a posterior mean point estimate. So the trade-off here is that ash is using an approximate likelihood, but a more flexible prior and arguably a more principled point estimate (posterior mean is optimal under mean squared error). I think many readers might benefit from this \"high-level\" summary of the differences. Another important point, which will come up later, is that when using ash the user has a choice of how to make the normal approximation. Specifically ash requires the user to provide point estimates (beta-hat) and standard errors (s-hat), with the goal that beta-hat approx \\sim N(beta, s-hat), where beta is true value that is being estimated. So there is not only one way to apply ash to a problem, but many different ways depending on the choice of point estimate beta-hat. The mle is one natural choice, but in this application there can be problems with infinite mles; see 2. Below. We agree that there are important methodological differences between the methods, and that a high-level summary of these differences would be beneficial to the readers. We have added a paragraph highlighting these differences in the second-to-last paragraph of the “Estimation methods” subsection of the “Methods” section. Among other differences, we highlight the increased flexibility of ash’s prior and its ability to handle non-ML estimators. Additional details regarding the methodology of these methods have also been added to the sections where apeglm and ash were initially introduced. 2. On dealing with infinite mles: To explain the issue with infinite mles, consider first a simple binomial experiment X \\sim Bin(n,p) in which we observe X=0. Then the mle for p is 0, and the mle for theta:=log(p/(1-p)) is -Infinity. Similarly, if X=n the mle for theta is Infinity. Also, in both cases. the standard error for theta is infinite. The same issue arises in the more complex beta-binomial models considered here. Essentially if all the reads in an experiment show the same allele then the mle for the allelic bias parameter (on the logit scale) is +-Infinity. This could happen due to low coverage, but it could also happen at high coverage sites if the allelic bias is very strong. This issue appears to arise in the data analyses used to produce Figure 3 (I did not check whether it arises in the simulations). In Figure 3 there appear many mles (y axis) taking values near +-(5 to 6); however, my brief investigations of the data suggested that most of these likely correspond to genes where all the reads come from one allele, and so the mle is actually +-Infinity as above. (That these infinite mles are computed to be near +-6 is presumably due to an issue with the numerical maximization method used to compute the mle.) I suspect that the problems with ash observed in Fig 3 stem from this issue: the mle for these situations where all the reads come from one allele are very unstable, and have a very large standard error (technically infinite, although for numeric reasons finite values are used) and these large standard errors cause these mles to be shrunk excessively. A simple fix for this problem, and one I suggest the authors try, is to add a pseudo-count (say 1, or 0.5) to the counts for *each* allele in the data before computing \"mles\" and corresponding standard errors. Pseudo-counts are commonly used to improve stability of mles in this type of situation. Indeed, adding pseudo-counts can be viewed as a simple kind of shrinkage method, so it seems reasonable to compare the more sophisticated EB methods with the simple pseudo-count method. For most genes the point estimates and standard errors will be very little affected by the addition of a small pseudo-count; but for the problematic genes with infinite mle the pseudo-count will stabilize the point estimate and reduce the standard error. I suspect entering the stabilized estimates + standard errors into ash will greatly reduce the problems observed with use of the mles in Figure 3. (Incidentally, Xing, Carbonetto and Stephens arXiv:1605.077871 encounter a closely-related issue when using ash to smooth Poisson data; they solved this using a slightly different approach that is conceptually similar to adding a pseudo-count.) As you suspected, there were indeed genes with “truly infinite” MLEs, but due to numerical reasons, were given finite estimates by the apeglm package. As you suggested, we have now performed additional analyses adding a pseudocount to each allele prior to computing MLEs, and compared the performance of the resulting ML, apeglm and ash estimates to those not involving pseudocounts for the simulations. We also attempted to remove the infinite ML genes prior to analysis. Results can be found in Table 1, Table 2 and Supplementary Figure 3. 3. Subsetting results based on shrinkage amounts and \"true\" values: In several places the paper reports error measures on subsets of the results. For example, in Table 1 lines 2-4 involve subsets of results chosen based on the true effect size or shrinkage amount (which depends on the true effect). Although tempting, this type of result is hard to interpret. For example, even the optimal shrinkage rule (i.e. the one that uses the correct prior, likelihood and loss function) may not perform uniformly better than the mle on subsets that are chosen in this way. Thus the sentence on p7 (\"For instance, among genes with effect sizes greater than two...\") may also be true for the optimal shrinkage rule, and so does not constitute direct evidence for \"overshrinkage\". (I agree there is overshrinkage, but this is not the right way to show it). Comparisons like p9 (\"Among genes that were shrunk...\"), which stratify by the amount of shrinkage, have the same problem because the amount of shrinkage depends on the true value and not only on the observed value. It is much cleaner and easier to interpret results if they are subsetted based on the *observed* effect (mle), rather than the true effect. This is because the optimal shrinkage rule is still optimal for *any subset chosen based only on the observed data*. (For this reason you could also subset based on other features of the observed data, like total allele count.) For example, if a method is worse than the mle for the subset of results where the mle is >4 then this is indeed evidence of a problem of some kind. Shrinkage in the first manuscript was defined as the movement of apeglm and ash estimates from the MLE toward zero. As apeglm, ash and ML estimates are all functions of the observed data, the degree of shrinkage is also a function of observed data and thus we felt that subsetting by shrinkage was valid. However, we do agree with your concern that subsetting by true effect sizes may cause difficulty in contrasting procedures with each other with respect to the optimal shrinkage rule, and thus have removed results of mean absolute error stratified by the true effect sizes. Moreover, per your suggestion, we have added stratification of MAE by total gene counts and MLE magnitude. We also added MA plots, which illustrates how the amount of shrinkage differs by total gene counts and MLE size (these plots were previously in the Supplemental Material, but have been moved to the main paper). 4. Computation: speed vs accuracy: When comparing with other methods/implementations there should be some assessment not only of speed, but of accuracy of the different implementations (meaning the accuracy with which they optimize the log-likelihood, rather than the accuracy of the point estimates). Fast answers are easy if you do not care about accuracy.... E.g. I suggest boxplots of loglik(method) - loglik(apeglm-new) for each method, to show that the apeglm-new solution is consistently as high in log-likelihood as other methods (or nearly so). Are there convergence criteria decisions to be made that might affect the trade-off between speed and accuracy? We agree that an assessment of numerical accuracy is important in showcasing our package, and have adding such assessments in the new version of the manuscript. We focused our analysis of numerical accuracy on genes such that the difference in an estimated coefficient between apeglm and the other packages were non-negligible (above 0.01), and among those genes reported the differences in log-likelihood. A high-level overview of the results is present in the last paragraph of the “Computational Performance of Apeglm” subsection of the “Results” section, and a detailed summary of the results was added to the Supplementary Methods section. Overall, we found that our package is, in addition to its estimation speed, also numerically accurate. 5. Reproducibility: I congratulate the authors on making all their code and data available. After a few tweaks to the code I was able to run the code used to produce Figures 1-3. However, my version of Fig 3 looked different from the one in the paper - my figure had different colors and some points seemed to be missing on my figure. I do not know the reasons for this. Reproducibility would have been made easier by avoiding the use of absolute file paths. I also suggest not defining functions that operate on global variables (e.g. subsetCalculations = function(sub){..,}) since they are more likely to lead to reproducibility problems. I was unable to run the code to perform the computation time comparisons (Figure 4), since it errored out. Again I do not know the reason, but it could be due to differences in the package versions I used compared with the authors. I did not have time to troubleshoot this. We apologize for the reproducibility issues in the first version of the paper. Briefly, the issues you reported stemmed from two underlying causes: 1) the version of the apeglm package in the devel branch at the time of publication did not match the version used in the manuscript; 2) we accidentally uploaded the wrong scripts to Zenodo. We have now correctly identified the apeglm package version in the manuscript (v1.11.2) and replaced the scripts in Zenodo with the correct ones. All scripts should now run without issues and output the same numbers and plots as shown in the paper. Moreover, we have removed absolute file paths and do not use global variables in our functions (some of the local variables defined within functions might share names with global variables created later on, but our functions no longer call global variables directly). 6. Miscellaneous other comments: For Table 3, I think it should be noted that the coverage probability is expected to be <0.95 because you are looking at how often the interval covers the *estimate* in the larger dataset, and not the *true* value. This makes it a hard to compare the methods here because it isn't clear what the right coverage is. Due to concerns posed by yourself and other reviewers, we have completely rewritten our analysis of real data to focus on more qualitative results, and have mostly left evaluations of accuracy to the simulations, where the true simulation parameters are known. Among other changes, we do not evaluate or assess coverage probabilities of estimators when analyzing the real data. p12: \"ash would most likely perform best in a situation where most effects were small\". I don't see any evidence for this here (e.g. in the normal simulation ash performs fine) and indeed no reason to expect it to be true a priori. I think this statement should be removed. We have removed this statement. 7. Minor comments: p3: \"When a subject is heterozygous for a gene at a particular SNP\"; this wording seemed awkward to me. We have changed the wording to “When a subject is heterozygous at a particular SNP within an exon of a gene” p3: \"... making it the most robust and reliable when dealing with small sample sizes\"; this conclusion (\"making it\") seemed not to follow directly from the first part of the sentence. We have changed this from “the most robust and reliable” to just “robust and reliable”. p4: \"Apeglm shrinks the effect of one predictor at a time\": I think this sentence might work better at the start of the paragraph, before specifying the prior used. We have made the suggested change. p5: \"guided by the author's claim\": this is not just a claim, it is a theorem dating back to the 1950s (see original paper for citations). Apologies for the confusion. We have changed it from “guided by the author’s claim” to “guided by the fact” and have cited both ash and the original 1950’s citation. p5: diallel typo? In our original manuscript, we had the term “diallel cross”, we did not find a typo. p5: use of beta for the mean of the exponential distribution is confusing as beta is already used elsewhere. We changed the notation for the mean parameter from beta to mu. p9: \"We also conducted...\" This did not seem worth reporting to me. The difference in sample size (5 vs 5 instead of 4 vs 4) is too small to expect that the results would be very different. We have removed this result. p9: In the paragraph \"Both apeglm and MLE...\" the acknowledgement that comparing against CAT in a hold-out set is potentially problematic is a bit buried in the middle of the paragraph. It would seem better to acknowledge this up front. Given the problems with CAT acknowledged here I suggest removing that figure (Fig 3d) or moving to an Appendix. Please see our response to your concerns in point #6. Figure 5: this should have a y axis that starts at 0. Unfortunately, the y-axis for figure 5 of the initial version of the paper (renamed Figure 8 in version 2) is on the log-scale, which means we cannot start it at zero. Using a log scale is necessary due to the very different computational times of the apeglm and aod packages and the difference in how well they scale with increasing numbers of covariates. We considered changing the figure to start the y-axis at a smaller positive number (eg 10, 1, 0.1 etc.) but we ultimately decided against this as the exact cut-point at which to start the y-axis would have been arbitrary and there would have been a large amount of unnecessary white space between the plots and the x-axis (due to the fact that the y-axis is measured on the log scale)."
}
]
},
{
"id": "58251",
"date": "04 Feb 2020",
"name": "Jarad Niemi",
"expertise": [
"Reviewer Expertise Bayesian statistics"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIs the rationale for developing the new method (or application) clearly explained?\nYes.\nIn our work a key issue is bias of allele reads toward a reference genome as explained in Sun and Hu (2014).1 The authors should mention if this bias is relevant for the applications in this manuscript and, if yes, how the methods deal with the bias.\n\nThe introduction argues against eliminating low count genes, yet the manuscript says \"Genes where at least three samples did not have at least 10 counts were removed...Genes without at least one count for both alleles across all individuals were removed...Genes with a marginally significant sex or parent effect were removed.\" Why the contradiction?\n\nIs the description of the method technically sound?\nNo.\nWhile the writing is clear, we generally found the order of content confusing. For example, normal-based CI construction should be explained immediately after point estimation and before competing methods, simulation details, and method comparison metrics. We also found there was a lack of details, some of which was in the Supplementary Material but seemed like it should be included in the main manuscript.\n\nIn addition, we have outlined concerns below:\nMajor concerns:\nIt isn't clear how MAE or CI coverage are calculated for the real data. For real data the truth is not known and therefore MAE and coverage cannot be calculated the way they can for the simulated data. Are you calculating MAE and coverage relative to the data? You comment \"we are treating the MLE of the held-out set as the truth\". Why? The simulation studies seemed to show this is a relatively poor estimate of the truth.\n\nMinor concerns:\nPlease provide some statements for why a beta-binomial model is assumed as opposed to alternative model assumptions, e.g. binomial, normal, Poisson.\n\nWe assume you are assume conditional independence in your beta-binomial likelihood and in your Cauchy distribution for the regression coefficients. If so, this should be stated explicitly, e.g. using \"ind\" above the tilde.\n\nHow often is \\phi_g estimated to be 500? How important is the value 500? Is this user specifiable in the package?\n\nIt is unclear what is meant by \"standard error\" in the statement \"apeglm provides Bayesian shrinkage estimates based on the mode of the posterior as well as standard errors.\" Is this the posterior standard deviation? Is it the (asymptotic) standard deviation of the estimator?\n\nThe manuscript states \"The scale parameter of the Cauchy prior, \\gamma_j, is estimated by pooling information across genes\". How exactly is this computed?\n\nIt seems odd to have the Supplementary Material on a site other than F1000. We're disappointed that the Estimation Procedure in the Supplementary Material is not included in the main body of the manuscript as this seems to be key to the methodology. If not included in the main manuscript, perhaps more specific references, say to equation numbers, could be included in the main manuscript.\n\nWe don't understand the statement \"Like apeglm, ash can only shrink estimates for one covariate at a time.\" Isn't the assumed hierarchical distribution a joint hierarchical distribution, albeit assuming independence, for all regression coefficients? If so, then isn't it jointly shrinking all the estimates? Or is the procedure a step-wise procedure where MLEs are shrunk one-at-time?\n\nIt is unclear why a Cauchy distribution is chosen. While a Cauchy distribution has the appealling property that it does not shrink large signals (very much), it generally does little shrinkage to small signals compared to alternative estimators, e.g. Bayesian LASSO (10.1198/016214508000000337,10.1093/biomet/asp047)2,3, horseshoe (10.1093/biomet/asq017)4, point-mass priors (10.1080/01621459.1993.10476353)5. In our applications, the true distribution of these regression coefficients often has a large spike around 0 which would suggest using a distribution with more mass than a Cauchy near 0.\n\nThe statement \"where 1 <= j <= K is chosen by the user\" is confusing. Does the user specify which predictors have a Cauchy distributions? What exactly is the user choosing?\n\nAre sufficient details provided to allow replication of the method development and its use by others?\nPartly.\nOne reason to provide code and data are to ensure ability to replicate even if the text is insufficient. So, ensuring the code is able to be run will provide sufficient details.\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility?\nYes.\nWe also applaud the authors for making their code and data available.\nReviewer 1 addressed this and we did not attempt to evaluate this further.\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article?\nPartly.\nIn the abstract, the article claims:\n\"Apeglm consistently performed better than ML[E] according to a variety of criteria, including mean absolute error (MAE) and concordance at the top (CAT).\"\nTable 1 and 2 provide supporting evidence for the claim that apeglm has lower MAE than MLE for a variety of simulation scenarios.\nFigures 1d and 2d shows apeglm and ash having similar CAT and ahead of the non-filtered MLE approach.\n\nIt might be helpful to point out that ash, another shrinkage estimator, also consistently performs better than the MLE.\n\n\"While ash had lower error and greater concordance than ML on the simulations, it also had a tendency to over-shrink large effects, and performed worse on the real data according to error and concordance.\"\nWe guess Figures 1a-c and 2a-c as well as line 4 in Table 1 were the evidence for this comment, but we find these figures extremely hard to interpret. The comment in the text is that \"some genes with estimates close to the truth were severely shrunk, and several genes with truly large effects were shrunk to zero.\", but it isn't clear that this is undesirable. Just because the truth is non-zero doesn't mean that the data randomly generated from this truth should suggest a non-zero result.\nWith this being said, we would not be surprised about ash shrinking large signals more than apeglm since the Cauchy distribution (used in apeglm) will shrink large signals less than a normal distribution (used in ash) will, but, as Reviewer 1 points out, there are differences in likelihood and estimation procedure between these two methods which make understanding why differences occur more difficult.\n\n\"2hen compared to five other packages that also fit beta-binomial models, the apeglm package was substantially faster, making our package useful for quick and reliable analyses of allelic imbalance.\"\nFigure 4 provides the computational cost comparison and seems to show that apeglm is faster than aod, aods3, gamlss, HRQoL, and VGAM under the tested scenario. An alternative version of this figure would provide the ratio of runtimes for these other methods compared to apeglm. While the current version allows for an understanding of the computation time involved, the main purpose of the figure is in comparison of times.\nIt does seem a bit odd that the authors compared these packages for computation but not for accuracy. In addition, why is ash not included in this comparison?\n\nOther:\nMinor issues:\nOnce you've defined an acronym, just use it, e.g. CAT.\n\nBe consistent with acronyms: choose ML or MLE and stick with it.\n\nFigure 5 seems unnecessary since an argument in this manuscript is to use \"shrinkage\" estimators rather than un-shrunk MLEs.\n\nAn updated reference for 29. Alvarez-Castro is 10.3934/mbe.20193896\n\nThe beta-binomial is a discrete random variable and thus it has a probability mass function rather than a probability density function.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? No\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly",
"responses": [
{
"c_id": "6083",
"date": "14 Dec 2020",
"name": "Josh Zitovsky",
"role": "Author Response",
"response": "Is the rationale for developing the new method (or application) clearly explained? Yes. In our work a key issue is bias of allele reads toward a reference genome as explained in Sun and Hu (2014).1 The authors should mention if this bias is relevant for the applications in this manuscript and, if yes, how the methods deal with the bias. Reference allele bias is indeed a potential problem when dealing with allelic counts from RNA-seq. However, the methods we benchmark in the manuscript cannot directly deal with such bias. Our simulation does not involve reference allele bias, and the RNA-seq study we examine took specific measures to avoid reference allele bias. We apologize for not clarifying this before and have added a paragraph at the end of the Introduction explaining these points. The introduction argues against eliminating low count genes, yet the manuscript says \"Genes where at least three samples did not have at least 10 counts were removed...Genes without at least one count for both alleles across all individuals were removed...Genes with a marginally significant sex or parent effect were removed.\" Why the contradiction? When filtering is done to remove genes with a high variance estimated allelic ratio, it is usually done with a threshold greater than e.g. 10 total counts per gene / one count per allele. Increased filtering may result in a loss of statistical power, when the optimal filtering rule is not known. Our minimal filtering was performed such that the metrics (e.g. error and ranking concordance) represent features for which there is some minimally detectable signal across alleles. Removing genes with a significant sex or parent effect was done for the purposes of performance analysis, as our analysis involved fitting intercept-only models. We did not want the extra variability induced from sex and/or parent effects in the set of genes used for evaluation. Is the description of the method technically sound? No. While the writing is clear, we generally found the order of content confusing. For example, normal-based CI construction should be explained immediately after point estimation and before competing methods, simulation details, and method comparison metrics. We also found there was a lack of details, some of which was in the Supplementary Material but seemed like it should be included in the main manuscript. We have moved the description of how the methods compute CIs as suggested. Moreover, we have added additional details about the estimation methods in both the main manuscript (under the “Estimation Methods” subsection of the “Methods” section) and the Supplementary Material. For example, in the main manuscript, we added more details regarding apeglm’s likelihood and prior, estimation of the overdispersion and qualitative differences between apeglm’s and ash’s methodologies. In the Supplemental material, we added more details regarding estimation of the overdispersion, estimation of the scale of the Cauchy prior and the numerical accuracy of our package. In addition, we have outlined concerns below: Major concerns: It isn't clear how MAE or CI coverage are calculated for the real data. For real data the truth is not known and therefore MAE and coverage cannot be calculated the way they can for the simulated data. Are you calculating MAE and coverage relative to the data? You comment \"we are treating the MLE of the held-out set as the truth\". Why? The simulation studies seemed to show this is a relatively poor estimate of the truth. We thank the reviewer for noting this drawback in our initial submission. Initially, our choice to use the MLE of the held-out set as the truth came from the fact that the ML estimators are consistent and asymptotically efficiency estimators of the regression parameters, and thus if the held-out sets are sufficiently large, the ML estimates will be very close to the truth. However, the held-out set only consists of 18 samples, which in practice may be too small to be useful. We agree with your concerns that many of the same problems of ML estimators that we address in our manuscript, such as instability in the presence of low information, would still be present in the held-out sets. After thinking about this more and conducting additional analysis, we came to the conclusion that even when using as many as 24 samples, the ML estimates are not close enough to the truth for some genes and using them as the truth may bias results. As a result, we have rewritten the real data analysis section to focus on qualitative assessments that do not require knowledge of the truth, such as differences in nature and extent of shrinkage between apeglm and ash and on estimation variance. Accuracy assessments have been largely left to simulations, where the true parameter values are known. Relatedly, we have changed the simulations so that the intercept is simulated from a standard normal distribution, as opposed to being drawn from ML estimates of intercept-only models fit to the genes of the real data set. The reason for this is similar: we have no reason to believe that the intercept ML estimates are close to the true intercepts, and upon investigation, we found that the distribution of ML estimates had several properties that would not realistically be demonstrated by a distribution of true effect sizes. Minor concerns: Please provide some statements for why a beta-binomial model is assumed as opposed to alternative model assumptions, e.g. binomial, normal, Poisson. We have added a justification for choosing a beta-binomial distribution to model allelic counts in the second paragraph of the “Estimation Methods” subsection of the “Methods” section. We assume you are assume conditional independence in your beta-binomial likelihood and in your Cauchy distribution for the regression coefficients. If so, this should be stated explicitly, e.g. using \"ind\" above the tilde. We have made the suggested changes to the notation so that the assumed conditional independence is clearer How often is \\phi_g estimated to be 500? How important is the value 500? Is this user specifiable in the package? It is difficult to give an exact frequency, as how often phi is estimated at 500 varies from dataset to dataset. The number of genes in a dataset where no or very little overdispersion is exhibited by the allelic proportions (conditional on the covariates) is roughly the number of times at which phi will be estimated at 500 for the dataset. As phi approaches infinity, the resulting regression parameter MLEs converge to the MLEs from a binomial distribution. We found that with phi=500, the ML estimates are already quite close to the ML estimates from a model with assumption of a binomial distribution, and setting the maximum above 500 led to only very small differences in the coefficients. However, the user can specify a different maximum (and minimum) than that used in this package as desired. Details have been added to the main manuscript and Supplemental Methods regarding our chosen minimum and maximum. It is unclear what is meant by \"standard error\" in the statement \"apeglm provides Bayesian shrinkage estimates based on the mode of the posterior as well as standard errors.\" Is this the posterior standard deviation? Is it the (asymptotic) standard deviation of the estimator? It is the posterior standard deviation. We clarified this in the second version. The manuscript states \"The scale parameter of the Cauchy prior, \\gamma_j, is estimated by pooling information across genes\". How exactly is this computed? We have added this information in the Supplemental Material section It seems odd to have the Supplementary Material on a site other than F1000. We're disappointed that the Estimation Procedure in the Supplementary Material is not included in the main body of the manuscript as this seems to be key to the methodology. If not included in the main manuscript, perhaps more specific references, say to equation numbers, could be included in the main manuscript. All references to the Supplemental Material have been made more specific, and are now references to the specific section of the Supplemental Material that is relevant. We don't understand the statement \"Like apeglm, ash can only shrink estimates for one covariate at a time.\" Isn't the assumed hierarchical distribution a joint hierarchical distribution, albeit assuming independence, for all regression coefficients? If so, then isn't it jointly shrinking all the estimates? Or is the procedure a step-wise procedure where MLEs are shrunk one-at-time? We apologize if this was not clear in the first version of the manuscript and have added clarifications in the new version of the manuscript and Supplemental Material. In summary, apeglm for allelic counts assumes a Beta-binomial likelihood for all regression coefficients, but it only assumes a Cauchy prior for one regression coefficient at a time (more specifically, the regression coefficients for only one covariate, across all genes). Thus only one covariate is being “shrunk” at a time. If Bayesian shrinkage of two coefficients was desired (for example), you would have to run apeglm twice: the first time choosing one coefficient, and the second time choosing the other. It is unclear why a Cauchy distribution is chosen. While a Cauchy distribution has the appealling property that it does not shrink large signals (very much), it generally does little shrinkage to small signals compared to alternative estimators, e.g. Bayesian LASSO (10.1198/016214508000000337,10.1093/biomet/asp047)2,3, horseshoe (10.1093/biomet/asq017)4, point-mass priors (10.1080/01621459.1993.10476353)5. In our applications, the true distribution of these regression coefficients often has a large spike around 0 which would suggest using a distribution with more mass than a Cauchy near 0. Our choice of a Cauchy prior was guided by the fact that a Cauchy prior tends to shrink large effect sizes less than other priors, and in a differential expression context was shown to produce estimates with lower error and better ranking be size than competing estimators (see reference 11). We agree that there are situations where a Cauchy prior would not be ideal, if sparsity of estimated coefficients (setting to exactly zero for certain genes) was desired for selection purposes. However apeglm follows and cites the ashr publication in providing the false sign rate (FSR) as a criterion for gene selection. A justification of our choice of a Cauchy prior and the flexibility of our software to handle other priors has also been added into the manuscript. The statement \"where 1 <= j <= K is chosen by the user\" is confusing. Does the user specify which predictors have a Cauchy distributions? What exactly is the user choosing? This is exactly right: The user is specifying which predictor (singular) is assumed to follow a Cauchy distribution for the purpose of shrinkage estimation. We have tried to make this clearer in the second version of the manuscript. See two responses above. Are sufficient details provided to allow replication of the method development and its use by others? Partly. One reason to provide code and data are to ensure ability to replicate even if the text is insufficient. So, ensuring the code is able to be run will provide sufficient details. We apologize for the reproducibility issues present in the first part of the paper. A detailed explanation of the problems and our fixes was given in our responses to the first reviewer. We believe all previous issues have been fixed and the code should now run without problems (assuming all of the relevant packages are installed and the right package versions are being used). If any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes. We also applaud the authors for making their code and data available. Reviewer 1 addressed this and we did not attempt to evaluate this further. Please see our response to your concern under “Are sufficient details provided to allow replication of the method development and its use by others?”. Are the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly. In the abstract, the article claims: \"Apeglm consistently performed better than ML[E] according to a variety of criteria, including mean absolute error (MAE) and concordance at the top (CAT).\" Table 1 and 2 provide supporting evidence for the claim that apeglm has lower MAE than MLE for a variety of simulation scenarios. Figures 1d and 2d shows apeglm and ash having similar CAT and ahead of the non-filtered MLE approach. It might be helpful to point out that ash, another shrinkage estimator, also consistently performs better than the MLE. Due to changes in the simulations (see our response to your “Major Concern” under “Is the description of the method technically sound?”), ash no longer performs better than maximum likelihood universally, though in general it still performs better. The abstract has been changed to accommodate the different results. We believe that our new abstract provides a succinct yet comprehensive and accurate summary of the new results. \"While ash had lower error and greater concordance than ML on the simulations, it also had a tendency to over-shrink large effects, and performed worse on the real data according to error and concordance.\" We guess Figures 1a-c and 2a-c as well as line 4 in Table 1 were the evidence for this comment, but we find these figures extremely hard to interpret. The comment in the text is that \"some genes with estimates close to the truth were severely shrunk, and several genes with truly large effects were shrunk to zero.\", but it isn't clear that this is undesirable. Just because the truth is non-zero doesn't mean that the data randomly generated from this truth should suggest a non-zero result. With this being said, we would not be surprised about ash shrinking large signals more than apeglm since the Cauchy distribution (used in apeglm) will shrink large signals less than a normal distribution (used in ash) will, but, as Reviewer 1 points out, there are differences in likelihood and estimation procedure between these two methods which make understanding why differences occur more difficult. Reviewer 1 voiced similar concerns, and you can see our detailed response to this concern in our responses to the first reviewer. To summarize, we have removed results of mean absolute error stratified by the true effect sizes. We also look more at subsets chosen based only on observed data (e.g. total allele counts and MLE size) to interpret results. We hope our new results are easier to interpret and our conclusions more convincing. \"When compared to five other packages that also fit beta-binomial models, the apeglm package was substantially faster, making our package useful for quick and reliable analyses of allelic imbalance.\" Figure 4 provides the computational cost comparison and seems to show that apeglm is faster than aod, aods3, gamlss, HRQoL, and VGAM under the tested scenario. An alternative version of this figure would provide the ratio of runtimes for these other methods compared to apeglm. While the current version allows for an understanding of the computation time involved, the main purpose of the figure is in comparison of times. It does seem a bit odd that the authors compared these packages for computation but not for accuracy. In addition, why is ash not included in this comparison? We have changed the Figure as suggested to better illustrate relative performance of the other packages compared to apeglm. Moreover, we have added comparisons of numerical accuracy to the main manuscript (last paragraph of “Computational performance of apeglm” subsection) and Supplemental Material. Our package is more numerically accurate and reliable than other packages compared. As to why ash is not included in the comparison, this is because ash requires a vector of initial parameter estimates and standard error estimates, and thus to use ash as we do in the manuscript, one has to perform ML estimation first, and then use ash to shrink the estimates. Comparing ash to apeglm or the ML-fitting packages would thus not be a same-to-same comparison. Other: Minor issues: Once you've defined an acronym, just use it, e.g. CAT. We have made the suggested changes to the manuscript. Be consistent with acronyms: choose ML or MLE and stick with it. We have made the suggested changes to the manuscript. Figure 5 seems unnecessary since an argument in this manuscript is to use \"shrinkage\" estimators rather than un-shrunk MLEs. Though our previous analysis showed that apeglm has higher accuracy than ML estimators, there are still reasons why one would prefer likelihood-based beta-binomial GLMs, such as if the sample size is large or if simplicity or unbiasedness is desired. Moreover, many shrinkage estimation packages like ash require a vector of initial ML estimates and standard errors. Finally, apeglm estimation is almost as fast as ML estimation when using the new apeglm package, and thus Figure 5 would be practically the same if we were to compare other packages to apeglm estimation speed instead. We have added this clarification in the “Computational performance of Apeglm” subsection of the “Results” section. An updated reference for 29. Alvarez-Castro is 10.3934/mbe.20193896 The reference has been updated. The beta-binomial is a discrete random variable and thus it has a probability mass function rather than a probability density function. In the new manuscript, we refer to the probability function of the beta-binomial as its “probability mass function” as opposed to a “density function”"
}
]
},
{
"id": "57280",
"date": "10 Feb 2020",
"name": "Ernest Turro",
"expertise": [
"Reviewer Expertise Biostatistics",
"genomics."
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper has two components:\n1) An advance in computational efficiency for estimating beta-binomial regression coefficients with shrinkage. The authors have produced a C++ implementation of the inference code previously written in R. Both versions of the code are implemented in the apeglm R package.\n2) An application of this new implementation of their method to the task of inferring allele-specific expression (ASE) and an assessment of its statistical performance in relation to two alternative approaches (ash and MLE).\nAs the authors start the paper by discussing ASE, rather than computational inference for shrinkage models, it is not immediately apparent that the innovation presented in this paper is computational rather than statistical. Distinguishing these two components clearly would make it more readily apparent that the paper does not present a novel statistical method.\nThe modelling of ASE has important facets that the authors do not discuss in the introduction (page 3) but which other (uncited) methods have addressed. For example, in a given sample, a gene may contain multiple heterozygous variants (potentially with uncertain phasing of alleles). Each heterozygous variant could overlap different sets of isoforms, each of which may have different levels of ASE. This phenomenon is modelled by the MMDIFF method (Turro et al, 2014, Bioinformatics1), for example. The authors should acknowledge this (unmodelled) complication in ASE and explain how they summarise allele-specific count data across multiple variants (e.g., SNPs or indels, which are possibly unphased) within genes to obtain the count pairs modelled by the beta-binomial shrinkage estimators.\nThe authors have performed several simulation studies and an analysis of a real ASE dataset. Both shrinkage estimators outperform MLE in the simulation studies. However, apeglm and MLE do approximately equally well in the real data set and both outperform ash by a significant margin. In addition, filtering of genes with low allele-specific read counts improves the MLE in the simulation studies but it does not do so in the real data analysis. This discordance demonstrates that the real data are very dissimilar from the simulated data. Although I don't think a major rewrite is warranted, if the authors could demarcate the computational advance (which can be demonstrated by simulation studies that are not representative of ASE, as the authors have done) from the specific application to ASE (using a real data set and perhaps a more faithful simulation study), the striking difference in performance shown in Figures 1-3 would be less incongruous.\nIn the introduction, the inability of other methods to model the effects of continuous covariates or estimate differences in allelic imbalance between groups (this is not the case though, see MMDIFF) is highlighted and contrasted with the proposed method. However, the authors' own analysis of real data only uses an intercept model. It would be desirable to demonstrate the flexibility afforded by the proposed approach.\nIn the assessment of statistical performance using the real data set, the MLEs obtained from the held-out data are treated as truth, even though earlier in the paper the authors demonstrate that MLEs have a particularly high mean absolute error. Presumably, this is the case (for genes with relatively low counts) even when the sample size is 18. The authors should consider alternative measures of performance that do not have this drawback.\nMinor comments:\np3: \"estimates for allelic expression proportions can be highly variable\" - estimates are fixed, the authors should write \"estimators\".\n\np3: a cancer dataset may not be the best choice of example to refer to the proportion of genes with allele-specific reads, due to the prevalence of somatic mutations.\n\np3: when discussing filtering as a \"remedy\" perhaps explain that this achieves a boost in specificity at the cost of power.\n\np3: \"the most robust and reliable when dealing with small sample sizes\" - this part of the sentence does not follow from the previous part, as there is no mention of ash's inadequacy.\n\np3: \"also introduced new source code\" - it is not clear what the \"also\" refers to.\n\np4: \"the probability that counts for a particular gene belong to a particular allele\" should be changed to \"the probability that a read for a particular gene belongs to a particular allele\" as the total \"counts\" will not be assigned to an allele as a block (the total counts derive from a heterogeneous mixture of reads from the two different alleles).\n\np4: more information should be given about how the scale parameter of the Cauchy prior is \"estimated by pooling information across genes\".\n\np4: the placement of the \\cdot indexing the bold face beta is unusual, as the j subscript corresponds to the first rather than the second index.\n\np9: rerunning the simulation study with 4 v 4 samples having run it with 5 v 5 samples seems unnecessary, as such a small change in sample size is unlikely to alter the conclusions.\n\np9: \"Figure 1d\" should read \"Figure 3d\".\n\nIs the rationale for developing the new method (or application) clearly explained? Partly\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly",
"responses": [
{
"c_id": "6084",
"date": "14 Dec 2020",
"name": "Josh Zitovsky",
"role": "Author Response",
"response": "This paper has two components: 1) An advance in computational efficiency for estimating beta-binomial regression coefficients with shrinkage. The authors have produced a C++ implementation of the inference code previously written in R. Both versions of the code are implemented in the apeglm R package. 2) An application of this new implementation of their method to the task of inferring allele-specific expression (ASE) and an assessment of its statistical performance in relation to two alternative approaches (ash and MLE). As the authors start the paper by discussing ASE, rather than computational inference for shrinkage models, it is not immediately apparent that the innovation presented in this paper is computational rather than statistical. Distinguishing these two components clearly would make it more readily apparent that the paper does not present a novel statistical method. We feel that the manuscript title referencing “software”, the abstract mentioning “we evaluated the accuracy of three different estimators” and “We also wrote C++ code to quickly calculate ... apeglm estimates”, the citation of the apeglm publication in the Introduction (“To this end, we look at three different estimation methods... approximate posterior estimation of GLM coefficients (apeglm)11”), and the note about the software in the Introduction (“We also introduced new source code for the apeglm package”) make it clear that the apeglm shrinkage method is not proposed as novel in this manuscript. The modelling of ASE has important facets that the authors do not discuss in the introduction (page 3) but which other (uncited) methods have addressed. For example, in a given sample, a gene may contain multiple heterozygous variants (potentially with uncertain phasing of alleles). Each heterozygous variant could overlap different sets of isoforms, each of which may have different levels of ASE. This phenomenon is modelled by the MMDIFF method (Turro et al, 2014, Bioinformatics1), for example. The authors should acknowledge this (unmodelled) complication in ASE and explain how they summarise allele-specific count data across multiple variants (e.g., SNPs or indels, which are possibly unphased) within genes to obtain the count pairs modelled by the beta-binomial shrinkage estimators. We thank the reviewer for bringing up this concern. Here we have focused exclusively on observed allelic counts, ignoring uncertainty of reads that align to both alleles and aggregation of read information across SNPs within a gene. Such data could feasibly be acquired with longer reads that are approaching the transcript length, but in general we agree this as a limitation of our manuscript. We have now added the following to our manuscript to address this unmodelled complication: “The methods and performance benchmarks we focus on here address issues stemming from low-count genes and small sample sizes. There are other important concerns in allele-specific analysis of short read RNA-seq datasets, such as reference allele bias, but we do not address such problems here and the methods discussed cannot directly account for them. Our simulation does not involve reference allele bias, and the RNA-seq study we examine took specific measures to avoid reference allele bias. For methods and analysis concerns involving reference allele bias, see Turro et. al.4 and Castel et. al.1.\" The authors have performed several simulation studies and an analysis of a real ASE dataset. Both shrinkage estimators outperform MLE in the simulation studies. However, apeglm and MLE do approximately equally well in the real data set and both outperform ash by a significant margin. In addition, filtering of genes with low allele-specific read counts improves the MLE in the simulation studies but it does not do so in the real data analysis. This discordance demonstrates that the real data are very dissimilar from the simulated data. Although I don't think a major rewrite is warranted, if the authors could demarcate the computational advance (which can be demonstrated by simulation studies that are not representative of ASE, as the authors have done) from the specific application to ASE (using a real data set and perhaps a more faithful simulation study), the striking difference in performance shown in Figures 1-3 would be less incongruous. We thank the reviewer for pointing out that the simulation and real data results may have been seen as contradicting each other. Based on concerns voiced by other reviewers and our own investigations, we have determined that the issue is not in the simulations, but rather in the real data analyses. Specifically, when benchmarking our methods on the real data set, we had treated the ML estimates from a held-out set as the truth, but as the held-out set only contains 18 samples, the inherent instability and estimation variance present in ML estimators could still present an issue in the accuracy of these estimates. In other words, it may not be reasonable to expect that these ML estimates are close to the true effect sizes, and treating them as such could bias results in favor of ML estimates and against ash (as ash estimates are further from the MLE than apeglm on average). The real data analyses now have been changed to focus more on qualitative comparisons where the truth need not be known (e.g. extent of shrinkage, estimation variance, etc.), and we have largely left estimation accuracy assessments to the simulations. With these changes in place, the simulation and real data results are no longer incongruous. In the introduction, the inability of other methods to model the effects of continuous covariates or estimate differences in allelic imbalance between groups (this is not the case though, see MMDIFF) is highlighted and contrasted with the proposed method. However, the authors' own analysis of real data only uses an intercept model. It would be desirable to demonstrate the flexibility afforded by the proposed approach. Thank you for bringing the Turro, Astle and Tavaré (2014) paper to our attention. We have added a mention of this paper in the Introduction as an example of a Bayesian method that can deal with allelic counts and arbitrary design matrices, and have removed the sentence that mentioned that methods do not exist to perform Bayesian analysis with arbitrary designs. Moreover, we agree that it would have been useful to showcase our method on more complicated design matrices to demonstrate the flexibility of our method. To this extent, we have extended our analysis of real data to include an application of apeglm and ash to a model with two binary covariates and an interaction. The results are discussed in the last paragraph of the “Sampling from the mouse dataset” subsection of the “Results” section. In the assessment of statistical performance using the real data set, the MLEs obtained from the held-out data are treated as truth, even though earlier in the paper the authors demonstrate that MLEs have a particularly high mean absolute error. Presumably, this is the case (for genes with relatively low counts) even when the sample size is 18. The authors should consider alternative measures of performance that do not have this drawback. We agree that treating the held-out MLEs as the truth is problematic and have changed the analyses of our real data set so that results do not depend on knowledge of the truth. See our previous response detailing this issue. Minor comments: p3: \"estimates for allelic expression proportions can be highly variable\" - estimates are fixed, the authors should write \"estimators\". This typo has been corrected. p3: a cancer dataset may not be the best choice of example to refer to the proportion of genes with allele-specific reads, due to the prevalence of somatic mutations. We now clarify that the TCGA dataset referenced here only used the normal breast tissue samples, not the tumor samples. p3: when discussing filtering as a \"remedy\" perhaps explain that this achieves a boost in specificity at the cost of power. We have added this explanation as suggested. p3: \"the most robust and reliable when dealing with small sample sizes\" - this part of the sentence does not follow from the previous part, as there is no mention of ash's inadequacy. We have changed this part of the sentence from “the most robust and reliable” to just “robust and reliable”. p3: \"also introduced new source code\" - it is not clear what the \"also\" refers to. We have changed this sentence to make it more clear. p4: \"the probability that counts for a particular gene belong to a particular allele\" should be changed to \"the probability that a read for a particular gene belongs to a particular allele\" as the total \"counts\" will not be assigned to an allele as a block (the total counts derive from a heterogeneous mixture of reads from the two different alleles). We have made the suggested change. p4: more information should be given about how the scale parameter of the Cauchy prior is \"estimated by pooling information across genes\". We have added the mathematical details regarding how the scale parameter is estimated in the Supplementary Methods section. p4: the placement of the \\cdot indexing the bold face beta is unusual, as the j subscript corresponds to the first rather than the second index. We have made notational changes so that the \\cdot appears after the j subscript and not before p9: rerunning the simulation study with 4 v 4 samples having run it with 5 v 5 samples seems unnecessary, as such a small change in sample size is unlikely to alter the conclusions. Another reviewer made a similar comment, and so this result has been removed. p9: \"Figure 1d\" should read \"Figure 3d\". The typo has been corrected."
}
]
}
] | 1
|
https://f1000research.com/articles/8-2024
|
https://f1000research.com/articles/9-1454/v1
|
14 Dec 20
|
{
"type": "Research Article",
"title": "Modelling the health and economic impacts of different testing and tracing strategies for COVID-19 in the UK",
"authors": [
"Tim Colbourn",
"William Waites",
"David Manheim",
"Derek Foster",
"Simone Sturniolo",
"Mark Sculpher",
"Cliff C Kerr",
"Greg Colbourn",
"Cam Bowie",
"Keith M Godfrey",
"Julian Peto",
"Rochelle A Burgess",
"David McCoy",
"Nisreen A Alwan",
"Guiqing Yao",
"Kang Ouyang",
"Paul J Roderick",
"Elena Pizzo",
"Tony Hill",
"Nuala McGrath",
"Miriam Orcutt",
"Owain Evans",
"Nathan J Cheetham",
"Chris Bonell",
"Manuel Gomes",
"Jasmina Panovska-Griffiths",
"Rosalind Raine",
"William Waites",
"David Manheim",
"Derek Foster",
"Simone Sturniolo",
"Mark Sculpher",
"Cliff C Kerr",
"Greg Colbourn",
"Cam Bowie",
"Keith M Godfrey",
"Julian Peto",
"Rochelle A Burgess",
"David McCoy",
"Nisreen A Alwan",
"Guiqing Yao",
"Kang Ouyang",
"Paul J Roderick",
"Elena Pizzo",
"Tony Hill",
"Nuala McGrath",
"Miriam Orcutt",
"Owain Evans",
"Nathan J Cheetham",
"Chris Bonell",
"Manuel Gomes",
"Jasmina Panovska-Griffiths",
"Rosalind Raine"
],
"abstract": "Background: Coronavirus disease 2019 (COVID-19) is resurgent in the UK and health and economic costs of the epidemic continue to rise. There is a need to understand the health and economic costs of different courses of action.\nMethods: We combine modelling, economic analysis and a user-friendly interface to contrast the impact and costs of different testing strategies: two levels of testing within the current test-trace-isolate (TTI) strategy (testing symptomatic people, tracing and isolating everyone) and a strategy where TTI is combined with universal testing (UT; i.e. additional population testing to identify asymptomatic cases). We also model effective coverage of face masks.\nResults: Increased testing is necessary to suppress the virus after lockdown. Partial reopening accompanied by scaled-up TTI (at 50% test and trace levels), full isolation and moderately effective coverage of masks (30% reduction in overall transmission) can reduce the current resurgence of the virus and protect the economy in the UK. Additional UT from December 2020 reduces the epidemic dramatically by Jan 2021 when combined with enhanced TTI (70% test-trace levels) and full isolation. UT could then be stopped; continued TTI would prevent rapid recurrence. This TTI+UT combination can suppress the virus further to save ~20,000 more lives and avoid ~£90bn economic losses, though costs ~£8bn more to deliver. We assume that all traced and lab-confirmed cases are isolated. The flexible interface we have developed allows exploration of additional scenarios, including different levels of reopening of society after the second lockdown in England as well as different levels of effective mask coverage.\nConclusions: Our findings suggest that increased TTI is necessary to suppress the virus and protect the economy after the second lockdown in England. Additional UT from December 2020 reduces the epidemic dramatically by Jan 2021 and could then be stopped, as continued TTI would prevent rapid recurrence.",
"keywords": [
"COVID-19",
"Test",
"Trace",
"Isolate",
"UK",
"Health",
"Economic",
"Impacts",
"Mathematical Model"
],
"content": "Introduction\n\nIn the UK, since the first confirmed case of coronavirus disease 2019 (COVID-19) on January 31, 2020, over 1,538,000 people have tested positive and over 55,838 deaths have been confirmed as of November 24, 20201 with a notable resurgence in the number of cases and deaths from September 2020. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the virus responsible for COVID-19, can be transmitted through inhaling viral droplets from an infectious person2, but infection by fomites3 and aerosols4 may also occur. Reducing contact via testing, tracing and isolation (TTI), restrictions on social contact (i.e. lockdown measures), instigating good hygiene and ventilation, and the use of face coverings (masks) are now widely used non-pharmaceutical interventions (NPIs) to reduce COVID-19 transmission. The magnitude of benefit varies between interventions, with emerging evidence from laboratory experiments5–7, clinical studies8–10, and epidemiological studies11–15 that face masks reduce infection risk.\n\nFollowing the resurgence in COVID-19 cases since September 2020 in the UK, the UK Government imposed a second lockdown in England from November 5, 2020 to suppress transmission and prevent hospitals being overwhelmed (the devolved governments in Wales and Northern Ireland introduced lockdowns in October and Scotland introduced additional restrictions in November). While rollout of an effective vaccine is awaited, intermittent lockdown may remain the main tool for SARS-CoV-2 suppression. But lockdown strategies have important economic, educational, social and psychological consequences16. Alternative means of SARS-CoV-2 suppression such as via effective TTI and high coverage of face masks are required to balance protecting the public from COVID-19 resurgence with protecting the economy and reopening society. Our study is the first that attempts to shed light on this question by combining epidemiological modelling with economic analysis.\n\nSince the onset of the pandemic, modelling has played an important role in understanding COVID-19 epidemic trends as well as assessing the effectiveness of different intervention strategies17–21. For example, recent modelling work19 highlighted the need for an effective TTI programme that tests sufficient numbers of people with symptoms, traces their contacts effectively and ensures they isolate to prevent COVID-19 resurgence, as schools reopened alongside society from September 2020. However, this study did not include an economic analysis. In fact, to our knowledge, only two models to date have considered the economic consequences of different interventions combined with projecting model outcomes from a model of disease progression, and neither are focused on the UK22,23. This is challenging because uncertainties in data and calibration are magnified when composing models: a combined model can contain more uncertainty than the individual model due to propagation of uncertainties. For example, balancing a detailed model of transmission and a detailed model of cost-effectiveness or net-benefit is very data reliant, but in the context of COVID-19, information on both fronts is scarce, though emerging.\n\nIn this study, we handle these limitations by directing our focus on understanding the most important features of the combined model. We use a Susceptible-Exposed-Infectious-Removed (SEIR) model24 and a simplified economic analysis to explore the pattern of infections and deaths from COVID-19 and the economic impact of different NPIs in the UK. To make a wide range of scenarios accessible, we provide a user-friendly web interface that allows policymakers to explore trade-offs between different scenarios. This tool is aimed to facilitate discussion of key proposed high-level policies and their impacts based on an easily comprehensible model. This is opposed to static pronouncements about the optimal reaction, in which a complex model such as Covasim used in Panovska-Griffiths et al.19 must be used to accurately represent the complexity of the projected course of the disease. Importantly, for the first time, we contrast TTI strategies that test only symptomatic people and trace their contacts, with combined TTI and additional universal testing (combined TTI+UT). We also look at the role of face masks in suppressing the epidemic in conjunction with TTI and UT.\n\nOn September 09, 2020, the UK government announced that universal testing added to the current TTI strategy was being considered as an option to reopen society. In general, a universal testing programme might include more frequent and mass testing, e.g. at schools, and before concerts and sports matches, as suggested in recent reports from the Scientific Advisory Group for Emergencies (SAGE), which advises the UK Government25. But it can also mean repeated testing of the entire population. In this paper, under our definition, universal testing (UT) means testing everyone routinely, in contrast to testing only people with COVID-19 symptoms (we assume 70% of cases are symptomatic19,26–28). Weekly SARS-CoV-2 testing of the entire population was proposed for the UK earlier in the epidemic29, via saliva testing30 that is now being piloted31, and has been followed by the recent initiative of the UK Government to undertake mass-testing in high-prevalence local authority areas32. Similar universal testing strategies have also been proposed for the USA33,34.\n\nThis study combines mathematical and economic modelling to contrast the health impact, in terms of mortality burden, the economic outcomes, in terms of economic costs of shutdowns and isolation, and the direct costs of targeted TTI strategy in absence and in presence of additional UT (TTI+UT strategy). For each testing strategy we model reopening of society with the contact rate increasing from December 02, 2020 to that before the lockdown in November 2020. The contact rate is defined as the daily number of contacts per person and assumed to be around 11 contacts per person in the pre-COVID-19 era35. For each scenario we incorporate the use of face masks under assumptions of mean effective coverage (EC), which is the product of the efficacy of face masks and the proportion of contacts in which they are worn (details on methods and sensitivity analysis are in Methods, at the end of the article).\n\n\nResults\n\nThe results of the modelling and economic analysis are shown in Figure 1 and Table 1. The graphs in Figure 1 were generated with Streamlit.\n\nFigure 1A (top): Scenario 1: Base testing scenario (Base Testing, Tracing and Isolation (TTI)), Testing level of 25%, tracing level of 50% and assumed full isolation. Figure 1B: Scenario 2: Enhanced symptomatic testing (Enhanced TTI), Testing level of 50%, tracing level of 50% and assumed full isolation. Figure 1C (bottom): Scenario 3: Enhanced symptomatic and asymptomatic testing (TTI + Universal Testing (UT)) Testing level of 70%, tracing level of 70% and assumed full isolation. Figures were generated under the combination of parameters that define these scenarios as described in the methods and using the Streamlit interface.\n\nThe modelling suggests that TTI interventions might prevent future epidemic waves if enough symptomatic people are tested and their contacts effectively traced and all isolated, in combination with at least moderate effective coverage (30% in our case) of masks (Figure 1A).\n\nSpecifically, if only 25% of symptomatic people are tested, and 50% of their contacts are traced and all isolated, (Figure 1A), then following the second lockdown in England, an increase in the number of cases is evident. This will push the effective reproduction number R above 1 and the epidemic will persist.\n\nTo prevent this, an adequate level of testing of symptomatic people is necessary; in Figure 1B we show the outcomes when the testing level is doubled to 50%, while tracing level remains the same as before (50%) and full isolation of people tested and their contacts is assumed. In this scenario, we see the declining trajectory in the number of cases in January 2021 onwards and the effective reproduction number dropping below 1.\n\nThe results for TTI+UT are shown in Figure 1C. In order for the predicted epidemic trajectory to decrease rather than increase after the second lockdown in England and from December 2020, even with nearly universal weekly testing, additional symptomatic testing and tracing is necessary. In Figure 1C we show the combination of 70% testing and 70% tracing that, for this combination of parameters, is necessary to suppress the virus in future and bring R rapidly down.\n\nOverall, increasing TTI testing clearly reduces the future mortality burden: the model projects 110,000 deaths with adequate TTI (50% testing of symptomatic, 50% tracing of their contacts and isolating all) compared to 140,000 deaths with weaker testing (25% testing and same trace-isolate assumptions), by May 31, 2021 (Table 1). Additional UT will reduce this burden even further: 90,000 deaths are predicted in this scenario by the end of May 2021.\n\nConsidering the economic impacts, increasing the TTI from testing 25% to 50% of symptomatic people reduces the economic loss (£543 billion vs £541 billion by the end of May 2021; Table 1). Additional UT reduces the economic loss further (£454 billion by the end of May 2021). Across all scenarios the majority of costs incurred are for tracing, whereas UT naturally has much higher testing costs (Table 1).\n\n\nDiscussion\n\nThis work combined epidemiological modelling with economic analysis and developed a user-friendly interface to project our results. We simulated different testing strategies after the lifting of the second lockdown in England in December 2020. Our findings suggest that loosening restrictions accompanied by scaled-up TTI (50% testing of symptomatic people, 50% tracing of their contacts and full isolation of those tested positive and their contacts) and moderate (30%) effective coverage of masks can significantly further limit spread of COVID-19 after the second lockdown, and protect the economy in the UK. Additional weekly universal testing of the entire population from December 2020, can suppress the virus further, save more lives, and avoid more economic losses, though costs more to deliver. To achieve this, UT would have to be combined with additional TTI covering symptomatic cases (70% testing of symptomatic, 70% of their contacts traced and full isolation), and combined TTI+UT is more costly than scaled-up TTI.\n\nThe particular value and novelty of our work lies both in combining the epidemiological impact with an economic model and in the comparison between two different testing strategies that are currently being considered by the UK Government. Our epidemiological model, described in detail in a separate paper24 is a version of the classic population-based SEIR model extended to include different testing and tracing strategies. SEIR models are well-established as a tool in modelling infectious diseases and have been widely used in modelling different COVID-19 questions36. Our economic model represents a simple economic analysis developed explicitly for this study to capture public sector costs of testing and tracing and wider economic impacts of shutdown and isolation. Our approach is deliberately simple and our interface is easy to use.\n\nThe epidemiological model is readily extendable to include varying degrees of symptomaticity, including the pre-symptomatic period and its duration, varying amounts of viral load correlated with levels of transmission, infectiousness levels, duration of exposure, infectiousness period, and test sensitivity. It can also be easily adapted to include different age and risk stratifications, describe more elaborate representation of the natural history of the disease and progression among different subpopulations, and indeed how contact and transmission varies among and across different cohorts. One aspect that needs to be explored further in future is our modelling assumption of perfect isolation following tracing. Adherence to isolation is an important aspect of a TTI strategy and the balance between testing, tracing and isolation levels needs to be untangled more. But this was beyond the scope of this paper. Finally, uncertainty can be added to the model by setting parameter distributions. But all of these complications add different levels of nonlinearity, which makes the model multidimensional and makes it more difficult to untangle the real contribution of various factors. Our intention was to keep the epidemiological model as simple as possible and combine with economic modelling. For this reason, we also do not consider spatial variation, the difference between cities and rural settings, or epidemics developing in different ways in different regions. For this study, we chose to focus specifically on testing, tracing and isolation, and the trade-offs between different strategies at an aggregate level. An important reason for this is the difficulty of obtaining sufficient good quality data with which to parametrise a more sophisticated model: doing so would have no guarantee of accuracy and risks distracting from our central point. Future work could usefully examine the extent to which the present conclusions hold in more complex models.\n\nIn particular, our model does not model specific populations at heightened risk of COVID-19 infection or death, e.g. males, people of black, Asian or mixed ethnicity, or people with co-morbidities37. Future work could look at risk-group and/or geographical stratification within the model, but this was beyond the scope of this study. Recent work has identified additional population cohorts, e.g. people with multiple morbidities or learning disabilities, and people at heightened risk of hospitalisation, intensive care unit admission and death from COVID-1938. Future extension of the work presented in this paper will combine these findings with the modelling and economic framework developed here, to explore the health impact and economic outcomes of different testing interventions on cohorts at higher risk from COVID-19. This will allow us to extend our outcome measures. We limited our examination of health outcomes to direct mortality, but recognise that chronic illness and organ damage from COVID-1939 may have long-term effects not only on the health and well-being of the people affected, but that COVID-19 also creates many second order impacts including but not limited to damage to the economy longer term, impacts on mental health, and social structures. We have not included these outcomes in order to keep the modelling framework as simple as possible, so our conclusions on the potential benefits of different interventions are likely to be conservative.\n\nOur economic model does not explicitly model any costs to enforce isolation, or costs to provide separate accommodation for people to isolate in. We also assume perfect isolation in our model and future work will explore this further27. Policies to support effective isolation, such as lost income reimbursement, community support, childcare, online education, and volunteers to run errands for those isolated, are important. Costs of enforcement may be covered by a combination of using existing policing systems and paying for additional measures with fines gathered from violators.\n\nDespite omission of important details such as detailed modelling of household and community layers, and assuming perfect isolation policies, the model indicates some priorities for immediate piloting with transparently calculated cost estimates. Tracing should focus on how to improve actual isolation rates among all recent contacts, not on targets. An important implication is that with high compliance regular testing might end the need for social distancing and should therefore be properly piloted. To minimise unnecessary isolation and encourage compliance the rapid tests that will be available to NHS and care home staff could be offered daily to anyone in isolation. Automatic payment of (say) £500 initially then £100 per day might achieve high compliance with isolation in those at highest risk including the unemployed and homeless. Average isolation until 2 or 3 days between negative tests would be about 3 days for uninfected contacts and up to 10 days for cases, so the average cost of full furlough would be less than £1,000 per person. Furlough for 10 times the half a million or so adults currently infected in the UK at an average of less than £1,000 would cost less than £5 billion and should be considered.\n\n\nConclusions\n\nIn summary, we have combined epidemiological modelling, economic analysis and a user-friendly interface to contrast the impact and costs of two levels of testing and tracing within the current UK TTI strategy and when universal testing is also included. Our findings suggest that increased TTI is necessary to suppress the virus after the second national lockdown in England. With reopening of society in December 2020, as before the second lockdown, scaled-up TTI and moderate (30%) effective coverage of masks can prevent further resurgence of the virus and protect the economy in the UK. Additional universal testing from December 2020 can save more lives and avoid more economic losses.\n\n\nMethods\n\nThe model is available from GitHub and is archived with Zenodo40. The impact analysis was conducted using the SEIR-TTI compartmental model shown schematically in Figure 2 and with details provided in Sturniolo et al.24 The model is an extended version of the classic SEIR model that incorporates probabilistically the effects of testing, contact tracing, and isolation24.\n\nSEIR is a compartmentalised model describing susceptible (S), exposed (E; infected but not infectious), infectious (I) and removed (R) population cohorts. Individuals move between these compartments in sequence as they become exposed, infected and infectious during disease progression until recovery. Each compartment comprises diagnosed (D) and undiagnosed (U) individuals with diagnosis leading to isolation. We assume that diagnosis happens through testing or putatively through tracing. A non-infectious individual that is “diagnosed” has effectively been misdiagnosed and the result is that they are needlessly required to isolate. Individuals transition between compartments X and Y at rates ∆X→Y which we derive in Sturniolo et al.24 from which this figure is reproduced.\n\nWithin the model, possible transitions between cohorts are indicated with arrows. Within each of these states, an individual can be unconfined or isolated. Infectious (I) individuals who are unconfined may be tested and become isolated. An individual in any state who is traced is isolated. Once isolated, individuals remain so for 14 days. Susceptible (S) isolated individuals cannot become infected due to their isolation, and return to the unconfined state after a 14-day delay. Exposed (E) and infectious individuals (I) do not return directly to the unconfined state and first progress to removed (R). Removed (R) and isolated individuals return, as with susceptible (S) individuals, to an unconfined state once 14 days has elapsed. Tracing is described by a rate of tracing eta and a probability of success chi.\n\nModel parameters and details of the calibration used are detailed below. Briefly, the model-projected deaths between January 21, 2020 and November 07, 2020 were matched to the publicly-available data on deaths reported within 28 days of positive tests, using the UK dashboard. This allowed us to determine the transmission probability beta, the numbers of contacts, the date of the onset of the epidemic, number of infectious people at the onset of the epidemic, the infection fatality rate (IFR) and the testing and tracing levels to constrain the model to mimic the reality of the COVID-19 epidemic in the UK until November 07, 2020.\n\nWe note that within the model an intervention changes the model parameters at a defined time. The principal parameters that are changed are the contact rate (average number of contacts per person per day) representing differing regimes of social distancing or lockdown, and the testing and tracing rates, representing building up capacity of TTI. A trigger changes parameters when a condition is met. The trigger conditions are the number of infections passing a set threshold. We use different thresholds according to whether the number of infections is increasing or decreasing to avoid rapidly oscillating between distancing regimes, which would not be politically or economically feasible. We use a threshold of < 10,000 infections to release lockdown as it approximates what may be a safe level of limited community transmission. We use a threshold of > 40,000 infections for beginning lockdown to reflect time elapsing between opening and closing given exponential growth. Lockdowns are not triggered in the scenarios shown in this paper.\n\nWe developed an economic model based on closures and isolation that uses the impacts from the SEIR-TTI model to calculate the cost on the economy41. Specifically, within the economic model we modelled GDP reduction as a function of the reduced contact rate and isolation requirements. That means reduced economic activity was modelled when people stay at home, rather than shop, work, or engage in other economic activities, as during the lockdown period for example. Because there are already estimates about the degree of contraction which occurred during the lockdown (e.g. from the Bank of England), we use reduction in interpersonal contacts as a valid proxy for the proportion of full shutdown effectively continuing to occur. This is more closely related to the epidemic evolution than direct estimates of shocks to supply and demand used in the past42, and has the advantage of using actual estimates of the economic impact of COVID-19 based on mitigation policy choices. Our economic model is further detailed below. Potential health and social costs of lockdown that are not included in our economic model are shown in Table 2 below.\n\nEstimated costs for TTI and UT are shown in Table 3 below. There are three principal components, which we also explain in detailed narratives below: (1) contact tracing using a network of public health community officers, mobile phone apps, and supervisors; (2) home-based saliva testing for active SARS-CoV-2 infection; and (3) follow-up and isolation of infected individuals and households. As per the economic model, total costs are variable depending on policy scenario and case numbers.\n\nShown are unit/daily costs. Total costs are variable dependent on policy scenario and case numbers. *Costs of testing are based on a pilot study in Southampton of mass home-based saliva testing31.\n\nWe contrasted three testing strategies from December 02, 2020 when the second lockdown in the England is lifted. Specifically, we compared the impact and costs of\n\nA) Baseline targeted test-trace-isolate (TTI) strategy from December 02, 2020 resembling the current strategy that tests only symptomatic people, traces their contacts and isolates those testing positive and their contacts. We model 25% testing, 50% tracing and full isolation.\n\nB) Enhanced TTI strategy from December 02, 2020 modelled by increased testing level compared to the baseline TTI. We model 50% testing, 50% tracing and full isolation.\n\nB) Universal testing in addition to the targeted TTI strategy from December 02, 2020. We model 70% testing across symptomatic and asymptomatic, 70% tracing of their contacts and full isolation.\n\nFor each of the strategies we include wearing masks and that the contact rate c increases to a value of 3.4 as the average c of the values between 2.8 and 4 from the CoMix study43 representing the values during the first lockdown and after the relaxing of it. While in this paper we present the results for a moderate effective coverage of masks, lower and higher effective coverage can be explored further on our interface.\n\nOn the interface that we have developed, we modelled three different levels of effective coverage of masks (defined as the size of the reduction in COVID-19 transmission in the population as a whole due to the use of face masks): 15%, 30% and 50%. These were estimated as a product of contact coverage (the proportion of infectious–susceptible contacts in which at least one person is wearing a mask), and per-contact effectiveness (the size of the risk reduction when at least one person is masked). Surveys and media reports suggest mask prevalence outside the home is around 40–80%44–47. Assuming no mask usage within the home, and 3–8 daily contacts of which 1–3 are at home34, population-wide prevalence is 30–60%. In a well-mixed population, this translates to contact coverage of 45–85%, because in many contacts involving unmasked individuals the other party will be wearing a mask. Reviews of mask effectiveness suggest a benefit of approximately 45% for uninfected wearers in non-healthcare settings, with a plausible range of about 20–70%8–10. Since these were mostly case-control studies, we adjust downwards for biases that may have inflated the effect size, and for mask type: most people in the UK use cloth masks48, which are probably less effective than the medical masks used by some participants in the reviewed studies7,8,49,50. However, we then adjust upwards for source control, which is hard to quantify but has been clearly demonstrated in the laboratory5–7. Taking all of this into account, and weighting by the proportion of single- and double-masked contacts, we end up with an effect size of 20–60%. Multiplying that by the contact coverage, we obtain an effective coverage estimate of 30%, with a range of 15–50%.\n\nIn the paper we present the results for this moderate 30% EC of masks, but the other scenarios can be explored on the interface described below.\n\nThe epidemiological model and the economic framework were combined into a user-friendly interface that we developed for the purpose of this analysis. The interface is available on Streamlit and the figures in this manuscript were directly imported from the interface plots. Within the interface there are clickable options that allow the user to explore different permutations of the scenarios we have considered in this study; a snapshot is shown in Figure 3.\n\nDetails of the compartmental SEIR-TTI model we used for the impact analysis are in Sturniolo et al.24 In summary, it is an extended version of the classic SEIR model that incorporates probabilistically the effects of testing, contact tracing, and isolation. For the purposes of the analysis here, we fixed the majority of the model parameters to the values from the literature as per Table 4. We fitted the four parameters: transmission probability β, the rate of contacts, the date of the onset of the epidemic, number of infectious people at the onset of the epidemic and the infection fatality rate (IFR) to match the model projected deaths to the publicly available mortality data from the UK government. To match the UK epidemic, we consider a single infectious individual introduced into the UK in late December 2019. This is simply a mathematical convenience and not a claim about the seeding of the actual epidemic in the UK. It is not our purpose to investigate the origins of the epidemic in this article. In reality it is likely that multiple infectious individuals were introduced into the UK at a later date. This distinction is immaterial to the functioning of the model. By calibrating to the mortality data, we obtain a transmission probability β of 0.0435 which translates to a basic reproduction number R0 of 3.3 when c is 11 contacts per day and under no interventions. The results of the calibration are shown in Figure 4 and further details are provided here.\n\n\nEconomic model details\n\nWe calculate reduction in GDP due to the pandemic and lockdown measures by relating GDP to the model parameter c (contacts per day) as a proxy for economic activity, for every day of the model scenario trajectory. GDP of £186 billion per month is taken as the pre-pandemic level74, when c = 11, whereas during lockdown GDP is 25% lower, when c = 3. For intermediate values of lockdown or distancing, GDP loss is scaled accordingly. The pandemic itself results in GDP loss, as c = 80% of baseline even when lockdown is fully released, i.e. the country is not back to c = 11 (100%) normal economic activity.\n\nIntervention costs are calculated by dividing the budget items shown in Table 3 by start-up costs and on-going costs: for tracing, and for testing. Costs to notify, enforce, and otherwise manage isolation are assumed to be covered by fines levied for breaches of isolation. Overall start-up costs for contact tracing are £10m for the app that supplements human contact tracing efforts, as well as a recruitment campaign to hire the number of needed contact tracers, supervisors, and managers. Start-up costs include recruitment and training costs for personnel, and app maintenance costs, for which we have made several assumptions detailed in Table 3 and below, though these are small enough not to significantly alter overall costs. On-going costs are scaled according to the numbers required by the intervention by estimating the cost per contact traced and the cost per test, as follows.\n\nContact tracing costs. Using our assumptions around number of contacts before lockdown (c0=11), during lockdown (c=0.3*c0), and after the lockdown is lifted (c=0.8*c0), we determine that over a period of seven days a total of 77 contacts need to be traced before lockdown, while during lockdown only 23 contacts will need to be traced.\n\nAs a policy design assumption for the model, we stipulate that contact tracers and supervisors are hired for a minimum of three months (90 days) for the system to function professionally, while team leads are hired for the entire term of contact tracing. Contact tracing costs are therefore blocked into three-month periods based on the anticipated maximum number of tracers needed in the subsequent three-month period. Recruitment and training costs for any additional tracers needed in the subsequent three-month period are added to the cost for that three-month period.\n\nThe recurring tracing costs can be used to determine a (marginal) cost per hour of tracing, which can then be used to determine the cost per trace given our estimate of 1.26 hours work per contact traced (Table 5). We estimate the cost per contact traced is approximately £18 (calculations as per ‘Tracing costs per case traced’ sheet here).\n\n*High-risk exposure contacts are people having had face-to-face contact with a COVID-19 case within two metres for more than 15 minutes; having had physical contact with a COVID-19 case; having had unprotected direct contact with infectious secretions of a COVID-19 case (e.g. being coughed on); having been in a closed environment (e.g. household, classroom, meeting room, hospital waiting room, etc.) with a COVID-19 case for more than 15 minutes; or a healthcare worker or other person providing care to a COVID-19 case, or laboratory workers handling specimens from a COVID-19 case, without recommended PPE or with a possible breach of PPE78.\n\n†Low-risk exposure contacts are people having had face-to-face contact with a COVID-19 case within two metres for less than 15 minutes; having been in a closed environment with a COVID-19 case for less than 15 minutes; having travelled together with a COVID-19 case in any mode of transport; or a healthcare worker or other person providing care to a COVID-19 case, or laboratory workers handling specimens from a COVID-19 case, wearing the recommended PPE78.\n\nTesting costs. We estimate that each test costs £4.79 including start-up and recurring costs. The vast majority of these costs are the £4.50 for each actual test (£3.50 for the test kit, £0.50 for mailing out the test kit, and £0.50 for the courier from the tested person’s address to the local lab). Start-up costs for testing are the cost of the RT-LAMP machines (£27,000 each). Each machine can run 96 tests every 30 minutes75 so if we assume they will be running for 18 hours per day (two 9-hour shifts) they will process 3,456 tests per day. We assume 10 machines per lab on average, each with £500 per day overheads, 40 lab workers (four per machine: two for each shift), and two supervisors (one for each shift).\n\nTesting personnel costs are blocked into six-month periods based on the anticipated numbers of tests per day over the subsequent six-month period. In a six-month period where only 100,000 tests are being done each day, costs per test would still be approximately £4.79, as the number of labs, maintenance costs, and lab workers would be scaled down accordingly, and the RT-LAMP machines would be amortized over the full period of use.\n\nWe assume that if people are unable to afford their own face coverings they will be wearing reusable face coverings made from materials to hand in the home, at little or no cost. The UK government has issued advice on how to make and properly use a face covering.\n\n\nRealising the Resources for different test-trace-isolate strategies\n\nThere is emerging evidence that mobile phone contact tracing apps have the potential to facilitate effective COVID-19 epidemic control at scale and at speed18. Nevertheless, personal follow-up on foot will also be required to ensure all contacts, including the most vulnerable, are reached79. The additional costs of such a system are relatively small in the context of the problem we are seeking to address.\n\nFor feasibility reasons, we assume that control of COVID-19 would be managed through local authorities by Consultants in Health Protection/Communicable Disease Control and Directors of Public Health. This was the approach used, with success, until the re-organisation in 2002 and it ensured effective control of communicable disease via local knowledge of and relationships with the community, the local politicians and leaders, the laboratory, the hospital and its consultants, and the general practitioners80,81. Legal powers to take such responsibility are available through Schedule 21 (powers relating to potentially infectious persons) of the Coronavirus Act 2020. Regional Health Protection Teams from Public Health England could take on management responsibilities for local authorities in England (public health functions are already devolved in Scotland, Wales, and Northern Ireland) and co-ordinate regionally and centrally through its established infrastructure. This includes regional epidemiologists who have a key role in understanding the epidemic at a regional level, identifying differences between local authorities, and sharing expertise.\n\nMovement of people between local authority areas could be accounted for by data sharing between contact tracing teams. China, while being different in many ways, demonstrates the ability for this hierarchical approach to succeed in identifying contacts82.\n\nCase finding and contact tracing. Contact tracing remains a key control measure for maintaining suppression of case counts83. Table 5 shows the staff needed to handle new cases and control spread through contact tracing and isolation84.\n\nThe NHS Test and Tracing Service was launched on 29th May. While information on the structure, duties, and means of collaborating with the contact tracing teams in local authorities has not been published, it is reasonable to assume that this centrally managed service will provide some of the hours required to run the case finding and contact tracing function shown in Table 5. It seems that the service is limited to phone and internet communication with individuals. Because the levels of ascertainment of cases of this approach remains unknown, it will be prudent for local authorities to assume that at least half the manpower shown in Table 5 will be required by them.\n\nLocal public health capacity. Each new case will require 38 hours of community health staff and volunteer time to trace an average of 30 contacts and test 3.7 symptomatic contacts, two thirds of whom will have COVID-1984 (these numbers reflect a situation when physical distancing measures are in place). The requirement for staff will vary with time as relaxation of physical distancing increases contact numbers or as subsequent physical distancing reduces contact numbers, and should decline if phone applications as used in South Korea85 are used by sufficient numbers of individuals here and their accuracy increases (though we do not assume any increase in efficiency or success of contact tracing resulting from use of phone apps). On average there will need to be 5.1 full time trained contact tracers (Public Health Community Officers, PHCO; Table 3) to cope with each additional concurrent case, though this will vary by the number of contacts per day. The numbers of contact tracers will need to be adjusted accordingly to accommodate part-time working and to cover all seven days of the week, as all contact tracing should be done within one day for each case.\n\nA fraction of health visitor (HV) and environmental health officer (EHO) staff can be redeployed initially to lead local teams of contact tracers86. Most local authorities have established volunteer registers87 and recently retired HVs and EHOs can also support the contact tracing effort. New staff will also need to be hired, given limited capacity and the existing important duties carried out by HVs and EHOs. The system of contact tracing could be up within weeks with sufficient political will and commitment. We assume that it will be possible for most Directors of Public Health alongside the Public Health Physician secondees from Public Health England to assess if they have control of the spread of the virus in their district a week later. The incidence of new cases will vary between local authorities and regions.\n\nInitially the number of cases can be best estimated from local deaths. As the system gets underway, new cases can be notified in the standard way for notifiable diseases, for which testing is helpful but not necessary. The number of cases will fall as physical distancing succeeds, as in China. An estimated 800 to 1,000 contact tracers would be needed two weeks after peak deaths in the averaged-sized local authority (population ~375,000). We assume this is achievable, given the 750,000 people who have already volunteered to help the NHS tackle the pandemic88. Training is assumed to take one day, as is setting up the administrative arrangements using local authority resources. Testing facilities can be negotiated with the local health laboratory (see Testing section below). The local authority will be assumed to take on the public information function.\n\nCommunity advisory committees and local health communication strategies\n\nThe overall success of this strategy rests on the willingness of citizens to engage with and accept the necessity of contact tracing and isolation for 14 symptom-free days if in contact with a case, and of home testing via spit (saliva) samples. Social psychological literature suggests that health communication messaging and health interventions are most effective when anchored to meaningful dimensions of identity and personal experience89,90, which has been affirmed by evidence from previous epidemics including HIV91,92 and Ebola93. Community-led and co-production approaches in the context of the COVID-19 response have been lacking94, but would be critical in ensuring that local engagement strategies result in significant uptake of testing, tracing and isolation over time. We therefore suggest that each local area develop a community advisory committee, whose role is to advise on the suitability of the national plan in their area, and to support the design of a local public health communications strategy tailored to specific subpopulations. It is critical that this group is composed of individuals from the full range of ethnic and cultural backgrounds within the area, given the importance of identity and context to the promotion of positive health behaviours, and the existing marginalisation of subgroups of the population. A life course approach would also ensure that any and all messaging was targeted to the specific needs and concerns facing individuals across the life course.\n\nAt the outset, community advisory committees may need to meet regularly (e.g. weekly to co-develop communication materials); but over time, its role could transition to helping provide an accountability loop between communities and implementers and managers of the TTI programme, which would require less regular contact. In this way, community members are able to feed details of emergent challenges and difficulties that people face in adhering to cycles of lockdown, real-time data on the efficacy of support systems, and ability to adhere to testing requirements over time. These groups could be coordinated by Public Health COVID-19 supervisors (see below).\n\nThere are relevant concerns about how much time it would take to set up these groups in each area. However, each local entity will have a range of third and voluntary sector organisations who are already working to support various communities affected by the crisis. Rapid assessments and mapping of existing community networks by public health agencies would allow for a quick deployment of existing and active community groups in each area, to take control of recruiting relevant people from various backgrounds to engage with the committee.\n\nThe task of the supervisor will be to create an overarching structure to coordinate their efforts in a unified structure. In times of lockdown where participatory engagement is limited or restricted, evolving frameworks for how to conduct remote participatory research and community engagement could be adapted95. Such a community mechanism will have wide-reaching benefits, including; maintaining local buy-in over time, appropriately tailoring engagement strategies and innovating over time to maintain engagement, and helping citizens to feel as though they are a part of a wider process for promoting collective wellbeing. The latter has been shown as critical in other crisis and recovery focused settings96,97 and can have positive knock on effects for mental health outcomes in the general population, which is a growing concern in the crisis98.\n\nContact tracing budget. One Public Health Community Officer (PHCO) will need to be recruited per 1,000 population (the exact number needed to be recruited in each three month block depends on the number of infections as explained in the economic model section), with budget for 20% extra posts included to cover sickness and absence to help ensure contact tracing always meets demand. These people should be familiar enough with their community to identify individuals disconnected from government reach and internet apps. They could be unemployed or under-employed lay people, including those made redundant due to the pandemic. No prior public health experience or skills will be required beyond minimal educational attainment and having been resident in their local area for at least a year, though ability to speak appropriate languages will be relevant for some communities. The PHCOs could be trained via a short online course delivered by public health professionals and will undergo online refresher training every month. PHCOs will be paid a living wage of £10 per hour, £80 per day for an 8hr shift.\n\nPHCOs will be supervised by full-time Public Health COVID-19 Supervisors (PHCS), at a ratio of 1 supervisor per 50 PHCOs. These PHCSs could be graduates of master’s degrees in public health or related disciplines and appointed if they can pass a simple test about control of the COVID-19 epidemic in line with this strategy; or, if sufficient numbers are available and they would not be taken away from important existing duties, they could be Environmental Health Officers. They will be based in COVID-19 offices in their local authority area. Given 343 local authorities in the UK, each will have around 3 or 4 PHCS. PHCS will be paid £20 per hour, £160 per day.\n\nEach local authority will need a COVID-19 response team lead overseeing this effort. The team lead will directly manage and supervise the PHCS and have an overview of the COVID-19 situation in their local authority area. They will be public health specialists with at least five years of experience, perhaps already in post in the local authority area. Importantly, their duties will only relate to the COVID-19 contact tracing, testing and isolation strategy. Therefore, if already in post they will be relieved of other public health duties (and an additional public health lead recruited to oversee such duties) – or perhaps less disruptively, individuals without existing duties will be recruited to lead the COVID-19 response in their local area.\n\nThe importance of an integrated system with all workers solely focusing on COVID-19 needs to be emphasised. It is likely to be necessary to ensure the consistently high levels of contact tracing, testing and isolation required.\n\nMobile phone costs and travel costs are included for all cadres as needed.\n\nA population-wide testing programme99 is a core component of population-wide TTI. This would require the following resources, which are either currently available or can be sourced from UK suppliers within a matter of weeks:\n\n1. A register of names, dates of birth, and addresses of all residents registered with a GP, to be updated as necessary with test results, changes of address and addition of unregistered subjects. Anonymous registration with local outlets for sample collection and delivery is needed for those reluctant to give name and address. “Ghost patients”100 can be dealt with using the strategy developed by the ONS.\n\n2. New 96-well machines running direct RT LAMP assays101 18hrs per day processing 96 samples every 30 minutes. Experienced staff to operate them are already in place in large and small academic and commercial labs throughout the UK, including possible demonstration sites. Posts for four 9-hour shifts for lab workers will be needed: 1 technician running each machine and 1 filling the wells with samples.\n\n3. Self-sample spit (saliva) test kits including sample transport tubes individually labelled with name, date of birth, and barcoded ID, LAMPreagents (note RT LAMP does not require the RNA extraction step so needs less reagents), and microtiter plates for 10 million tests per day. Additional production facilities must be commissioned if necessary (Box 1).\n\n4. Arrangements to deliver and collect samples from every household once a week, with delivery to a testing lab within a few hours. Results would be directly uploaded online automatically by the RT LAMP machine into a LIMS system as the sample is diagnosed by the machine, coupled with auto texting of negative results using software already in place. Positive results in those without phone or email would be delivered by courier.\n\n5. This high throughput would depend on various regulatory emergency waivers:\n\n1. Lab staff would wear PPE where necessary but would not be accredited to conduct medical tests.\n\n2. Laboratories would be advised on precautions but not accredited for handling infectious samples.\n\n3. LAMP reagent production with normal non-medical quality control cannot be hampered by patents or regulations on medical test manufacture.\n\nWe recommend evaluation of regular COVID-19 saliva testing of the whole population in an entire city as a demonstration site (preferably several towns and cities), with strict household isolation following a positive test. Isolation ends when all residents test negative at the same time. Everyone else can resume normal life if they choose to. This should be assessed for feasibility in one or more cities with populations of 200,000–300,000. This experiment could only be achieved after extensive, transparent public engagement leading to widespread public acceptability across all social and economic groups. Economic and educational measures would need to be provided to ensure equity with the non-quarantined population. Although this is an ambitious proposal, it does need to begin as soon as possible, whilst the infection rate is fairly low but rising. The rate at which it then rises or falls compared with the rest of the UK will be apparent within a few weeks. A decision can then be taken on national roll-out, beginning in high-risk areas.\n\nA local population of 200,000 with 90% compliance will require 26,000 tests per day, plus an excess to offer more regular testing for NHS staff and care workers. Whatever the results, these data will enable policy to be based on real-time evidence (instead of modelling assumptions) on new infection rates in the expanding regularly-tested population and the untested remainder. The latter can be monitored by testing population samples as well as by NHS number linkage to hospital diagnoses and GP records. Complementary aspects of PTTI: contact tracing and phone apps will be critical in the unscreened population and may enable testing to be done less frequently as prevalence falls. Testing would be voluntary, but incentives for staying in isolation following a positive test in a household could be considered in line with those suggested by community advisory committees. Helplines would be provided to support households in isolation with access to income compensation, mental health support and food delivery.\n\nThese pilot studies, one of which has started on a smaller scale in Southampton with 14,000 people31, will show whether PTTI is a practicable way of responding to the COVID-19 epidemic. Even if the epidemic is not completely controlled in pilot studies the establishment of far greater testing and tracing capacity will facilitate other initiatives. Different households would return samples on different days, giving a daily sample of each small area. Depending on the proportion of people tested and cases detected a local outbreak could therefore be detected soon after it occurs, as test results would be automatically uploaded online by each LAMP machine.\n\nA register of everyone registered with a GP (suitably amended to deal with unregistered people and “ghost patients”) would be used to deliver and collect saliva (and nasal/throat in a subsample) self-samplers in bar-coded tubes labelled with name and date of birth of all residents to every household once a week. The register would be expanded to include any missing people who are subsequently identified (with unique ID numbers for those with no NHS number) and continuously updated to assign people to the household of their current address. Many “households” would have one resident.\n\nHouseholds would self-isolate on the day that any resident gets a positive test, with earlier self-isolation of a household when anyone in it is thought to have COVID-19 based on a publicised list of diagnostic symptoms, pending the household’s next test results.\n\nContact tracing (above) could be focused on the “hard to reach” population that the uncontrolled epidemic will then be confined to. Anyone not possessing a negative test result dated in the past week would be required to provide a saliva/nasal/throat sample and their name, address and date of birth. They would be added to the register and sent weekly self-sample kits like everyone else. There will be challenges with this, for example, inclusion of the homeless population, that may need to be overcome.\n\nSamples would be analysed on machines in university and commercial labs, if necessary by continuous (24-hour) operation (with very occasional down-time for maintenance), though we have costed 18hr per day operation. Laboratory and testing regulations would have to be set aside to enable the laboratory staff currently using these machines for other purposes to do the testing supported by additional assistants. Strategic planning to identify essential laboratory work that needs to be continued during the COVID-19 crisis will be required. This should consider the opportunity costs of not doing such work, whilst also considering the opportunities and costs of extra shifts to utilise the same equipment, recruitment and training of extra lab staff and potential efficiency gains to existing processes (including those that could be gained via relaxing regulations, along with the potential costs of relaxing such regulations).\n\nOne of the key bottlenecks for ramping up testing to such a large scale is the availability of reagents and test kit supplies for the tests. Creative ways of resolving this issue are urgently needed (Box 1).\n\n\n\nTTI and UT are ambitious compared to the number of tests currently conducted each day. However, it is in line with international estimates of the scale of testing required33,34. The UK government’s five-pillar plan for scaling up COVID-19 testing102 reaches out to local manufacturers to ramp up testing capability and pharmaceutical companies are also offering to help103. The extent to which such capacity can be transformed into sustained delivery of the government’s current target of 500,000 swab and antibody tests per day is still unclear.\n\nStudies are underway to confirm that saliva samples collected into simple specimen pots can reliably be used for mass population SARS-CoV-2 testing; if confirmed this would remove the current bottleneck in swab availability. The main testing reagents in short supply are not likely to be the non-biological chemicals used, large enough quantities of which could fairly easily be produced in around three months by industrial chemical companies. Some of these materials are already supplied by large companies such as BASF. The bespoke formulations of the mixtures of bio-based reagents, such as proprietary mastermixes and primers specific to each test kit, are potentially the main bottlenecks104. It will likely be easier and quicker for the existing manufacturers to scale up production than for a new company to attempt to do so, as the new company will require all of the same ingredients in order to exactly match the bespoke formulation of the specific test kit.\n\nTherefore, the UK government probably needs to coordinate industrial consortia of companies with relevant scale-up capabilities and Good Manufacturing Practice approval, such as Robinson brothers105 (based in the midlands), and test kit manufacturers, such as New England Biolabs and OptiGene, to ensure there is adequate supply of key reagents. In this way, test kit manufacturers will be enabled to create the quantities of the bespoke proprietary formulations needed for millions of tests a day in the UK.\n\nTo ensure manufacturers have adequate incentive to participate, the government could issue “put options” that allow the companies to recoup most of their losses in the event the kits are never used106. More traditional methods of reducing commercial risk, such as direct purchase orders and public-private partnerships, can also be considered so long as they can be arranged quickly enough.\n\nInitial estimates from an industrial chemist suggest the costs to cover the UK demand, per type of reagent, are on the order of £5-10m. It would require short bespoke use of manufacturing units (equipment) per component, the blending of the final formulation, and finally the development of appropriate logistics. The total cost is estimated to be less than £100m.\n\nRapid efforts will also be needed to source the swabs required to collect nasal/throat self-samplers and the bar-coded tubes labelled with name and date of birth of all residents, to deliver to every household once a week. Again, option-based guarantees and other de-risking measures could play an important role in ensuring the demand is met106.\n\nTTI: Test-Trace-Isolate UT: Universal Testing\n\nThe team of PHCO and PHCS will follow up all those who test SARS-CoV-2 positive and who therefore require isolation. They will ensure that the people requiring isolation understand they need to stay at home for the required period in order to not spread the virus, and steps will be taken to ensure that households have the resources necessary to comply with isolation in the first instance. The costs of policing any infringements will be met by the fines levied for such infringements (likely with surplus funds left over). Therefore, no costs are added for isolation encouragement and enforcement.\n\nFor isolation support and enforcement to work without disadvantaging marginalised groups further the following will need to be put in place:\n\n1) financial compensation for time off work to comply with a 14-day isolation order following tracing;\n\n2) clear guidelines on the roles and powers that police and other authorities have in enforcing isolation;\n\n3) a means-based fine system for infringements of isolation, based on household income levels/earnings;\n\n4) development of minimum packages of support that are streamlined to specific vulnerable populations – so support that is provided is bespoke for the needs of each household during an isolation period (i.e. houses where earning levels are not impacted will be offered a different resource package than those where earnings are impacted);\n\n5) assurances that basic resources (heating, water, electricity, internet access) will be guaranteed during the period of isolation, and for a one-month period post isolation.\n\nOn rare instances where households still break isolation rules, police officers will be put in touch with households in breach of guidelines. Fines will be levied in line with household income levels (there is precedence for this with speeding fines107).\n\n\nData availability\n\nAll data used in this paper is included in the tables of the paper or otherwise publicly available and included in our software, linked below.\n\n\nSoftware availability\n\nSource code available from: https://github.com/ptti/ptti/tree/F1000-final\n\nArchived source code at time of publication: https://zenodo.org/record/4298847\n\nLicense: GNU GPL Version 3+",
"appendix": "Contributions\n\nThis study was developed by TC, JPG, WW and DM with contributions from DF and RR. A previous version of the work was conceptualised by TC, JP, NA, KMG and PR and initially developed with DF, GY, RB, DM, CaB, EP, MO, MS, MG and RR, with early analysis done by GC and TC. The mathematical model used here was developed by WW, SS and JPG with input from TC and DM. The economic model used here was developed by DM and TC with input from WW, SS, JPG, EP, MG and MS. TC, WW, JPG, SS, GC, CaB, KMG, DF, EP, TH, NG, NC, MS, and MG contributed parameter values used in the model or interpretation. The scenarios used in the study were developed by JPG, TC, DM, WW and KG in discussion with all co-authors. WW, DW and JPG ran the modelling analysis with input from TC. DM and WW ran the economic analysis with input from TC, JPG, EP, MG and MS. JPG drafted this version of the paper with substantial contributions from TC, WW, DM, DF and RR. CaB, KMG, JP, RAB, GY, KO, PJR, TH and GC contributed specific sections of the manuscript. Co-authors provided critical feedback across developing iterations of the paper and have read and approved the final manuscript.\n\n\nAcknowledgements\n\nThe following people have provided helpful comments or suggestions, or have otherwise enabled this work: Katie Brennan, Martin J Buxton, Majid Ezzati, Amy Gimma, Gabriel Goh, Grace Grove, Agnese Iuliano, Christopher Jarvis, Paula Lorgelly, Tara Mangal, Ofir Reich, Justin Shovelain, Mark E Thomas.\n\n\nReferences\n\nCoronavirus (COVID-19) in the UK. Reference Source\n\nWorld Health Organisation: Getting your workplace ready for COVID-19. 2020. Reference Source\n\nAboubakr HA, Sharafeldin TA, Goyal SM: Stability of SARS-CoV-2 and other coronaviruses in the environment and on common touch surfaces and the influence of climatic conditions: A review. Transbound Emerg Dis. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorawska L, Milton DK: It Is Time to Address Airborne Transmission of Coronavirus Disease 2019 (COVID-19). Clin Infect Dis. 2020; 71(9): 2311–2313. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeung NHL, Chu DKW, Shiu EYC, et al.: Respiratory virus shedding in exhaled breath and efficacy of face masks. Nat Med. 2020; 26(5): 676–680. PubMed Abstract | Publisher Full Text\n\nSommerstein R, Fux CA, Vuichard-Gysin D, et al.: Risk of SARS-CoV-2 transmission by aerosols, the rational use of masks, and protection of healthcare workers from COVID-19. Antimicrob Resist Infect Control. 2020; 9(1): 100. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFischer EP, Fischer MC, Grass D, et al.: Low-cost measurement of face mask efficacy for filtering expelled droplets during speech. Sci Adv. 2020; 6(36): eabd3083. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChu DK, Akl EA, Duda S, et al.: Physical distancing, face masks, and eye protection to prevent person-to-person transmission of SARS-CoV-2 and COVID-19: a systematic review and meta-analysis. Lancet. 2020; 395(10242): 1973–1987. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiang M, Gao L, Cheng C, et al.: Efficacy of face mask in preventing respiratory virus transmission: A systematic review and meta-analysis. Travel Med Infect Dis. 2020; 36: 101751. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChou R, Dana T, Jungbauer R, et al.: Masks for Prevention of Respiratory Virus Infections, Including SARS-CoV-2, in Health Care and Community Settings : A Living Rapid Review. Ann Intern Med. 2020; 173(7): 542–555. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChernozhukov V, Kasahara H, Schrimpf P: Causal impact of masks, policies, behavior on early covid-19 pandemic in the U.S. medRxiv. 2020. Publisher Full Text\n\nMitze T, Kosfeld R, Rode J, et al.: to wear a face mask or not. 2020. Reference Source\n\nWang X, Ferro EG, Zhou G, et al.: Association Between Universal Masking in a Health Care System and SARS-CoV-2 Positivity Among Health Care Workers. JAMA. 2020; 324(7): 703–704. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXu J, Hussain S, Lu G, et al.: Associations of Stay-at-Home Order and Face-Masking Recommendation with Trends in Daily New Cases and Deaths of Laboratory-Confirmed COVID-19 in the United States. Explor Res Hypothesis Med. 2020; 1–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWong SH, Teoh JYC, Leung CH, et al.: COVID-19 and Public Interest in Face Mask Use. Am J Respir Crit Care Med. 2020; 202(3): 453–455. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBalancing lives and livelihoods: relaxation of lockdown needs to be finely balanced - The BMJ. 2020. Reference Source\n\nDavies NG, Kucharski AJ, Eggo RM, et al.: Effects of non-pharmaceutical interventions on COVID-19 cases, deaths, and demand for hospital services in the UK: a modelling study. Lancet Public Health. 2020; 5(7): e375–e385. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFerretti L, Wymant C, Kendall M, et al.: Quantifying SARS-CoV-2 transmission suggests epidemic control with digital contact tracing. Science. 2020; 368(6491): eabb6936. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPanovska-Griffiths J, Kerr CC, Stuart RM, et al.: Determining the optimal strategy for reopening schools, the impact of test and trace interventions, and the risk of occurrence of a second COVID-19 epidemic wave in the UK: a modelling study. Lancet Child Adolesc Health. 2020; 4(11): 817–827. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTeslya A, Pham TM, Godijk NG, et al.: Impact of self-imposed prevention measures and short-term government-imposed social distancing on mitigating and delaying a COVID-19 epidemic: A modelling study. PLoS Med. 2020; 17(7): e1003166. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGiordano G, Blanchini F, Bruno R, et al.: Modelling the COVID-19 epidemic and implementation of population-wide interventions in Italy. Nat Med. 2020; 26(6): 855–860. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFogli A, Perri F, Ponder M, et al.: Pandemic Control in ECON-EPI Networks. medRxiv. NBER Working Paper 27741. 2020. Publisher Full Text\n\nSilva PCL, Batista PVC, Lima HS, et al.: COVID-ABS: An agent-based model of COVID-19 epidemic to simulate health and economic effects of social distancing interventions. Chaos Solitons Fractals. 2020; 139: 110088. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSturniolo S, Waites W, Colbourn T, et al.: Testing, tracing and isolation in compartmental models [In Press]. PLoS Computational Biology. 2020. Publisher Full Text\n\nScientific Advisory Group for Emergencies: SAGE meetings. 2020. Reference Source\n\nKucharski AJ, Klepac P, Conlan AJK, et al.: Effectiveness of isolation, testing, contact tracing, and physical distancing on reducing transmission of SARS-CoV-2 in different settings: a mathematical modelling study. Lancet Infect. Dis. 2020; 20(10): 1151–1160. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHellewell J, Abbott S, Gimma A, et al.: Feasibility of controlling COVID-19 outbreaks by isolation of cases and contacts. Lancet Glob Health. 2020; 8(4): e488–e496. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFlaxman S, Mishra S, Gandy A, et al.: Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe. Nature. 2020; 584(7820): 257–261. PubMed Abstract | Publisher Full Text\n\nPeto J: Covid-19 mass testing facilities could end the epidemic rapidly. BMJ. 2020; 368: m1163. PubMed Abstract | Publisher Full Text\n\nPeto J, Carpenter J, Smith GD, et al.: Weekly COVID-19 testing with household quarantine and contact tracing is feasible and would probably end the epidemic. R Soc Open Sci. 2009; 7(6): 200915. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNew saliva test for coronavirus piloted in Southampton. Reference Source\n\nIacobucci G: Covid-19: Mass population testing is rolled out in Liverpool. BMJ. 2020; 371: m4268. PubMed Abstract | Publisher Full Text\n\nRoadmap to Pandemic Resilience. Reference Source\n\nRoadmap to Responsibly Reopen America. Reference Source\n\nKlepac P, Kucharski AJ, Conlan AJK, et al.: Contacts in context: large-scale setting-specific social mixing matrices from the BBC Pandemic project. medRxiv. 2020; 2020.02.16.20023754. Publisher Full Text\n\nA Compendium of Models that Predict the Spread of COVID-19. Reference Source\n\nO’Driscoll M, Santos GRD, Wang L, et al.: Age-specific mortality and immunity patterns of SARS-CoV-2. Nature. 2020. PubMed Abstract | Publisher Full Text\n\nCummins L, Ebyarimpa I, Cheetham N, et al.: Factors associated with COVID-19 related hospitalisation, critical care admission and mortality using linked primary and secondary care data. Medrxiv preprint (to appear). MEDRXIV/2020/241844.\n\nYong E: COVID-19 Can Last for Several Months. The Atlantic. 2020. Reference Source\n\nColbourn T, Waites W, Manheim D, et al.: ptti/ptti: Modelling the health and economic impacts of different testing and tracing strategies for COVID-19 in the UK (Version F1000-final). Zenodo. 2020. Publisher Full Text\n\nColbourn T, Waites W, Raine R, et al.: Modelling the Health and Economic Impacts of Population-Wide Testing, Contact Tracing and Isolation (PTTI) Strategies for COVID-19 in the UK. 2020. Publisher Full Text\n\nBloom EA, de Wit V, Carangal-San Jose MJF: Potential Economic Impact of an Avian Flu Pandemic on Asia. 2005. Reference Source\n\nCoMix study - Social contact survey in the UK. 2020. Reference Source\n\nhttps://coviddatahub.com.\n\nUnmasked retail staff putting off customers, says survey. Reference Source\n\nTaylor D, Brooks L, Halliday J: Nine out of ten public transport users wearing face masks - survey. The Guardian. 2020. Reference Source\n\nBTP. Reference Source\n\nSmith M: COVID-19: Britons still won’t wear face masks. 2020. Reference Source\n\nVerma S, Dhanak M, Frankenfield J: Visualizing the effectiveness of face masks in obstructing respiratory jets. Phys Fluids (1994). 2020; 32(6): 061708. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMacIntyre CR, Seale H, Dung TC, et al.: A cluster randomised trial of cloth masks compared with medical masks in healthcare workers. BMJ Open. 2015; 5(4): e006577. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDouglas M, Katikireddi SV, Taulbut M, et al.: Mitigating the wider health effects of covid-19 pandemic response. BMJ. 2020; 369: m1557. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRichards M, Anderson M, Carter P, et al.: The impact of the COVID-19 pandemic on cancer care. Nature Cancer. 2020; 1: 565–567.\n\nMaringe C, Spicer J, Morris M, et al.: The impact of the COVID-19 pandemic on cancer deaths due to delays in diagnosis in England, UK:a national, population-based, modelling study. Lancet Oncol. 2020; 21(8): 1023–1034. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSaxena S, Skirrow H, Bedford H: Routine vaccination during covid-19 pandemic response. BMJ. 2020; 369: m2392. PubMed Abstract | Publisher Full Text\n\nLee J: Mental health effects of school closures during COVID-19. Lancet Child Adolesc Health. 2020; 4(6): 421. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHall KS, Samari G, Garbers S, et al.: Centring sexual and reproductive health and justice in the global COVID-19 response. Lancet. 2020; 395(10231): 1175–1177. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKluge HHP, Wickramasinghe K, Rippin HL, et al.: Prevention and control of non-communicable diseases in the COVID-19 response. Lancet. 2020; 395(10238): 1678–1680. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTsai J, Wilson M: COVID-19: a potential public health problem for homeless populations. Lancet Public Health. 2020; 5(5): e186–e187. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang H, Li T, Barbarino P, et al.: Dementia care during COVID-19. Lancet. 2020; 395(10231): 1190–1191. PubMed Abstract | Publisher Full Text | Free Full Text\n\nArmitage R, Nellums LB: The COVID-19 response must be disability inclusive. Lancet Public Health. 2020; 5(5): e257. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKinner SA, et al.: Prisons and custodial settings are part of a comprehensive response to COVID-19. Lancet Public Health. 2020; 5(4): e188–e189. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartinez-Juarez LA, Sedas AC, Orcutt M, et al.: Governments and international institutions should urgently attend to the unjust disparities that COVID-19 is exposing and causing. EClinicalMedicine. 2020; 23: 100376. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAldridge RW, Lewer D, Katikireddi SV, et al.: Black, Asian and Minority Ethnic groups in England are at increased risk of death from COVID-19: indirect standardisation of NHS mortality data [version 2; peer review: 3 approved]. Wellcome Open Res. 2020; 5: 88. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRogers JP, Chesney E, Oliver D, et al.: Psychiatric and neuropsychiatric presentations associated with severe coronavirus infections: a systematic review and meta-analysis with comparison to the COVID-19 pandemic. Lancet Psychiatry. 2020; 7(7): 611–627. PubMed Abstract | Publisher Full Text | Free Full Text\n\nU.K. Population (2020)- Worldometer. Reference Source\n\nCoronavirus (COVID-19) in the UK. Reference Source\n\nDavies NG, Klepac P, Liu Y, et al.: Age-dependent effects in the transmission and control of COVID-19 epidemics. medRxiv. 2020; 2020.03.24.20043018. Publisher Full Text\n\nLauer SA, Grantz KH, Bi Q, et al.: The Incubation Period of Coronavirus Disease 2019 (COVID-19) From Publicly Reported Confirmed Cases: Estimation and Application. Ann Intern Med. 2020; 172(9): 577–582. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWilder-Smith A, Chiew CJ, Lee VJ: Can we contain the COVID-19 outbreak with the same measures as for SARS? Lancet Infect Dis. 2020; 20(5): e102–e107. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi Q, Guan X, Wu P, et al.: Early Transmission Dynamics in Wuhan, China, of Novel Coronavirus-Infected Pneumonia. N Engl J Med. 2020; 382(13): 1199–1207. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWoelfel R, Corman VM, Guggemos W, et al.: Clinical presentation and virological assessment of hospitalized cases of coronavirus disease 2019 in a travel-associated transmission cluster. Infectious Diseases (except HIV/AIDS). 2020. Publisher Full Text\n\nBhatraju PK, Ghassemieh BJ, Nichols M, et al.: Covid-19 in Critically Ill Patients in the Seattle Region - Case Series. N Engl J Med. 2020; 382(21): 2012–2022. PubMed Abstract | Publisher Full Text | Free Full Text\n\nConsiderations for quarantine of individuals in the context of containment for coronavirus disease (COVID-19). Reference Source\n\nFocus S: UK Economic Outlook. Reference Source\n\nGenie® HT OptiGene: OptiGene. Reference Source\n\nGlobal Health Data Exchange: GHDx. 2020; [cited 2020 Apr 8]. Reference Source\n\nEuropean Centre for Disease Prevention and Control: Rapid risk assessment: Coronavirus disease 2019 (COVID-19) pandemic: increased transmission in the EU/EEA and the UK - eighth update. 2020; [cited 2020 Apr 9]. Reference Source\n\nLeung K, Wu JT, Liu D, et al.: First-wave COVID-19 transmissibility and severity in China outside Hubei after control measures, and second-wave scenario planning: a modelling impact assessment. Lancet. 2020; 395(10233): 1382–1393. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEames KT, Webb C, Thomas K, et al.: Assessing the role of contact tracing in a suspected H7N2 influenza A outbreak in humans in Wales. BMC Infect Dis. 2010; 10: 141. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPollock AM, Roderick P, Cheng KK, et al.: Covid-19: why is the UK government ignoring WHO’s advice? BMJ. 2020; 368: m1284. PubMed Abstract | Publisher Full Text\n\nBowie C, Hill T: Rapid response to:Covid-19: why is the UK government ignoring WHO’s advice? BMJ. 2020; 368: m1284. Publisher Full Text\n\nWHO-China Joint Mission: Report of the WHO-China Joint Mission on Coronavirus Disease 2019 (COVID-19). 2020; [cited 2020 Apr 7]. Reference Source\n\nEuropean Centre for Disease Prevention and Control: Contact tracing: Public health management of persons, including healthcare workers, having had contact with COVID-19 cases in the European Union - first update.2020. [cited 2020 Apr 5]. Reference Source\n\nEuropean Centre for Disease Prevention and Control: Contact tracing for COVID-19: current evidence, options for scale-up and an assessment of resources needed. 2020; [cited 2020 May 8]. Reference Source\n\nFlattening the curve on COVID-19 - How Korea responded to a pandemic using ICT 상세보기NoticeConsulate General of the Republic of Korea in Houston. Reference Source\n\nPublic Health England: Best start to life and beyond; improving public health outcomes for children, young people and families. 2018. Reference Source\n\nDevon County Council: Community advice, Coronavirus advice in Devon. 2020; [cited 2020 Apr 8]. Reference Source\n\nNHS England: NHS England » NHS volunteer responders: 250,000 target smashed with three quarters of a million committing to volunteer. Reference Source\n\nEhret PJ, LaBrie JW, Santerre C, et al.: Self-affirmation and motivational interviewing: integrating perspectives to reduce resistance and increase efficacy of alcohol interventions. Health Psychol Rev. 2015; 9(1): 83–102. PubMed Abstract | Publisher Full Text\n\nMcDonald RI, Fielding KS, Louis WR: Conflicting social norms and community conservation compliance. J Nat Conserv. 2014; 22(3): 212–216. Publisher Full Text\n\nCornish F, Campbell C: The social conditions for successful peer education: a comparison of two HIV prevention programs run by sex workers in India and South Africa. Am J Community Psychol. 2009; 44(1–2): 123–135. PubMed Abstract | Publisher Full Text\n\nJaspal R, Lopes B, Maatouk I: Social Identity and Attitudes toward HIV Pre-exposure Prophylaxis: A Structural Equation Model. J Soc Serv Res. 2020; 46(3): 331–344. Publisher Full Text\n\nParker M, Hanson TM, Vandi A, et al.: Ebola and Public Authority: Saving Loved Ones in Sierra Leone. Med Anthropol. 2019; 38(5): 440–454. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarston C, Renedo A, Miles S: Community participation is crucial in a pandemic. Lancet. 2020; 395(10238): 1676–1678. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLupton (editor) D: Doing Fieldwork in a Pandemic (crowd-sourced document). Reference Source\n\nBurgess RA, Fonseca L: Re-thinking recovery in post-conflict settings: Supporting the mental well-being of communities in Colombia. Glob Public Health. 2020; 15(2): 200–219. PubMed Abstract | Publisher Full Text\n\nHargreaves J, Davey C, Group for lessons from pandemic HIV prevention for the COVID-19 response: Three lessons for the COVID-19 response from pandemic HIV. Lancet HIV. 2020; 7(5): e309–e311. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHolmes EA, O'Connor RC, Perry VH, et al.: Multidisciplinary research priorities for the COVID-19 pandemic: a call for action for mental health science. Lancet Psychiatry. 2020; 7(6): 547–560. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPeto J, Alwan NA, Godfrey KM, et al.: Universal weekly testing as the UK COVID-19 lockdown exit strategy. Lancet. 2020; 395(10234): 1420–1421. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBaker C: Population estimates & GP registers: why the difference? 2016. Reference Source\n\nUdugama B, Kadhiresan P, Kozlowski HN, et al.: Diagnosing COVID-19: The Disease and Tools for Detection. ACS Nano. 2020; 14(4): 3822–3835. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDepartment of Health and Social Care [UK government]: Coronavirus (COVID-19) Scaling up our testing programmes. 2020. Reference Source\n\nGsk: Supporting UK national effort to boost COVID-19 testing. 2020. Reference Source\n\nMehta A: Mystery surrounds UK claim of Covid-19 test reagent ‘shortage’. Chemistry World. 2020. Reference Source\n\nRobinson Brothers: Contract Manufacture. 2020. Reference Source\n\nManheim D, Foster D: Option-based guarantees to accelerate urgent, high-risk vaccines: a new market-shaping approach [version 1; peer review: 1 approved, 1 approved with reservations]. F1000Res. 2020; 9: 1154. Publisher Full Text\n\nGovernment Digital Service: Speeding penalties. 2011. Reference Source"
}
|
[
{
"id": "78273",
"date": "15 Mar 2021",
"name": "Katharina Hauck",
"expertise": [
"Reviewer Expertise Economics of infectious diseases"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper models a test-trace-isolate (TTI) strategy (testing symptomatic people, tracing and isolating everyone) and a strategy where TTI is combined with universal testing (UT; i.e. additional population testing to identify asymptomatic cases). The study projects public health impacts, as the number of infected, isolated and dead individuals, for the United Kingdom between December 2020 and May 2021. The model also projects some economic impacts, comprising of the economic costs of shutdowns in terms of GDP loss, and the direct costs of the TTI strategy. The findings are that TTI and particular a month-long UT campaign over December have a positive impact on the pandemic trajectory, and economic savings. The authors conclude that increased TTI is necessary to suppress the virus and protect the economy, and UT would have a dramatic impact and suppress the effective reproductive number below one. It is great that the authors attempt to assess the combined public health and economic impact of the pandemic and testing strategies. However, I have several concerns with the modelling that undermines my confidence in the optimistic assessment of the efficacy of TTI in sending the pandemic into reverse. Transmission of SARS-CoV-2 is modelled with a deterministic compartmental model of diagnosed and undiagnosed compartments. The model is fitted to deaths between January 2020 and October 2020, which means transmission probability, the uniform contact rate, the TTI efficacy required to constrain the pandemic, and other parameters are fitted values. A problem with fitting to earlier deaths data is that the meanwhile prevalent variants with their higher transmissibility are not considered, potentially resulting in an underestimate of transmission dynamics in the model. Neither asymptomatic infection nor varying severity are considered as separate compartments. I see this as a limitation of the model because the modelling of asymptomatic infections seems crucial in projections of a TTI strategy, the efficacy of which is strongly affected by the proportion of infected individuals coming forward to be tested. The paper assumes that 70% of cases are symptomatic. This estimate is likely too optimistic; recent studies have found that this proportion is probably much lower, maybe only 14% of cases are symptomatic.1 This would greatly reduce the number of infected that are diagnosed, and the projected efficacy of the TTI intervention in the model. A few other assumptions suggest that the efficacy of TTI and UT may be overestimated in the model, including 100% isolation of all those tested and traced is assumed. There is evidence that adherence to self-isolation is much lower, possibly as low as 18%.2 The optimistic assumption on 100% isolation would overestimate the efficacy of TTI and UT in reducing transmissions. I believe the authors would need to revisit their parameter assumptions in light of recent evidence and redo their projections before they can draw reliable policy conclusions. There is little detail given on the nature of the weekly UT intervention, which according to the findings prevents a further 90,000 deaths and reduces economic loss by £454 billion over the projection horizon. It is not clearly mentioned whether the authors assume weekly testing of the whole UK population over the month of December, i.e., 4 testing rounds of 67 million individuals? Table 1 suggests that UT would require maximum daily tests of 10 million (compared to 69,000 for TTI alone), add £1bn to the total tracing costs (over £13bn of TTI alone), require 20,000 additional tracers (over the 140,000 of TTI alone), and cost £7bn (compared to £55m of TTI). I wonder whether it is worthwhile to analyse UT of the whole population, considering that implementation of such an extensive policy seems infeasible. What about analysing UT of school children, as currently implemented in the UK, or of visitors of high-contact events?\n\nGDP loss is determined by the contact rate as an indicator of economic exchange, with the minimum ‘lockdown’ contact rate scaled to a maximum GDP loss of 15% as observed during full lockdown. The GDP loss calculation is very simplified and does not consider recent advances on the sophisticated integrated modelling of epidemiological and economic outcomes that consider differential contact rates by economic sectors, sectoral interdependencies, demand shocks, reduced work productivity due to sickness and home working, monetary valuation of loss of life, disruptions of international trade,3,4 and others. The intervention costs of TTI and UT are added to the economic assessment, but apart from that, no other costs are considered. The methods section in the appendix contains a very comprehensive and careful exposition of the testing strategies, the resources required and their costs. This part of the paper strikes me as a strong contribution to the existing literature on SARS-CoV-2 testing.\n\nOverall, I am sceptical that the optimistic conclusions on the efficacy and cost-savings associated with TTI and UT are justified by the modelling as it is currently specified. An obvious reality not considered by the model is the current vaccination rollout. The policy relevance of the study is compromised by not including vaccinations, as they will have a substantial impact on the comparative attractiveness of investments into TTI and UT.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": [
{
"c_id": "6525",
"date": "14 Apr 2021",
"name": "Tim Colbourn",
"role": "Author Response",
"response": "Author response (bold) to reviewers comments (plain text) The paper models a test-trace-isolate (TTI) strategy (testing symptomatic people, tracing and isolating everyone) and a strategy where TTI is combined with universal testing (UT; i.e. additional population testing to identify asymptomatic cases). The study projects public health impacts, as the number of infected, isolated and dead individuals, for the United Kingdom between December 2020 and May 2021. The model also projects some economic impacts, comprising of the economic costs of shutdowns in terms of GDP loss, and the direct costs of the TTI strategy. The findings are that TTI and particular a month-long UT campaign over December have a positive impact on the pandemic trajectory, and economic savings. The authors conclude that increased TTI is necessary to suppress the virus and protect the economy, and UT would have a dramatic impact and suppress the effective reproductive number below one. It is great that the authors attempt to assess the combined public health and economic impact of the pandemic and testing strategies. However, I have several concerns with the modelling that undermines my confidence in the optimistic assessment of the efficacy of TTI in sending the pandemic into reverse. Thank you for your review of our paper. We are glad you like the ambition of our paper though we disagree that the paper should be not approved based on our choice of parameter values, when those values reflect the situation in November 2020 when we finalised this paper. Further changes to the model parameters are possible via the web interface we developed for our model, associated with and referenced in the paper. The underlying compartmental model estimating TTI was published in PLOS Computational Biology on 4th March 2021. Thanks also for highlighting recent advances in joint epidemiological and economic modelling. A recent review of this was recently prepared and submitted for publication by some of the authors (Jasmina Panovska-Griffiths and Mark Sculpher) and hence we are aware of these. But we note that including these amendments is beyond the scope of this piece, which was only intended to illustrate the joint impact on health and the economy of TTI strategies. We agree that with the reality of vaccines and variants since mid-late December 2020 the nature of the epidemic has moved on considerably since we finished this paper, but while in principle we can extend the work to accommodate these, we are unable to significantly revise the model. Doing so is of course possible in theory, but unfortunately requires a complete reworking to account for several changes that have occurred which our calibration would need to change to reflect. Hence, we ask the reviewer to please view this piece of work as a standalone assessment of the COVID-19 situation in the UK in early autumn 2020, which we strongly feel is publication worthy. Transmission of SARS-CoV-2 is modelled with a deterministic compartmental model of diagnosed and undiagnosed compartments. The model is fitted to deaths between January 2020 and October 2020, which means transmission probability, the uniform contact rate, the TTI efficacy required to constrain the pandemic, and other parameters are fitted values. A problem with fitting to earlier deaths data is that the meanwhile prevalent variants with their higher transmissibility are not considered, potentially resulting in an underestimate of transmission dynamics in the model. As per our previous comment, this study was completed in mid-November 2020, before the emergence of the B.1.1.7. UK (Kent) variant, or any other virus variants and hence the spillover effect from these on transmissibility and severity is not incorporated. We therefore do not consider this to be an issue making our paper unworthy of publication. Neither asymptomatic infection nor varying severity are considered as separate compartments. I see this as a limitation of the model because the modelling of asymptomatic infections seems crucial in projections of a TTI strategy, the efficacy of which is strongly affected by the proportion of infected individuals coming forward to be tested. Our universal testing (UT) scenario covers asymptomatic and symptomatic infected people alike and does not require individuals to “come forward”. It is akin to the twice weekly LFD tests currently recommended for all secondary school children to identify asymptomatic as well as symptomatic cases. For testing targeted by symptoms, this simply means that a smaller proportion of cases are tested. It is correct that we do not represent a difference in propensity to transmit between these two groups as that was not known at the time and there remains uncertainty about this. The paper assumes that 70% of cases are symptomatic. This estimate is likely too optimistic; recent studies have found that this proportion is probably much lower, maybe only 14% of cases are symptomatic.1 This would greatly reduce the number of infected that are diagnosed, and the projected efficacy of the TTI intervention in the model. This assumption is in line with all modelling studies to date- both published and preprints. We agree with the reviewer that if the proportion who were symptomatic was lower then TTI would indeed be less effective, though UT would remain effective as would still identify cases whether they were asymptomatic or symptomatic. A few other assumptions suggest that the efficacy of TTI and UT may be overestimated in the model, including 100% isolation of all those tested and traced is assumed. There is evidence that adherence to self-isolation is much lower, possibly as low as 18%.2 The optimistic assumption on 100% isolation would overestimate the efficacy of TTI and UT in reducing transmissions. I believe the authors would need to revisit their parameter assumptions in light of recent evidence and redo their projections before they can draw reliable policy conclusions. Our assumption of 100% isolation is a modelling simplification only. Lower proportions isolated are easily though implicitly captured by lowering the testing or tracing rates. We refer the reviewer and readers to the accompanying web interface linked in the paper (generated with Streamlit) where the testing rate, tracing rate and other parameters can be set. We appreciate the reviewer's comments that a broader parameter sweep across the n-dimensional parameter space would give uncertainty bounds and allow us to incorporate a number of other scenarios including balancing test-trace and isolate assumptions. Undertaking such parameter sweeps would only be useful if we were able to rework the model in light of the new reality of vaccines and variants and consequent new policy questions. All of this is beyond the scope of this paper however unfortunately. There is little detail given on the nature of the weekly UT intervention, which according to the findings prevents a further 90,000 deaths and reduces economic loss by £454 billion over the projection horizon. It is not clearly mentioned whether the authors assume weekly testing of the whole UK population over the month of December, i.e., 4 testing rounds of 67 million individuals? Table 1 suggests that UT would require maximum daily tests of 10 million (compared to 69,000 for TTI alone), add £1bn to the total tracing costs (over £13bn of TTI alone), require 20,000 additional tracers (over the 140,000 of TTI alone), and cost £7bn (compared to £55m of TTI). I wonder whether it is worthwhile to analyse UT of the whole population, considering that implementation of such an extensive policy seems infeasible. Testing the whole population regularly was proposed and discussed by the UK government as a potential option via “Operation Moonshot” so we don’t believe it was beyond the realms of feasibility. Therefore we stand by our decision to model it- again noting we are simulating specific modelling scenarios and not predicting current or future reality. What about analysing UT of school children, as currently implemented in the UK, or of visitors of high-contact events? Our model does not distinguish between different groups of the population such as school children or visitors of high-contact events, therefore whilst this would indeed be useful to include, it is beyond the scope of our work. GDP loss is determined by the contact rate as an indicator of economic exchange, with the minimum ‘lockdown’ contact rate scaled to a maximum GDP loss of 15% as observed during full lockdown. The GDP loss calculation is very simplified and does not consider recent advances on the sophisticated integrated modelling of epidemiological and economic outcomes that consider differential contact rates by economic sectors, sectoral interdependencies, demand shocks, reduced work productivity due to sickness and home working, monetary valuation of loss of life, disruptions of international trade,3,4 and others. Thank you for highlighting this recent and more advanced work. Our work was only intended to be a simple model to highlight the parallels between health outcomes and economic outcomes and how TTI/UT could be used to achieve both health and economic benefits. It is only a first-order approximation of the economic effects and it is possible to imagine more sophisticated models that capture higher-order effects and feedback mechanisms. These are beyond the scope of our analysis. The intervention costs of TTI and UT are added to the economic assessment, but apart from that, no other costs are considered. The methods section in the appendix contains a very comprehensive and careful exposition of the testing strategies, the resources required and their costs. This part of the paper strikes me as a strong contribution to the existing literature on SARS-CoV-2 testing. Thank you – we hope this, and our responses to your other points, would be sufficient for you to “approve” the paper, with reservations. Given we are out of time to work on it further the paper may otherwise remain as a pre- print / grey literature. Overall, I am sceptical that the optimistic conclusions on the efficacy and cost-savings associated with TTI and UT are justified by the modelling as it is currently specified. An obvious reality not considered by the model is the current vaccination rollout. The policy relevance of the study is compromised by not including vaccinations, as they will have a substantial impact on the comparative attractiveness of investments into TTI and UT. We again stress to the reviewer that our paper was written before any vaccinations were approved and indeed before vaccine trial results were published, so we were unable to include vaccinations in our study. Is the work clearly and accurately presented and does it cite the current literature? Partly Is the study design appropriate and is the work technically sound? Partly Are sufficient details of methods and analysis provided to allow replication by others? Partly If applicable, is the statistical analysis and its interpretation appropriate? Partly Are all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions drawn adequately supported by the results? No We disagree that differences in parameter values should be grounds to say the conclusions are not supported by the results. We believe the conclusions do follow from our results even if you disagree with some of the specific results as indicated above. We also refer the reviewer to our web interface which allows different parameter values to be set and simulations re-run. Tim Colbourn, David Manheim, William Waites, Jasmina Panovska-Griffiths"
}
]
}
] | 1
|
https://f1000research.com/articles/9-1454
|
https://f1000research.com/articles/8-1786/v1
|
22 Oct 19
|
{
"type": "Case Report",
"title": "Case Report: A castration-resistant terminal prostate cancer patient’s survival prolonged after practicing Falun Gong",
"authors": [
"Yu-Hong Dong",
"Shawn Wu",
"Ann Corson",
"Kai-Hsiung Hsu",
"Yu-Hong Dong",
"Shawn Wu",
"Ann Corson"
],
"abstract": "Background: Most metastatic prostate cancer patients receive androgen deprivation therapy (ADT) as the mainstay of treatment. Unfortunately, patients ultimately progress to castration resistance. Clinical finding and diagnosis: We describe a man in his eighties who developed stage IV, M1b prostate cancer, and multiple (≥5) bone metastases who required the aid of a walker to ambulate. Without treatment, his treating physician predicted he would survive 6 months. Interventions and outcomes: The patient initially responded well to treatment with ADT, but during the 80-89th week of treatment he developed castration resistance. ADT was then discontinued. He subsequently began practicing Falun Gong (FLG) as an alternative form of care. Within one year of constant practice, he became able to walk independently, his bone metastases disappeared, and he also enjoyed better psychosocial functioning. His treating physician assessed that his prostate malignancy was “clinically, under control” and “his overall functional status is excellent.” The patient survived a total of 263 weeks (61.4 months) post diagnosis, including 174 weeks (40.6 months) after developing castration resistance. Conclusion: This castration-resistant terminal prostate cancer patient gained significant clinical benefits after practicing Falun Gong.",
"keywords": [
"Castration-resistant prostate cancer",
"Metastatic prostate cancer",
"Survival",
"Falun Gong",
"Eudaimonic well-being",
"Alternative medicine",
"Mind-body Medicine",
"Mindfulness"
],
"content": "Introduction\n\nProstate cancer is the second most common cancera and the fifth leading cause of cancer death in menb worldwide. In the United States, it is the most prevalent cancer and the second leading cause of cancer death in menc. The five-year overall survival in metastatic prostate cancer patients aged ≥ 75 years is 13.6%1. Androgen deprivation therapy (ADT), either by means of bilateral orchiectomy or the administration of a gonadotropin-releasing hormone (GnRH) agonist or antagonist, is the mainstay of treatment for metastatic prostate cancer2. In most metastatic prostate cancer patients, ADT is initially effective, but its response is not sustained and patients ultimately progress to castration resistance3.\n\nIn recent years there has been a growing awareness of oriental meditative approaches for health and wellness in the West. Deeply rooted in thousands of years of Chinese cultural history, Qigong is a traditional Chinese mind-body practice that uses qi (vital energy) and consciousness cultivation to achieve an optimal health state. Falun Gong (FLG) is a unique type of ancient qigong practice based on the principles of “truthfulness–compassion–tolerance”. The practice of FLG includes reading the main teachings, Zhuan Falund and performing five sets of slow, gentle, meditative exercisese.\n\nRecently, we prospectively followed a terminal stage IV prostate cancer patient in his eighties, who relapsed after ADT and then began practicing FLG. This is the first well documented castration-resistant prostate cancer case whose survival was significantly prolonged after practicing FLG.\n\n\nCase report\n\nAn 80-year-old Chinese man was a retired professor at a university in the USA. He presented to a urology clinic in 2014 complaining of bone pain, fatigue, and an elevated prostate specific antigen (PSA; ranging from 94.7 to 126.6 ng/ml; normal range: 0.6–12.4 ng/ml for 75–79 years born in China4) for one year. He had a 20-year history of hypothyroidism treated with Synthroid™ (75µg/day), a hernia operation in 2003, and an implanted pacemaker in 2011. He had been diagnosed with benign prostate hyperplasia in 2004 and his baseline PSA values ranged from 4.1 to 8.6 ng/ml between 1996 and 2006. There was a family history of cancer. The patient required the aid of a walker to ambulate. Digital rectal exam (DRE) revealed a “very hard, very nodular and fixed prostate.” PSA was 468 ng/ml on presentation. Transrectal ultrasound image showed the prostate to be 63×57×53 mm3 with disrupted capsule, abnormal inhomogeneous peripheral zone, scattered calcifications, and no median lobe. The prostate biopsy pathology report revealed an adenocarcinoma with a Gleason Score 9 (4+5 or 5+4) in all lobes with 40–80% cancer infiltration (Figure 1). Whole-body bone scan was highly suspicious for numerous osseous metastases in the lumbar and thoracic spine as well as in the left iliac bone of the pelvis, some areas of the sacrum and in the left 9th rib (Figure 2). CT scan findings were consistent with the bone scan (Figure 2). The patient was diagnosed with terminal stage IV, M1b metastatic prostate adenocarcinoma with extensive (≥5) bone metastases. His treating physician predicted that without effective treatment he would have a life expectancy of only 6 months.\n\nThe adenocarcinoma consists of poorly formed or fused glandular acini (Gleason grade 4) and aggregates, cords and solid foci not recognizable as acini (Gleason grade 5), resulting in the Gleason score of 9 (4+5). The Gleason score is the sum of two most prevalent histologic grades in a prostate adenocarcinoma, rated on a scale of 1-5, with 5 being the most clinically aggressive.\n\nUpper Row: before FLG practice (May-2014); Lower Row: after FLG practice (Nov-2016); Left and Middle Columns: anterior and posterior views of the bone scan; Right Column: post-processed (Multiplanar Reformation) Spine CT. The yellow arrows denote the normal uptake level of tracer corresponding to the original bone metastatic foci in the upper row, as shown at the sites of the left 9th rib region, thoracic vertebra (T7, also shown on CT image), sacral vertebrae region, and left iliac region, indicating the disappearance of cancer. The blue arrows denote degenerative bone regions with mild increased tracer uptake in sternum region and lumbar spine region but without evidence of neoplastic lesions. Degenerative appearances of bone and joint were also observed in the upper and lower extremities (bone scan) as well as in the spine (CT image).\n\nThe patient was started on degarelix (240mg, 80mg, 80mg at month intervals) and bicalutamide 50mg daily. Survival was monitored by symptoms and serum PSA. Quality of life (QoL) was assessed with the SF-8™ questionnaire (Table 1). Psychosocial functioning was evaluated using an adapted, simplified six-part questionnaire that assessed the patient’s status of positive thinking, happiness, altruism, emotional control and purpose of life. All these parameters have been systemically researched in a number of mindfulness psychological studies and commonly used in psychosocial functioning outcomes5 (Table 2).\n\n* After onset of cancer and before practicing FLG, the patient was short of breath after climbing a 14-step staircase and had to rest midway to catch his breath. At week 187, he could go up the same 14-step staircase without stopping halfway and was no longer short of breath after reaching the top.\n\nThe patient’s bone pain was gone within 5 days. After five weeks, his PSA dropped to 4.5 ng/ml. At week 14, degarelix was replaced with leuprolide acetate at a dose of 22.5 mg quarterly. PSA levels remained at 0-0.2 ng/ml until week 68. At week 80, his PSA rose to 1.4 ng/ml, and his physician discontinued bicalutamide due to suspicion of cancer relapse. PSA continued to increase to 2.4 ng/ml at week 84 and to 3.7 ng/ml at week 89. (Figure 3, Table 3)\n\nIn 1996-2006 the patient’s PSA was 4.1-8.6 ng/ml. Before ADT, his PSA was 468 ng/ml. PSA decreased to 0 within five weeks after ADT and remained at < 0.9 ng/ml. At week 80, PSA increased to 1.4 ng/ml, and bicalutamide was stopped. At week 89, PSA increased to 3.7 ng/ml, and he started FLG practice. PSA declined to 0.6 ng/ml at week 94. The last PSA were 4.4–8.2 ng/ml, within his baseline levels (4.1–8.6 ng/ml) in 1996–2006. Each point stands for the PSA recorded at various times and the step size in time is not even.\n\nThe physician planned to use abiraterone acetate plus prednisone or enzalutamide if the PSA rose to 10 ng/ml. The patient was not pleased with either choice because of the increased risk of side effects and limited survival benefits. ADT was then discontinued. At week 89, the patient began practicing FLG. When he watched the FLG video for the first time, the patient reported that he felt as if substantial high energy matter was plugged into his two lower legs. He could sense the flow of energy and the warmth induced by the energy during practicing. In the past, during cold winter nights, the patient had a tendency to get cramps in his legs. To prevent this, he had been using a hot water bottle under the quilt while sleeping for many years. Soon after starting to practice FLG, the patient reported he did not need a hot water bottle anymore.\n\nAt week 94, the patient’s PSA reversed its increasing trend by dropping to 0.6 ng/ml (Figure 3, Table 3). At week 106, he no longer needed the assistance of a walker. At weeks 130 and 133, no osseous metastases were seen on either CT or whole-body bone scan (Figure 2). At week 151, DRE revealed that the prostate was softer and smaller than before.\n\nFollowing ADT, the patient experienced adverse events including right foot swelling, wheezing, and hot flashes for about one year, all of which, the patient reported, had gradually disappeared after practicing FLG.\n\nThe patient’s QoL improved greatly with FLG practice (Table 1). He was even able to go up a 14-step stairway without stopping halfway or developing shortness of breath with exertion on stairs he had before practicing FLG (Extended data: Video S1). Additionally, the patient reported the following sequential trend of psychosocial functioning improvement: starting FLG practice, finding the purpose of life, developing positive thinking, having a better temperament, experiencing symptom improvement, feeling happier and less depressed, and becoming more altruistic (Table 2). At week 219, based on the extended survival, absence of symptoms, disappearance of bone metastases, improved QoL and psychosocial functioning, his treating physician assessed that his prostate malignancy was “clinically, under control” and “his overall functional status is excellent.” The patient reported to be living a healthy life with good physical and mental status. He shared his experience as a terminal cancer survivor in the hospital support group as well as with the local medical community.\n\nOn week 228, the patient had a car accident which resulted in a severe neck injury and decreased ability to walk. On week 263, the patient was found dead at home from unknown causes. Three days before his death, he had spoken with us and denied any signs or symptoms of cancer relapse.\n\n\nDiscussion\n\nThe case presented here is a castration-resistant metastatic prostate cancer patient in his eighties. Prostate cancer patients receiving ADT inevitably develop castration-resistance within a median relapse time of 18 months6 due to prostate cancer cells developing mechanisms to proliferate despite castrate levels of testosterone. M1 castrate-resistant prostate cancer patients with a high number of skeletal metastasis (≥5) have a considerably poorer progression-free survival of 8.4 months and overall survival (OS) of 18.7 months, compared to patients with 1-4 bone metastases7.\n\nThe terminal prostate cancer patient in our case had a Gleason Score 9, a PSA of 468 ng/ml, and high number of bone metastases (≥5). Even with an initial response to ADT treatment, he still became castration resistant. After starting FLG practice, his PSA came down, he could walk independently, and his bone metastases disappeared within a short time frame. After developing castrate resistance and stopping ADT, he survived for an additional 174 weeks (40.6 months), which is far longer than the reported OS of 18.7 months in patients with similar clinical conditions7. The close temporal relationship between FLG practice and significant cancer improvement and the absence of any other treatment after starting FLG practice suggest that FLG practice resulted in a holistic and beneficial effect in this terminal prostate cancer patient. Moreover, the patient reported a good QoL free of cancer symptoms until he died of a non-cancer related cause, possibly due to the sequela of his severe car accident and/or his pacemaker-related cardiac condition.\n\nAs the pituitary-gonadal system is usually restored within 4 to 12 weeks after leuprolide acetate is discontinuedf, the later rise in the patient’s most recent PSA level might reflect his baseline PSA levels of 20 years ago. Additionally, PSA levels are also known to increase with age. A cohort study (the Concord Health and Ageing in Men Project, CHAMP) involving a representative sample of 1434 eligible community-dwelling men with no diagnosis of prostate cancer reported a 5th to 95th percentile range of 0.6–12.4 ng/ml of serum PSA level in men aged 75–79 and born in China4.\n\nWhile there are anecdotical testimonies from terminal cancer patients revealing they survived longer than expected after practicing FLG8,9, this is the first case report documented with detailed medical records. According to the canon of traditional Chinese medicine, the Yellow Emperor’s Classic of Medicine, people may enjoy a long, disease-free life by cultivating the mind and following upright moral principles10. The emerging discipline of psychoneuroimmunology has begun to disclose the relationship between psychological, neurological and immunological systems. For example, eudaimonic well–being is found to be associated with decreased expression of transcriptional response to adversity profile involving inflammation mediated neoplastic diseases11, decreased tumour growth and progression–related norepinephrine12. The patient presented here experienced symptom improvement soon after reading Zhuan Falun and accepting the principles of “truthfulness–compassion–tolerance”, suggesting that the improvement of moral and spiritual character also plays a significant role in the improvement of health.\n\n\nConclusions\n\nWe describe in detail for the first time in the medical literature, the clinical course of a terminal prostate cancer case with relapse to ADT followed by FLG practice, an ancient Qigong practice originating from China. Most of the patient’s prostate cancer signs and symptoms disappeared quickly after FLG practice, accompanied with greatly improved psychosocial functioning and quality of life. We conclude that this terminal prostate cancer case gained clinical benefits through practicing FLG. Although the mechanism by which FLG practice caused improvement is not yet well understood, its effects are identifiable and measurable. Thus, the practice of FLG deserves more investigation by the medical community.\n\n\nPatient perspective\n\n“My name is XXX. I had a teaching career as an Organic Chemistry Professor for 37 years at XXX before retiring in 2000. In May of 2014, I became a final stage, metastatic prostate cancer patient with bone pain, a 468 PSA, and an estimated life expectancy of less than half a year. I also required the aid of a walker to ambulate. After 19 months of androgen deprivation therapy (ADT), my cancer relapsed. Shortly prior to that relapse, I was very fortunate to have discovered a relatively new but rapidly growing mind and body cultivation practice known as Falun Gong that came originally from China.\n\nOn January 2016, I began to seriously engage in FLG cultivation practices involving reading moral teachings from a principle guidebook of Falun Dafa entitled “Zhuan Falun” and performing five different sets of Qigong exercises daily. A little over a month later, my PSA reversed its rising trend and dropped immediately to 0.6. Three months after that, I took my last Lupron treatment, thus ending the ADT therapy. After another five months, a new bone scan revealed the disappearance of metastases lesions. Today, I have been off any medicine, including cancer fighting medicine for 22 months. I am a healthy person with normal quality of life, no bone pain, full of energy, better tempered, more optimistic and positive. My own experience so far has demonstrated to my own satisfaction that cultivation of mind and body through practicing Falun Gong can be an effective healing and cost-effective way of combating cancer which is also free of undesirable side effects…\n\nThe magical effect of practicing Falun Gong had on my health made me appreciate the importance of elevating one’s moral character to be in tune with the fundamental characteristics of the universe, namely, truthfulness, compassion, and tolerance. It also made me realize that there may very well be a set of supernormal science on human body that is quite unique, subtle, and highly deserving of greater attention and further studies by modern scientists.”\n\nPatient’s Signature, July 2018\n\n\nPhysician perspective\n\nWednesday, July 2018: Patient seen on July XX. Overall, he is doing well. He is not depressed. He has no pain. He is currently on no medical therapy for his prostate cancer. He is optimistic and functional. All other systems reviewed and otherwise negative.\n\nAssessment: Prostate malignancy: Clinically, under control. His overall functional status is excellent. Patient is comfortable with his spiritual approach to care.\n\n\nConsent\n\nWritten informed consent for the publication of this case report and any associated images was obtained from the patient before his death.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\nFigshare: A castration-resistant terminal prostate cancer patient survived significantly longer and walked independently after practicing Falun Gong, https://doi.org/10.6084/m9.figshare.993329613\n\nThis project contains the following extended data:\n\n- Video S1: This patient was diagnosed with terminal stage IV, M1b metastatic prostate adenocarcinoma with extensive (≥5) bone metastases requiring the aid of a walker to ambulate. During treatment, he developed castration-resistance. After practicing Falun Gong, his bone metastases disappeared and he became able to walk independently. This video was taken three and a half years after the diagnosis of terminal cancer was made and less than two years after he started to practice Falun Gong. The patient walked up a 14-step stairway without stopping halfway and did not experience the shortness of breath with exertion on stairs he had before practicing Falun Gong.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Acknowledgements\n\nWe thank Dr. Robert A. Goulart for sharing the pathological photos and reviewing the figure legend for the pathological photos in this paper. We thank Dr. Pamela Ellsworth for giving suggestions for the manuscript. We thank Rong-Sen Yang, MD, PhD, Dr. Chian-Feng Huang and Dr. Shi-Wei Huang for reviewing the manuscript. We thank Ms. June Fakkert for proof–reading.\n\n\nFootnotes\n\nahttp://gco.iarc.fr/today/online-analysis-pie?v=2018&mode=cancer&mode_population=continents&population=900&populations=900&key=total&sex=1&cancer=39&type=0&statistic=5&prevalence=0&population_group=0&ages_group%5B%5D=0&ages_group%5B%5D=17&nb_items=7&group_cancer=1&include_nmsc=1&include_nmsc_other=1&half_pie=0&donut=0&population_group_globocan_id=\n\nbhttp://gco.iarc.fr/today/online-analysis-pie?v=2018&mode=cancer&mode_population=continents&population=900&populations=900&key=total&sex=1&cancer=39&type=1&statistic=5&prevalence=0&population_group=0&ages_group%5B%5D=0&ages_group%5B%5D=17&nb_items=7&group_cancer=1&include_nmsc=1&include_nmsc_other=1&half_pie=0&donut=0&population_group_globocan_id=\n\nchttps://gis.cdc.gov/Cancer/USCS/DataViz.html\n\ndhttps://falundafa.org/eng/eng/pdf/ZFL2014.pdf\n\nehttp://en.falundafa.org/eng/html/dymf_2014/dymf_2014_2.htm\n\nfhttps://www.accessdata.fda.gov/drugsatfda_docs/label/2014/020517s036_019732s041lbl.pdf\n\n\nReferences\n\nKeating MJ, Giscombe L, Tannous T, et al.: Age-dependent overall survival benefit of androgen deprivation therapy for metastatic prostate cancer. J Oncol Pharm Pract. 2019; 1078155219835597. PubMed Abstract | Publisher Full Text\n\nLoblaw DA, Virgo KS, Nam R, et al.: Initial hormonal management of androgen-sensitive metastatic, recurrent, or progressive prostate cancer: 2006 update of an American Society of Clinical Oncology practice guideline. J Clin Oncol. 2007; 25(12): 1596–605. PubMed Abstract | Publisher Full Text\n\nRitch C, Cookson M: Recent trends in the management of advanced prostate cancer [version 1; peer review: 3 approved]. F1000Res. 2018; 7: pii: F1000 Faculty Rev-1513. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLitchfield MJ, Cumming RG, Smith DP, et al.: Prostate-specific antigen levels in men aged 70 years and over: findings from the CHAMP study. Med J Aust. 2012; 196(6): 395–8. PubMed Abstract | Publisher Full Text\n\nLuberto CM, Shinday N, Song R, et al.: A Systematic Review and Meta-analysis of the Effects of Meditation on Empathy, Compassion, and Prosocial Behaviors. Mindfulness (N Y). 2018; 9(3): 708–724. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCrawford ED, Eisenberger MA, McLeod DG, et al.: A controlled trial of leuprolide with and without flutamide in prostatic carcinoma. N Engl J Med. 1989; 321(7): 419–24. PubMed Abstract | Publisher Full Text\n\nTait C, Moore D, Hodgson C, et al.: Quantification of skeletal metastases in castrate-resistant prostate cancer predicts progression-free and overall survival. BJU Int. 2014; 114(6b): E70–E73. PubMed Abstract | Publisher Full Text\n\nClearwisdom editors: Life and Hope Renewed: The Healing Power of Falun Dafa. Mahwah, New Jersy: MINGHUI Publishing. 2008. Reference Source\n\nMcCoy WF, Zhang L: A Journey to Ultimate Health: Falun Gong Stories. First printing edition, Golden Lotus Press, 1998. Reference Source\n\nMaoshing N: The Yellow Emperor's Classic of Medicine: A New Translation of the Neijing Suwen with Commentary. Boston, Massachusetts: Shambhala Publications. 1995; 1–2. Reference Source\n\nFredrickson BL, Grewen KM, Coffey KA, et al.: A functional genomic perspective on human well-being. Proc Natl Acad Sci U S A. 2013; 110(33): 13684–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDavis LZ, Slavich GM, Thaker PH, et al.: Eudaimonic well-being and tumor norepinephrine in patients with epithelial ovarian cancer. Cancer. 2015; 121(19): 3543–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDong YH, Wu S, Corson A, et al.: Figshare: A castration-resistant terminal prostate cancer patient survived significantly longer and walked independently after practicing Falun Gong. 2019. http://www.doi.org/10.6084/m9.figshare.9933296"
}
|
[
{
"id": "56932",
"date": "02 Dec 2019",
"name": "Laurence Klotz",
"expertise": [
"Reviewer Expertise Prostate cancer"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis case could readily be explained by prolonged androgen suppression, which is common in elderly men after discontinuation of LHRH agonist. In the absence of serum testosterone levels after ADT was stopped, the case is not convincing. While I'm open to the effect of interventions such as this, it requires very convincing evidence.\n\nOn this basis I'd elect to reject this paper.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? No\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? No",
"responses": [
{
"c_id": "5134",
"date": "30 Dec 2019",
"name": "Yuhong Dong",
"role": "Author Response",
"response": "We thank Dr. Klotz for his valuable review, and summarize his questions into three parts and answer them accordingly below:(1) This case could readily be explained by prolonged androgen suppression, which is common in elderly men after discontinuation of LHRH agonist.Answer:In this case, the use of LHRH agonist (leuprolide acetate) was discontinued at week 116 while the use of antiandrogen (bicalutamide) was discontinued at week 80. The reviewer seemed to attribute the cascade of clinical responses that occurred in the patient to the discontinuation of LHRH agonist therapy. However, the patient already started to walk independently at week 106, indicating improvement in his bone metastases occurred before discontinuation of LHRH agonist and could not be attributed to the effect of prolonged androgen suppression.Alternatively, the reviewer may refer to Antiandrogen Withdrawal Syndrome (AAWS) in the question instead of prolonged androgen suppression. AAWS, usually defined as >50% decline in PSA following cessation of the antiandrogen, has been increasingly recognized and reviewed1,2,3. A large multi-institutional clinical trial (CALGB 9583) to investigate AAWS in androgen-independent prostate cancer patients (with a median age of 72 years) has showed that 11% of patients experienced AAWS with the median time to PSA progression at 5.9 months, and objective responses in measurable disease were observed in 2% of patients4. In this case in his eighties, the time to PSA progression is 19.8 months (week 94-179) following the same criteria in that trial, while the survival time is 42.3 months (week 80-263), significantly greater than the median survival time of 16.7 months in the trial. We believe that even if AAWS might have played some role in this patient’s clinical course, AAWS was not the sole factor contributing to longer survival and time to PSA progression. As indicated in that trial, patients in the AAWS + ketoconazole treatment arm have greater median time to PSA progression (8.6 months) and objective response rate (20%) than do patients in the AAWS alone arm (5.9 months). FLG practice in this case started at week 89 might play a role comparable to ketoconazole therapy in the CALGB 9583 trial by adding anticancer activity. However, this enhancing role of ketoconazole therapy does not help prolong the median survival time (15.3 months) in the AAWS + ketoconazole treatment arm, actually a bit shorter than that in the AAWS alone arm (16.7 months). In contrast, FLG practice did help prolong the survival time to at least 42.3 months.More importantly, without any effective drug treatment after developing castration resistance, the patent in this case lived without any symptoms, in excellent mental and spiritual health with a good quality of life without any cancer complications (such as bone pain, skeletal-related events, and urinary symptoms) or side effects associated with drug therapies. Additionally, a series of improvements in bone metastases occurred sequentially at week 106 (ambulating without a walker), week 130 (CT not showing osseous metastatic findings), week 132 (bone scan showing disappearance of bone metastases) and week 187 (walking up a 14-step stairway without stopping halfway). It is unlikely that all these beneficial effects could be attributed to AAWS alone. Holistic mind-body improvement, however, is commonly observed in the practice of FLG.(2) In the absence of serum testosterone levels after ADT was stopped, the case is not convincing.Answer:In Table 3, serum testosterone levels were tested at 44, 31, <3 and <3 (ng/dL) at week 1, 5, 84 and 89, respectively. Testosterone levels dropped quickly at the beginning of ADT and stayed low (<3 ng/dL) after discontinuing antiandrogen at week 84 and 89, as expected with the use of the LHRH agonist. However, PSA levels progressively rose between week 68 (0.2 ng/ml) and week 89 (3.7 ng/ml) despite low testosterone, indicating the cancer had become resistant to ADT. Thus, the testosterone level was not considered to have any role in the subsequent clinical events.(3) “Are enough details provided of any physical examination and diagnostic tests, treatment given and outcomes?” “Is the case presented with sufficient detail to be useful for other practitioners?” The reviewer ticked “no” to both questions.Answer:Based on this case’s well documented hospital records, we have provided adequate physical examinations, diagnostic tests (biopsy, PSA levels, testosterone levels, bone scan, CT scan), details of the treatments given, as well as the clinical outcomes. We believe this case report has been presented with ample medical details.Clinical physicians and patients are likely to appreciate the significance of this case study’s findings, especially if they consider the following three perspectives:I. SurvivalSuch a significantly prolonged survival time is rarely reported in the medical literature and should be of great interest for those physicians who has managed similar patients.In the CHAARTED trial, a high volume of metastatic prostate cancer was defined by the presence of visceral metastases or four or more bone lesions with at least one beyond the vertebral bodies and pelvis5. This 80-year-old patient with a Gleason score of 9 met the definition of high volume metastatic prostate cancer as he had 5 bone metastases including one site beyond the vertebral body and pelvis (left 9th rib region). In the CHAARTED trial, the median life expectancy for an analogous patient was 32.2 months with patients seldom surviving longer than 60 months. Despite the multiple effective treatment for castration resistant prostate cancer, such as chemotherapy and second-generation anti-androgen therapy (zytiga, xtandi), provided by in this large clinical trial, the median life expectancy was just 32.2 months. In the non-academic medical community, the median life expectancy may be even shorter, perhaps around 27 months6. The patient in this study, who practiced FLG, survived a total of 61.4 months (263 weeks) post-diagnosis, which is longer than comparable patients in both the CALGB 9583 and CHAARTED studies.II. Quality of lifeWithout any effective drug treatment options after developing castration resistance, the patent in this case lived without any symptoms, in excellent mental and spiritual health, with a good quality of life, without any cancer complications (such as bone pain, skeletal-related events, and urinary symptoms), or medication side effects. We believe that this case has been presented with sufficient evidence of the valuable clinical benefits this patient experienced from practicing FLG. This information is useful for physicians and patients alike who are searching for effective treatment options.III. CostFLG practice is free, and anyone can start or stop the practice on his own. This may alleviate physicians and patients’ financial concerns.The patient himself was a rigorous scientist who received orthodox scientific training and had a very good reputation in the organic chemistry community. His personal experience of practicing Falun Gong made him realize that “there may very well be a set of supernormal science on human body that is quite unique, subtle, and highly deserving of greater attention and further studies by modern scientists.”We appreciate Dr. Klotz’s comments and hope this response answers his questions and concerns properly. References: Sartor AO, Tangen CM, Hussain MHA, et al.: Antiandrogen withdrawal in castrate-refractory prostate cancer: a Southwest Oncology Group trial (SWOG 9426). Cancer 2008; 112(11): 2393–2400. PubMed l Full text Leone G, Tucci M, Buttigliero C, et al.: Antiandrogen withdrawal syndrome (AAWS) in the treatment of patients with prostate cancer. Endocr Relat Cancer. 2018;25(1):R1-R9. PubMed l Full text Lau YK, Chadha MK, Litwin A, Trump DL: A dramatic, objective antiandrogen withdrawal response: case report and review of the literature. J Hematol Oncol. 2008; 5(1):21. PubMed l Full text Small EJ, Halabi S, Dawson NA, et al.: Antiandrogen withdrawal alone or in combination with ketoconazole in androgen-independent prostate cancer patients: a phase III trial (CALGB 9583). J Clin Oncol 2004; 22 (6):1025-33. PubMed l Full text Sweeney CJ, Chen YH, Carducci M, et al.: Chemohormonal therapy in metastatic hormone-sensitive prostate cancer. N Engl J Med. 2015;373(8):737-46. PubMed l Full text Alhanafy AM, Zanaty F, Ibrahem R, Omar S. Prognostic Factors for Hormone Sensitive Metastatic Prostate Cancer: Impact of Disease Volume. Asian Pac J Cancer Prev. 2018;19(4):1113-1118. PubMed l Full text"
}
]
}
] | 1
|
https://f1000research.com/articles/8-1786
|
https://f1000research.com/articles/9-1447/v1
|
11 Dec 20
|
{
"type": "Method Article",
"title": "A pipeline to analyse time-course gene expression data",
"authors": [
"Nelle Varoquaux",
"Elizabeth Purdom"
],
"abstract": "The phenotypic diversity of cells is governed by a complex equilibrium between their genetic identity and their environmental interactions: Understanding the dynamics of gene expression is a fundamental question of biology. However, analysing time-course transcriptomic data raises unique challenging statistical and computational questions, requiring the development of novel methods and software. This workflow provides a step-by-step tutorial of the methodology used to analyse time-course data: (1) quality control and normalization of the dataset; (2) differential expression analysis using functional data analysis; (3) clustering of time-course data; (4) interpreting clusters with GO term and KEGG pathway enrichment analysis. As a case study, we apply this workflow to time-course transcriptomic data from mice exposed to four strains of influenza to showcase every step of the pipeline.",
"keywords": [
"time-course gene expression data",
"clustering",
"differential expression",
"workflow"
],
"content": "Introduction\n\nGene expression studies provide simultaneous quantification of the level of mRNA from all genes in a sample. High-throughput studies of gene expression have a long history, starting with microarray technologies in the 1990s through to single-cell technologies. While many expression studies are designed to compare the gene expression between distinct groups, there is also a long history of time-course expression studies. Such studies compare gene expression across time by measuring mRNA levels from samples collected at different timepoints1. Such time-course studies can vary from measuring a few distinct timepoints, to sampling ten to 20 time points. These longer time series are particularly interesting for investigating development over time. More recently, a new variety of time course studies have come from single-cell sequencing experiments (Habib et al., 2016; Shalek et al., 2014; Trapnell et al., 2014) which can sequence single cells at different stages of development; in this case, the time point is the stage of the cell in the process of development -- a value that is not know but estimated from the data as its \"pseudo-time.\"\n\nWhile there are many methods that have been proposed for discrete aspects of time course data, the entire workflow for analysis of such data remains difficult, particularly for long, developmental time series. Most methods proposed for time course data are concerned with detecting genes that are changing over time (differential expression analysis), examples being edge (Storey et al., 2005), functional component analysis based models (Wu & Wu, 2013), time-course permutation tests (Park et al., 2003a), and multiple testing strategies to combine single time point differential expression analysis (Wenguang & Zhi, 2011). However, with long time course datasets, particularly in developmental systems, a massive number of genes will show some change (Varoquaux et al., 2019). For example, in a study of mice lung tissues infected with influenza that we consider in this workflow, we show that over 50% of genes are changing over time. The task in these settings is often not to detect changes in genes, but to categorize them into biologically interpretable patterns.\n\nWe present here a workflow for such an analysis that consists of 4 main parts (Figure 1):\n\nQuality control and normalization;\n\nIdentification of genes that are differentially expressed;\n\nClustering of genes into distinct temporal patterns;\n\nBiological interpretation of the clusters.\n\nThis workflow represents an integration of both novel implementations of previously established methods and new methodologies for the settings of developmental time series. It relies on several standard packages for analysing gene expression data, some specific for time-course data, others broadly used by the community. We provide the various steps of the workflow as functions in a R package called moanin.\n\n\nInstallation and setup\n\nmoanin and timecoursedata are available from bioconductor, and can be installed using the install function in the package BiocManager, along with the corresponding package that contains time course datasets we will use:\n\n\n\nThe following additional packages are needed for this workflow:\n\n\n\nIf Bioconductor is installed, the CRAN and Bioconductor packages above can be installed via\n\n\n\n\nAnalysis of the dynamical response of mouse lung tissue to influenza\n\nThis workflow is illustrated using data from a micro-array time-course experiment, exposing mice to three different strains of influenza, and collecting lung tissue during 14 time-points after infection (0, 3, 6, 9, 12, 18, 24, 30, 36, 48, 60 hours, then 3, 5, and 7 days later) (Shoemaker et al., 2015). The three strains of influenza used in the study are (1) a low pathogenicity seasonal H1N1 influenza virus (A/Kawasaki/UTK4/2009 [H1N1]), a mildly pathogenic virus from the 2009 pandemic season (A/California/04/2009 [H1N1]), and a highly pathogenic H5N1 avian influenza virus (A/Vietnam/1203/2004 [H5N1]. Mice were injected with 105 PFU of each virus. An additional 42 mice were injected with a lower dose of the Vietnam avian influenza virus (103 PFU).\n\nBy combining gene expression time-course data with virus growth data, the authors show that the inflammatory response of lung tissue is gated until a threshold of the virus concentration is exceeded in the lung. Once this threshold is exceeded, a strong inflammatory and cytokine production occurs. These results provide evidence that the pathology response is non-linearly regulated by virus concentration.\n\nWhile we showcase this pipeline on micro-array data, Varoquaux et al. (2019) leverages a similar set of steps to analyse RNA-seq data of the lifetime transcriptomic response of the crop S. bicolor to drought.\n\n\nOverview of the data\n\nFirst let's load the data. The package moanin contains the normalized data and metadata of (Shoemaker et al., 2015).\n\n\n\nThe meta data contains information about the treatment group, the replicate, and the timepoint for each observation (see Table 1).\n\nBefore we dive into the exploratory analysis and quality control, let us define color schemes for our data that we will use across the whole analysis. We define color schemes for groups and time points as named vectors. We also define a series of markers (or plotting symbols) to distinguish replicate samples in scatter plots. We also reorder the factor Group which describes the treatments so that the treatments are ordered from low to high pathogeny (with Control being first).\n\n\n\n\nQuality control and normalization\n\nThe first steps of analysis of gene expression data is always to do normalization and quality control checks of the data. However, in what follows, we do not show the steps for normalization, as these are specific to the platform (microarray); the code for the normalization is available from GitHub (Abrams et al., 2019; Park et al., 2003b).\n\nInstead, we focus on common steps for exploratory analysis of the data, including for the purpose of quality control. These steps are not specific to time course data, but could be applied for any gene expression analysis. For this reason, we will not print the detailed code that is needed for this part of the analysis; interested readers can examine the rmarkdown code that accompanies this workflow.\n\nTypically, two quality control and exploratory analysis steps are performed before and after normalization: (1) low dimensionality embedding of the samples; (2) correlation plots between each sample. In both cases, we expect a strong biological signal, while replicate samples should be strongly clustered or correlated with one another.\n\nBefore performing any additional exploratory analysis, let us only keep highly variable genes: for this step, we keep only the top 50% most variable genes. In Figure 2, we plot the distribution of variance across all genes.\n\n\n\nLet us first perform the Principal Component Analysis (PCA) analysis. Here, we perform a PCA of rank 3 of the centered and scaled gene expression data.\n\nIn Figure 3, we visualize the data by plotting the Principal Components (PC), with samples colored by either its condition (top row) or its sampling time (bottom row) and each replicate a different symbol. We can see a large difference in later time points.\n\nSamples are colored by condition (top row) and sampling time (bottom row).\n\nWe also plot in Figure 4 the Pearson correlation between each sample as a heatmap diagram (using the function aheatmap in NMF). We order the samples by their group (treatment) and timepoint (the time of sampling).\n\nWe can already see interesting patterns emerging from the correlation plot. First, the cross-correlation amongst samples taken from the control mice is higher than the cross correlation amongst the rest of the treatments. Second, the influenza-infected mice mildly react until time point 36. Third, the less pathogenic the strain is, the closer the samples are to the control condition. Fourth, the Vietnam samples at time point 120 and 168 are the one that are the most different from control samples.\n\n\nDifferential expression (DE) analysis of time-course data\n\nThe next step in a gene expression analysis is typically to run a differential expression analysis, generally to find genes different between different conditions. For time-course data, there are two different approaches for determining differentially expressed genes,\n\n1) Per-time point analysis, where we consider each time point a different condition and determine what genes are changing between specific time points, or between conditions at a single time-point.\n\n2) Global analysis, where we consider the expression pattern globally over time, and consider what genes have either different patterns between conditions or a changing pattern (i.e. non-constant) over time. A common approach first step is to fit a spline model to each gene (Storey et al., 2005), and then use that spline model to test for different kinds of differential expression across time.\n\nThe per-time point analysis uses classical differential expression approaches and is often the approach advocated when dealing with small time-course datasets, where there are only a few time points (Love et al., 2014; Ritchie et al., 2015; Robinson et al., 2010). For long time-course datasets, however, a separate test for each time point results in creating many different tests, for example one for every time point, the results of which are difficult to integrate. We find in practice that the global analysis simplifies analysis and interpretation of longer time courses data, with per-time point analysis reserved for particularly interesting comparisons of individual time-points.\n\nTime course data can either be on a single condition (to identify genes changing over time) or on multiple conditions (such as the influenza dataset we are considering), which will alter slightly the types of questions we are interested in.\n\nDE analysis in moanin. Our package moanin provides functionality for performing both of these types of approaches, though our focus is on the global approach, specifically by fitting spline models to the genes.\n\nIn both situations, we first need to set up an object (a moanin object) to hold the meta data, as well as information for fitting the spline model (formula, number of degrees of freedom of the splines, ...). The moanin object will contain information used throughout this analysis, in particular the condition and timepoints of each sample and the basis matrix for fitting splines models.\n\nWe start by creating the moanin object using the create_moanin_model function. We need to provide two things to the function: a data.frame with the metadata, and the number of degrees of freedom of the splines to be used in the functional modeling. The metadata data.frame object should contain at least two columns: one named Group, containing the treatment effect, and a second one named Timepoint containing the timepoint information (see Table 1).\n\nWe create a moanin object for our data:\n\n\n\nThe main operation of the create_moanin_model function, in addition to holding the necessary meta data of the samples, is to create a basis matrix for the splines fit. This matrix gives the evaluation of all of the spline basis functions for each sample (as such, replicate samples will have the same values). By default, create_moanin_model will create the spline basis functions which will lead to a different spline fit for every group (as defined by the Group variable in the meta data.frame). This is done by the following R formula syntax:\n\n\n\nAlternatively, the user can provide a formula of their own, or simply provide the basis matrix themselves.\n\nWe can see this information when we print the object:\n\n\n\n\n\nThe moanin class extends the SummarizedExperiment class of Bioconductor, so that the data as well as the meta data and our spline information are all held in one object. This means that the object can be subsetted just like the data matrix, and the corresponding meta data and basis function evaluations will be similarly subsetted:\n\n\n\n\n\n\n\n\n\nmoanin provides a simple interface to perform a timepoint by timepoint differential expression analysis. Comparison between groups is traditionally done by defining the group comparisons (called contrasts in linear models) as a linear combination of the coefficients of the model. Comparing groups within each timepoint can create many contrasts, and thus moanin provides functionality to create these contrasts in an automatic way, and then calls limma (Ritchie et al., 2015) on the set of contrasts provided. By default, moanin expects RNA-Seq gene counts, and will estimate voom weights, so for microarray data we will set use_voom_weights=FALSE.\n\nHere, we show an example where we create contrasts that will be the difference between the control mouse (\"M\") and the mouse infected with the high dose of the influenza strain A/Vietnam/1203/04 (H5N1) (\"VL\") for each time point (but the function works with any form contrasts (Ritchie et al., 2015)).\n\nFirst, create the contrasts for all timepoints between the two groups of interest:\n\n\n\nThis creates a character vector of contrasts to be tested, one for each timepoint, in the format required by limma:\n\n\n\nThen moanin will run the differential expression analysis on all of those timepoints jointly using the function DE_timepoints.\n\n\n\nThe output is a table of results, where each row corresponds to a gene and the columns correspond to the p-value (pval), log-fold change (lfc) and adjusted p-value (qval) of the sets of contrasts; the order of the genes in the table is the same as the input data. Table 2 contains the results for the first timepoint (i.e. first three columns of the output) and the first ten genes.\n\nShown are the three columns giving results corresponding to timepoint 0 (the first time point).\n\nAdditional timepoints are in the additional columns of the output.\n\nWe will repeat this, comparing each of the remaining three treatments to the control (\"M\") (code not printed here, as it is a replicate of the above, see accompanying rmarkdown document).\n\nIn Figure 5, we show the distribution of genes found differentially expressed per week between control and each of the influenza strains. Such an analysis can demonstrate some general trends, with clearly more genes being differentially expressed at later time points, and the Vietnam high-dose showing perhaps an earlier onset than the low-dose.\n\nHowever, the distribution of the number of genes found differentially expressed by considering each time-point independently highlights the challenges of such approach. We can see that some timepoints have many less genes found significantly differentially expressed (e.g. timepoint 6H and 18H of the Kawasaki strain). While there may be biological differences at those time points for some genes, it seems unlikely that the large majority of genes differentially expressed at timepoint 3H stop being differentially expressed at 6H and then jump back to being differentially expressed at 9H. A more likely explanation is that there are some technical or biological artifacts about the samples for 6H that are creating higher variation and thus less ability to detect significance.\n\nAnother difficulty with such an approach is making sense of the general temporal structure for any particular gene, as different genes will have different combinations of timepoints DE. For the comparison of the Kawasaki strain to the control, for example, there are 26534 genes found DE in some timepoints, and there are 1590 different combinations of timepoints for which they are DE. Some of these make sense, such as DE in timepoints 48H–168H (509 genes), but many are very fragmentary. For example there are 330 genes which are DE in timepoints 48, 60, 120, 168H, but not in the 72H. Many of these genes are likely to have not made the cutoff for significance in 72H, but don't show real differences in the overall trend between 48H-168H. In Figure 6, we show the plots of the first 10 such genes. We can see that most genes show overall changes across the timecourse, but have either overall increased variability at 72H often due to one single replicate behaving as an outlier: This increased variability results at 72H in a lack of significance for this timepoint despite showing an overall different expression pattern.\n\nAs a summary, classic differential expression methods are appropriate for unordered treatments, but fail to make use of the temporal structure of the data.\n\nTo leverage this temporal structure, Storey et al. (2005) proposed to model each gene in a time-course micro-array with a splines function, and to use a log-ratio likelihood test to detect differentially expressed genes.\n\nmoanin extends this idea by not only fitting a splines function for each gene, but also providing functionality to compare time course data between different treatment conditions, using a similar mechanism of contrasts -- only now the contrasts are differences between the estimated mean functions. This is done with the function DE_timecourse, which takes a similar input that of DE_timepoints, only now it will fit spline functions for each gene and test the entire mean function (and unlike DE_timepoints, therefore does not require the extra step of expanding the contrasts into contrasts for individual timepoints).\n\n\n\nThe output from DE_timecourse is a matrix of (raw) p-values and Benjamini-Hochberg corrected q-values for each comparison.\n\n\n\nFor convenience we will separate these into two matrices.\n\n\n\nThe number of genes found differentially expressed ranges from around 12000 to 29000 depending on the strain and dosage of influenza virus given to the mice (Figure 7). This corresponds to between 30% to 70% of the genes found differentially expressed in this time-course experiment.\n\nThe next step in a classical differential expression analysis is typically to assess the effect of the treatment by calculating for each gene the log fold change in the gene expression between the treatment and control.\n\nComputing the log fold change on a time-course experiment is not trivial: one can be interested in the average log-fold change across time, or the cumulative log-fold change. Sometimes a gene can be over-expressed at the beginning of the time-course data, and then under-expressed at the end of the experiment. As a result, moanin provides a number of possible ways to compute the log fold change across the whole time-course. This is done via the function estimate_log_fold_change which takes as arguments the data, the moanin object, the contrasts to evaluate, and the method to use to estimate the log-fold change.\n\nIndividual timepoints. The first method (\"timely\") gives a simple interface to compute the log-fold change for each individual timepoints (see Table 3 for a sample output).\n\n\n\nThis matrix can then be used to visualize the log-fold change for each contrast per timepoint.\n\nCumulative Effect. Sometimes, a single value per gene for each contrast is more useful, and estimate_log_fold_change used above provides several options for this as well. See Table 4 for all the possible ways to compute log-fold change values with estimate_log_fold_change (including timely discussed above).\n\nThe method \"timecourse\" tries to capture the overall strength and direction of the response in the following way: we leverage the timepoint by timepoint log-fold change lfc(t), and apply the following formula:\n\nNote, however, that when a gene is not consistently up- or down-regulated the estimation of the direction will not accurately represent the changes observed.\n\nWe demonstrate several of these methods on our data.\n\n\n\nThe returning object is a matrix, where each row corresponds to a gene, each column to a contrast, and each entry to the log-fold change for this pair of contrast and gene (see Table 5).\n\nIn Figure 8, we plot the log-fold change summary of each of these methods against each other. We see that each method captures different elements of the time-course data, for example, overall change versus the largest change.\n\nWith a single measure of log-fold change and the p-value, we can now look at the traditional volcano plot. In Figure 9, we show the example of a volcano plot for the comparison of the control to the Kawasaki strain, using the \"timecourse\" method of calculating log fold change.\n\n\n\nThe package moanin also provides a simple utility function (plot_splines_data) to visualize gene time-course data from different conditions. In Figure 10, we plot the 10 genes with the smallest p-values.\n\n\n\nFor each gene, the individual data points are plotted against time and color coded by their condition. Further, a fitted spline function for each group is plotted to aid in comparing global trends across conditions.\n\nFigure 11, we visualize the genes with the largest absolute timecourse log-fold change.\n\n\n\nIn examining these visualizations, we can see that genes often follow similar patterns of expression, although on a different scale for each gene. We can leverage this observation to cluster the genes into groups of similar patterns of transcriptomic response.\n\n\nClustering of time-course data\n\nThe very large number of genes found differentially expressed impairs any interpretation one would attempt: with 70% of the genome found differentially expressed, all pathways are affected by the treatment. Hence the next step of the workflow to cluster gene expression according to their dynamical response to the treatment.\n\nBefore clustering the genes, we first reduce the set of genes of interest to genes that (1) are found to be significantly differentially expressed; (2) have a large-fold change between conditions. Reducing the set of genes on which to perform the clustering allows the estimation of the centroids of the clusters with more stability.\n\nTo do this, we first aggregate all p-values obtained during the time-course differential expression step in a single p-value using Fisher's method (Fisher, 1925) (pvalues_fisher_method). Next we select all the genes which have a Fisher-adjusted p-value below 0.05 and a log-fold change of at least two for at least one condition and one time-point.\n\n\n\nAfter filtering, we are left with 5521 genes. We can then apply a clustering routine. As observed by looking at genes found differentially expressed, many genes share a similar gene expression pattern, but on different scales.\n\nWe thus propose the following adaptation of k-means:\n\n1. Splines estimation: for each gene, fit the splines function with the basis of your choice (as contained in the moanin object).\n\n2. Rescaling splines: for each gene, rescale the estimated splines function such that the values are bounded between 0 and 1 and thus comparable between genes.\n\n3. K-means: apply k-means on the rescaled fitted values of the splines to estimate the cluster centroids.\n\nThe rescaling of the splines aims at bringing each gene onto a comparable scale, akin to the centering and scaling performed on gene expression typically done on gene expression studies without a time component.\n\nThese clustering steps are performed by the splines_kmeans function in moanin. For now, we will set the number of clusters to be 8, though we will return to the question of picking the best number of clusters below.\n\n\n\nThe splines_kmeans function returns a named list with:\n\ncentroids: a matrix containing the cluster centroids. The matrix is of shape (n_centroid, n_samples).\n\nclusters: a vector of size n_genes, containing the cluster assignments given by the kmeans step of each gene\n\nWe then use the plot_splines_data function, only now applied to the centroids, to visualize the centroids of each cluster obtained with the splines k-means model (Figure 12).\n\n\n\nThese centroids are on a 0-1 scale, because of our rescaling of the spline fits, and do not represent the actual gene expression level. In Figure 13, we plot a few of the actual genes assigned to cluster 2 with the estimated centroid overlaid:\n\n\n\nAs we can see, while these genes have some similarity with the pattern of the cluster centroids, these particular genes are not the best examples of the cluster, in the sense of matching the centroid estimates. Because of this variability in how well the genes fit a cluster, we would like to be able to score how well a gene fits a cluster.\n\nFurthermore, we arbitrarily chose a subset of genes based on our filter, and we would like to have a mechanism to assign all genes to a cluster.\n\nThus, the next step in the clustering portion of the workflow is a scoring and label step. Each gene is given a score that corresponds to a goodness-of-fit between each gene and each cluster, computed as follows:\n\nwhere μk is the centroid of cluster k and yi the gene of interest. The scoring function thus returns a value between 0 and 1, 0 being the best score possible and 1 the worst score possible (no correlation between the gene and the cluster centroid).\n\nThe score then allows us to assign all genes to a cluster (i.e. a label) based on the cluster for which they have the best score, regardless of whether the gene was used in the clustering procedure.\n\nThe scoring and labeling is done via the splines_kmeans_score_and_label function. This function calculates the goodness of fit of the gene to the cluster centroid and gives a cluster label to the gene if they have a sufficiently high score, as we explain above.\n\n\n\nThe scores_and_labels list contains three elements:\n\nscores: the matrix of shape n_cluster x n_genes, containing for each gene and each cluster, the goodness of fit as described above.\n\nlabels: the labels for all of the genes with a sufficiently good goodness-of-fit score.\n\nscore_cutoff: the cutoff used on scores to determine whether to assign a label\n\nAssigning cluster labels: We could just assign each gene to a cluster based on which cluster gave the minimum score. By default, splines_kmeans_score_and_label does not do that, but rather requires a sufficiently low enough goodness-of-fit score. The criteria for being \"sufficiently low\" is based on looking at the distribution of the scores of all genes on all clusters (i.e. the entire scores matrix returned by splines_kmeans_score_and_label). A gene is then assigned to a cluster only if their best score is above the 50% percentile of that distribution, with the remaining genes getting NA as their assignment (this choice can be changed by proportion_genes_to_label). Note that because the distribution of scores of genes across all clusters is used -- this is not equivalent to assigning 50% of genes to a cluster -- it is possible that all genes are assigned to a cluster.\n\n\n\nAfter running scores_and_labels on our data, we now have 19772 genes that are assigned to a cluster.\n\nWe can visualize the impact of this filtering process by considering the distribution of goodness-of-fit scores for each cluster if we did not have the filtering cutoff of splines_kmeans_score_and_label, i.e. if we simply assign every gene to a cluster based on which gives them their minimum score. We display for each cluster, the scores of the genes assigned to that cluster (Figure 14), as compared to the cutoff value for what that score must be under our filtering procedure. We can do this by rerunning splines_kmeans_score_and_label and setting the filter percentage_genes_to_label=1 (and we can speed up the calculation by giving our previously calculated score via the argument previous_scores)\n\n\n\nThe dashed red line indicates the scoring threshold: all genes with a score above this threshold will not be labeled.\n\nNote that for some of the genes, their scores in all clusters is 1. Such genes fit poorly to all clusters and the assignment of genes to a single cluster that we did above to one cluster is done arbitrarily to the first matching cluster for such genes. This underlines the importance of filtering genes that fit poorly to clusters.\n\nWe can also investigate the differences between the labels provided by the splines k-means model and our scoring and labeling step, in terms of the number of genes assigned to each cluster (Figure 15).\n\nOf the original 5521 genes used in the clustering, 4946 are still assigned to a cluster based on their goodness of fit score. We can compare whether they are still assigned to the same clusters based on a confusion matrix shown below (Figure 16). The k-means cluster assignments are shown on the rows, and the goodness-of-fit assignments in the columns, with the number in each cell indicating the number of genes in the intersection of the two clusters.\n\nOf the four genes we plotted before, two of them remained assigned to cluster 2 after our scoring. We can again look at a few genes in cluster 2, only now pick the best scoring genes (Figure 17)\n\n\n\nWe see that, as expected, these genes are a much better match to the cluster centroids.\n\nNow, let us look more in detail at some specific clusters. Cluster 6 seems particularly interesting: it captures genes with strong differences between the different influenza treatments and the control, while the control remains relatively flat.\n\nWe've already shown how we can plot a few example genes, however, it can be hard to make sense of individual genes given the amount of noise, as well as being hard to draw conclusions based on a few genes.\n\nHeatmaps are useful to investigate the range of expression patterns for specific genes. Here, we are going to plot heatmaps of the normalized gene expression patterns and the rescaled gene expression patterns side by side.\n\nFirst, we select the genes of interest, namely those in cluster 6, based on our goodness-of-fit assignment.\n\n\n\nNow we will create the heatmaps of these genes (3746 genes, Figure 18).\n\n\n\n\n\nThose two heatmaps demonstrate that the clustering method successfully clusters genes that are on different scales, and yet share the same dynamical response to the treatments.\n\nA common question that arises when performing clustering is how to choose the number of clusters. A choice for the number of clusters K depends on the goal. In this particular case, the end goal is not the clustering, but to facilitate interpretation of the differential expression analysis step. As a result, the number of clusters should not exceed the number of gene sets the user wants to interpret. This allows to set a maximum number of clusters. Let us assume here that this number is 20 clusters.\n\nOnce the maximum number of clusters is set, several strategies allow the identification of the number of clusters:\n\n• Elbow method. First introduced in 1953 by Thorndike (Thorndike, 1953), the elbow method looks at the total within cluster sum of squares as a function of the number of clusters (WCSS). When adding clusters doesn't decrease the WCSS by a sufficient amount, one can consider stopping. This method thus provides visual aid to the user to choose the number of clusters, but often the \"elbow\" is hard to see on real data, where the number of clusters is not clearly defined.\n\n• Silhouette method. Similarly to the Elbow method, the Silhouette method refers to a method of validation of consistency within clusters, and provides visual aid to choose the number of clusters.\n\n• Stability methods Stability methods are more computationally intensive than any other method, as they rely on assessing the stability of the clustering for every k to a small randomization of the data. The user is then invited to choose the number of clusters based on a number of similarity measures.\n\nFirst, let us run the clustering for all values of k of interest. We will, for each clustering, conserve (1) with within cluster sum of squares; (2) the clustering assignment (or label) for each gene.\n\nBelow we run the clustering for k equal to 2 – 20. The splines_kmeans function returns the WCSS for each cluster, which we sum to get the total WCSS.\n\n\n\nElbow method. The Elbow method to choose the number of clusters relies on visualization aid to choose the number of clusters. The method relies on plotting the within cluster sum of squares (WCSS) as a function of the number of clusters. At some point, the WCSS will start decreasing more slowly, giving an angle or \"elbow\" in the graph. The number of clusters is chosen at this \"Elbow point.\"\n\nWe plot the WCSS for k = 2 – 20 here (Figure 19). We see that as expected the WCSS continues to drop, but there is no clear drop in the decrease, except for very small values of k (3-4 clusters). However, 3-4 seems a very small number of gene clusters to find, given the complexity\n\n\n\nAverage silhouette method. The silhouette value is a measure of how similar a data point is to its own cluster (cohesion) compared to other clusters (separation), shown in Figure 20.\n\n\n\n\n\n\n\nLooking at the stability of the clustering. On real data, the number of clusters is not only unknown but also ambiguous: it will depend on the desired clustering resolution of the user. Yet, in the case of biological data, stability and reproducibility of the results is necessary to ensure that the biological interpretation of the results hold when the data or the model is exposed to reasonable perturbations.\n\nMethods that rely on the stability of the clustering results to choose k thus ensure that the biological interpretation of the clusters holds after perturbations to the data. In addition, simulation where the data is generated with a well defined k show that the clustering is more stable for the correct number of the clusters.\n\nMost methods to find the number of clusters with stability measures only provide visual aids to guide the user. The first element often visualized is the consensus matrix: the consensus matrix is an n × n matrix that stores the proportion of clustering in which two items are clustered together. A perfect consensus matrix ordered such that all elements that belong to the same cluster are adjacent to one another which show blocks along the diagonal close to 1.\n\nTo perform such analysis, the first step is run the clustering several times on a resampled dataset--either using bootstrap or subsampling.\n\nUsing the bootstrapping strategy, we sample with replacement a sample of the same size as our original clustering (i.e. 5521 genes):\n\n\n\nUsing the subsampling strategy, we take a unique subset of the genes, keeping 80% of the genes:\n\n\n\nWe run the bootstrap method on all genes differentially expressed and with a log-fold-change higher than 2 (computed with the lfc_max method), and do it B = 20 times for each of k = 2 – 10. We show the code below, but because of the time the computations take, we have evaluated these values separately and provided the results in separate files (one per k) for users to explore more quickly.\n\n\n\nFor now, we bring in the results for k = 5 and k = 20.\n\n\n\nEach column corresponds to a bootstrap sample, each row to a gene, and each entry to the label found for that particular clustering. Thus, for each gene, we have an assignment to a cluster over the 25 resampling runs. See Table 6 for a sample of the results of the clustering for 10 bootstrap and k = 5.\n\nNow we will use the function consensus_matrix in the moanin package to calculate proportion of times each pair of samples was clustered together across the 25 resampling runs, and plot the heatmap of those consensus matrices for k = 5 and k = 20.\n\n\n\nWe can see (Figure 21) that the choice of K = 5 seems much more stable across resampling runs than that of K = 20.\n\nThe model explorer strategy. The model explorer algorithm (Ben-Hur et al., 2001) proposes to estimate the number of clusters exploiting the observation that if the number of clusters is correct, the clustering results are stable to bootstrap resampling, as described above. The distribution of similarities between bootstrapped results for each k can thus be compared for different values of k and guide the user in the choice of number of clusters.\n\nThe model explorer strategy works as follows. For a single choice of k, first perform n bootstrap experiments to estimate the cluster centroids, followed by a step of assigning a label to all data points. Then, choose a similarity measure between two partitions or clusters S(B1, B2). Examples are the normalized mutual information or Fowlkes-Mallows. Finally, compute the pairwise similarity measure between all bootstrapped partitions (i.e. B choose 2 pairs). Repeat this procedure for different k and plot per k the cumulative density of the obtained scores.\n\nWe have already run the bootstrap resampling for values k = 2 – 20 and saved the results. We will read in that data in, and use it to calculate the pairwise similarity scores.\n\nThe function plot_model_explorer takes in input a list of the bootstrapped results for all labels.\n\n\n\nFrom this plot (Figure 22), we can deduce that k = 5 is more stable than k = 3 and k =4, but not as stable as k = 2. The model explorer strategy, in addition to visualizing the diversity of the centroids, can thus help assess an adequate number of clusters.\n\nNow, replot the same model explorer, but only for the clustering experiments k = 6, k = 7, k = 8, k = 9, and k = 10 so that we can see more clearly the stability measures in that range (Figure 23).\n\n\n\nThis analysis doesn't necessarily pick a particular k, but can help decide between k within a desired range, for example, or to avoid k that degrade the stability.\n\nConsensus clustering as a way to find k. The consensus clustering (Monti et al., 2003) relies on a similar idea but instead of looking at the cumulative density of similarity measures of bootstrapped clustering, the authors suggest plotting the cumulative density of elements of the consensus matrix. We provide a function plot_cdf_consensus in moanin to do this (Figure 24):\n\n\n\nThe stability of the clustering based on the consensus matrix can then be measured via a single number by looking at the area under the curve (AUC). Indeed, the more stable the clustering, the closer to 0 or 1 will be the entries of the consensus matrix, and thus, the higher the area under the curve (AUC). To choose a good balance between the model complexity (the number of clusters) and the stability of the clustering, the consensus clustering strategy thus suggests looking at the AUC as the function of the number of clusters, or at the \"improvement\" in AUC as a function of the number of clusters to identify the highest increase.\n\nHere we calculate the AUC (Figure 25).\n\n\n\nWhile here we calculate the improvement in the AUC (Figure 26).\n\n\n\nThe consensus clustering method suggest that the most stable is k = 2, which separates over-expressed genes from under-expressed genes. While it is indeed a very stable clustering, it does not capture the range of gene expression patterns present in the data. This shows the limitation of such methods on real data, where the number of clusters is not clearly defined.\n\n\nDownstream analysis of clusters\n\nOnce good clusters are obtained, the next step is to leverage the clustering to ease interpretation. Classic enrichment analysis can then be performed on the gene set defined by each cluster. Examples include KEGG pathway enrichment analysis, GO term enrichment analysis, and motif enrichment analysis.\n\nFirst, let us clean up the genes we work with and only select the genes we are going to use in the enrichment analysis. One can either consider (1) the results of the clustering without any further filtering, (2) only the set of differentially expressed genes in each cluster, or (3) the subset of genes that fit well to a cluster (based on some criterion).\n\nLet us first tackle the case of pathway enrichment analysis. We will leverage the packages biomaRt (Durinck et al., 2005) and KEGGprofile (Zhao et al., 2017) for this step. KEGGprofile is a package that facilitates enrichment analysis on a set of genes labeled with the Ensembl annotation (Yates et al., 2019) based on the set of biological pathways described in the KEGG database (Kanehisa et al., 2015).\n\nWe thus need to convert the gene names, which are given in the Refseq annotation, into the corresponding Ensembl name. This is where biomaRt comes in handy: it enables easy conversion from one gene annotation to another. Here, we will use the function getBM in biomaRt to convert the gene names from cluster 8.\n\n\n\nThen we use the function find_enriched_pathway in the KEGGprofile package to determine whether any KEGG pathways are enriched in cluster 6, i.e. whether a higher percentage of genes from a single pathway are found in cluster 6 than we would expect by simply proportional assignment of genes to clusters (see Table 7).\n\n\n\nThe Gene Ontology (GO) database (Consortium, 2018), also categorizes genes into meaningful biological ontologies and can be used for enrichment analysis via the package topGO (Alexa & Rahnenfuhrer, 2016). We again use biomaRt to find the mapping between genes and the GO terms to which they match.\n\n\n\nThe biomaRt query results in a matrix with two columns: gene names and GO term ID. The package topGO (Alexa & Rahnenfuhrer, 2016) expects the GO term to gene mapping to be a list where each item is a mapping between a gene name and a GO term ID vector, for example:\n\n\n\nmoanin provides a simple function (create_go_term_mapping) to make this conversion:\n\n\n\nOnce the gene ID to GO mapping list is created, moanin provides an interface to topGO to determine enriched GO terms. In particular, it performs a p-value correction and only returns the significant GO-term enrichment in an easy to use data.frame object. Here, we show an example of running a GO term enrichment on the \"Biological process\" ontology (BP) (see Table 8 for results).\n\n\n\n\nConclusion\n\nThis workflow provides a tutorial for the analysis of lengthy time-course gene expression data in R using the package moanin, which aids in implementing common timecourse analyses. We illustrate the workflow through the analysis of mice lung tissue exposed to different influenza strains and measured over time. The proposed workflow consists of three common analysis main steps generally performed after quality control and normalization: (1) differential expression analysis; (2) clustering of time-course gene expression data; (3) downstream analysis of clusters. We demonstrate how the use of the package moanin allows for easy implementation of these procedures in the setting of time-course data.\n\n\nData availability\n\nData used in this workflow are available from NCBI GEO, accession GSE63786. Normalized data can be found in timecoursedata. Normalization information is provided as supplementary information.\n\n\nSoftware availability\n\nSource code is available from GitHub: https://github.com/NelleV/2019timecourse-rnaseq-pipeline.\n\nArchived source code at the time of publication: https://doi.org/10.17605/OSF.IO/U2DQP (Varoquaux & Purdom, 2020)\n\nLicense: BSD-3\n\nAll packages used in the workflow are available on GitHub, CRAN, or Bioconductor.\n\nFinally, we use sessionInfo() to display all packages used in this pipeline and their version numbers.\n\n",
"appendix": "Acknowledgments\n\nThe authors thank Karthik Ram and the Ropensci community for valuable feedback.\n\n\nFootnotes\n\n1Because the collection of the mRNA is often destructive, samples at different time points are generally from different biological samples; longitudinal studies, for example tracking the same subject over time, are certainly possible, but not directly considered here.\n\n\nReferences\n\nAbrams ZB, Johnson TS, Huang K, et al.: A protocol to evaluate RNA sequencing normalization methods. BMC Bioinformatics. 2019; 20(Suppl 24): 679. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlexa A, Rahnenfuhrer J: TopGO: Enrichment Analysis for Gene Ontology. 2016. Reference Source\n\nBen-Hur A, Elisseeff A, Guyon I: A Stability Based Method for Discovering Structure in Clustered Data. Pac Symp Biocomput. 2001; 6–17. PubMed Abstract\n\nThe Gene Ontology Consortium: The Gene Ontology Resource: 20 years and still GOing strong. Nucleic Acids Res. 2018; 47(D1): D330–D338. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDurinck S, Moreau Y, Kasprzyk A, et al.: BioMart and Bioconductor: a powerful link between biological databases and microarray data analysis. Bioinformatics. 2005; 21(16): 3439–40. PubMed Abstract | Publisher Full Text\n\nFisher RA: Statistical Methods for Research Workers. Edinburgh Oliver & Boyd. 1925. Reference Source\n\nHabib N, Li Y, Heidenreich M, et al.: Div-Seq: Single-nucleus RNA-Seq reveals dynamics of rare adult newborn neurons. Science. 2016; 353(6302): 925–28. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKanehisa M, Sato Y, Kawashima M, et al.: KEGG as a reference resource for gene and protein annotation. Nucleic Acids Res. 2015; 44(D1): D457–D462. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLove MI, Huber W, Anders S: Moderated Estimation of Fold Change and Dispersion for Rna-Seq Data with Deseq2. Genome Biol. 2014; 15(12): 550. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMonti S, Tamayo P, Mesirov J, et al.: Consensus Clustering: A Resampling-Based Method for Class Discovery and Visualization of Gene Expression Microarray Data. Mach Learn. 2003; 52(1): 91–118. Publisher Full Text\n\nPark T, Yoo DH, Ahn JI, et al.: Statistical tests for identifying differentially expressed genes in time-course microarray experiments. Bioinformatics. 2003a; 19(6): 694–703. PubMed Abstract | Publisher Full Text\n\nPark T, Yi SG, Kang SH, et al.: Evaluation of normalization methods for microarray data. BMC Bioinformatics. 2003b; 4: 33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRitchie ME, Phipson B, Wu D, et al.: limma Powers Differential Expression Analyses for RNA-Sequencing and Microarray Studies. Nucleic Acids Res. 2015; 43(7): e47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobinson MD, McCarthy DJ, Smyth GK: edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010; 26(1): 139–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShalek AK, Satija R, Shuga J, et al.: Single-cell RNA-seq reveals dynamic paracrine control of cellular variation. Nature. 2014; 510(7505): 363–69. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShoemaker JE, Fukuyama S, Eisfeld AJ: An Ultrasensitive Mechanism Regulates Influenza Virus-Induced Inflammation. PLoS Pathog. 2015; 11(6): e1004856. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStorey JD, Xiao W, Leek JT, et al.: Significance analysis of time course microarray experiments. Proc Natl Acad Sci U S A. 2005; 102(36): 12837–42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nThorndike RL: Who belongs in the family? Psychometrika. 1953; 18(4): 267–76. Publisher Full Text\n\nTrapnell C, Cacchiarelli D, Grimsby J, et al.: The dynamics and regulators of cell fate decisions are revealed by pseudotemporal ordering of single cells. Nat Biotechnol. 2014; 32(4): 381–86. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVaroquaux N, Cole B, Gao C, et al.: Transcriptomic Analysis of Field-Droughted Sorghum from Seedling to Maturity Reveals Biotic and Metabolic Responses. Proc Natl Acad Sci U S A. 2019; 116(52): 27124–27132. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVaroquaux N, Purdom E: A pipeline to analyse transcriptomic time-course data. 2020. http://www.doi.org/10.17605/OSF.IO/U2DQP\n\nWenguang S, Zhi W: Multiple Testing for Pattern Identification, with Applications to Microarray Time-Course Experiments. J Am Stat Assoc. Taylor & Francis. 2011; 106(493): 73–88. Publisher Full Text\n\nWu S, Wu H: More Powerful Significant Testing for Time Course Gene Expression Data Using Functional Principal Component Analysis Approaches. BMC Bioinformatics. 2013; 14: 6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYates AD, Achuthan P, Akanni W, et al.: Ensembl 2020. Nucleic Acids Res. 2019; 48(D1): D682–D688. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhao S, Guo Y, Shyr Y: KEGGprofile: An Annotation and Visualization Package for Multi-Types and Multi-Groups Expression Data in Kegg Pathway. 2017. Reference Source"
}
|
[
{
"id": "77190",
"date": "02 Feb 2021",
"name": "Michael I. Love",
"expertise": [
"Reviewer Expertise Developer of statistical methods for RNA-seq data analysis"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors provide an R package and accompanying workflow to analyze time course gene expression data, with a particular focus toward longer time courses, such that pairwise DE analysis at each time point is cumbersome. The package and workflow importantly emphasizes quality control at a number of steps throughout the analysis, to ensure that the results are reliable (e.g. not due to artifact) and biologically meaningful. I was able to download the workflow and run the entire workflow using the Make script provided in the GitHub repo. I obtained the same results and figures using R 4.0.3 and Bioconductor 3.12 (the workflow uses a previous version of R and Bioconductor). The workflow is easy to read, and the package provides a number of useful features that are not available in other packages to my knowledge. I found the functions and section describing the assignment of genes to clusters to be particularly useful.\nMajor comments:\nHaving communicated with a number of bioinformatic analysts working with time course data, I would say the most common approach is a LRT of ~condition + time + condition:time compared to ~condition + time to obtain a single p-value for the inclusion of the interaction terms. The current article implies that the most common approach is to perform pairwise DE at every time point, but in my experience this is not the case. I agree that the methods and workflow presented here are useful and go well beyond the simple LRT with a single p-value per gene. I think what \"saves\" most users when performing the LRT is that most time course designs have a limited number of replicates (e.g. n=2 or 3), and so they are under-powered for detecting all of the condition-specific dynamics over time. Otherwise, as mentioned in the article, as the power increases, it is entirely possible to obtain a large majority of genes rejecting the null hypothesis.\n\nIt wasn't clear to me when or if the differential analyses were controlling for a t0 baseline or not. E.g. in the M vs VL comparisons, do the comparisons after time point 0 control for the difference at time point 0? Is this an option? Likewise, for the LFC matrix provided by estimate_log_fold_change().\n\nA number of time course packages for RNA-seq are not mentioned, though I thought for completeness, it would warrant a brief description of how existing methods/workflows differ from the methods presented here, as part of the Introduction. For example, Spies et al 2017 compares a number of methods1.\nMinor comments:\nIt would be good for stylistic purposes to maintain consistency between \"x = 3\", \"x=3\", and \"x <- 3\" throughout the workflow.\n\n\"bioconductor\" should be capitalized\n\n\"micro-array\" is used, elsewhere \"microarray\"\n\n\"meta data\" and \"metadata\"\n\n\"The first steps ... is always to do ...\"\n\nThe Rmarkdown code is mentioned while the link does not appear until the end of the article.\n\n\"samples colored by ... its condition\"\n\n\"model formula\" and \"basis matrix\" are mentioned without much accompanying description what these terms mean. Many readers may not know what these refer to.\n\nHow would one provide the basis matrix to the object creation step?\n\nMay worth noting that if the sum of the estimated LFC is close to 0 for a gene, slight fluctuations in the data or scaling/normalization could flip the sign of the timecourse value. It is mentioned that the direction of this estimator may not be accurate, but its variance could also be mentioned.\n\nIt is mentioned that 70% of the _genome_ is DE, presumably this is 70% of the genes passing expression filters.\n\nMay be worth mentioning that scaling the splines and clustering works in this case because the genes have been filtered to those with small p-value and large LFC. Otherwise, rescaling by the range can increase the noise in expression data.\n\nIn the goodness of fit scoring equation, it wasn't clear the scale of y_ij.\n\n\"--\" instead of en dash in \"Assigning cluster labels\".\n\nMissing period after \"...previous scores)\"\n\nWhen it is mentioned that \"Cluster 6 seems particularly interesting\" may be worth referring back to Figure 12.\n\nWhat is the scale of the data in Figure 18 (left panel)? Log CPM?\n\nIn the section on choosing a number of clusters, various terms are used to describe what I imagine is referring to a gene: \"data point\", \"item\", \"samples\". E.g. use of \"samples\" in Figure 21 caption. Perhaps only use \"gene\" throughout.\n\nRecommend citation for RefSeq and Ensembl.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "120341",
"date": "09 Feb 2022",
"name": "Rodolphe Thiébaut",
"expertise": [
"Reviewer Expertise Transcriptomics",
"statistics",
"immunology"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSummary:\n\nThe authors propose a workflow for analysing time-course gene expression data. The whole workflow is implemented in an R/Bioconductor package. They present a step-by-step pipeline of analysis from quality control to enrichment analysis which is illustrated with a real data set.\nGeneral comments:\n\nThe manuscript is well written and easy to read. It presents an interesting workflow to analyze time-course gene expression data. This type of data raises specific difficulties that do need specific approaches. Practical solutions are proposed and implemented such as various measures of the log-fold change for time-course data or a score for the contribution of genes to data-driven gene clusters.\nAll functions of the package, as well as their outputs, are well illustrated and interpreted using a real data set. The integration of existing functions and new functions for all steps of time-course gene expression data analysis into a single package represents a useful contribution to the field of bioinformatics.\nThe code is working well with the exception of the package KEGGprofile which is not available with R 4.1.2.\nMain comments:\nAll the manuscript is illustrated using a study on the dynamical response of mouse lung tissue to influenza but this data set is not clearly described. In particular, there is no information concerning the sample size and the number of genes measured. Furthermore, the meaning of the labels of the 5 mice groups used on the figures (e.g. Figures 3 and 4) is not explained in the text of the manuscript.\n\nThe gene expression data were generated with a microarray technology that is not used anymore. An illustration using RNA-seq data would have been more relevant. It does not compromise the interest of the package but may demand some adjustment for this type of data. The behavior of some statistical tests can clearly be problematic with RNAseq data (see for instance Gauthier et al., 20201).\n\nThe context of the work which seems to exclude longitudinal studies, i.e. repeated measurements among the same subject, as compared to repeated cross-sectional analyses where mice are sacrificed at each time point, should be better highlighted and not just mentioned in a note. It would even be of interest to better pinpoint where the pipeline is not relevant for longitudinal data (e.g. the spline regression would require random effects).\n\nFisher’s method is used to aggregate all the p-values obtained in the DE step. Fisher’s method only works with independent p-values. Is it also the case here? Other methods to merge p-values according to the strength of the dependency between p-values exist, see Vovk & Wang, 2020: Combining p-values via averaging2.\n\nIn regards to the workflow for selecting the optimal number of clusters (k), quantitative criteria to select the optimal value for k could have been provided. See for instance Charrad et al., (2014): NbClust: an R package for determining the relevant number of clusters in a data set3.\n\nShort Time Series distance accounts for both the dependency across time points as well as interval length between two measurements. It could have been suggested for the clustering. See seminal work on STS and FSTS with an example on microarray in Möller-Levet et al., 2003: Fuzzy Clustering of Short Time-Series and Unevenly Distributed Sampling Points4.\n\nIn the last step consisting of “Downstream analysis of clusters”, it could be of interest to refer to other catalogs than GO & KEGG, such as Altman et al., 20215 or BTM from Li et al., 20146. Even though a concurrent approach would have been to start from a direct gene set analysis as in Hejblum et al., 20157.\nMinor comments:\nPage 1: In the first chunk (at the top of the page), there is an extra comma.\n\nPage 10: “and adjusted p-value”. Which method?\n\nFigure 7 (page 13): We can read that: “The number of genes found differentially expressed ranges from around 12000 to 29000 […] (Figure 7)”, while on the barplot (Figure 7) there are cases where the bars exceed 35000 (e.g Vietname HD).\n\nPage 21: In the equation, S0k, aiGj, biGj, tj are not defined.\n\nPage 23: How do the authors conclude that Cluster 6 is particularly interesting?\n\nPage 26 (Figure 18): The color legend for the left heatmap is missing: it could be concluded that raw expression only varies from 0 to 0.6.\n\nPage 26: The authors should cite Rousseeuw (1987)7, Silhouettes: a graphical aid to the interpretation and validation of cluster analysis, for the Silhouettes method.\n\nPage 27: The authors do not explain how to choose the optimal number of clusters according to the Silhouette method and they do not discuss the optimal number of clusters they would choose for their application with this criterion.\n\nPage 33: “from cluster 8” while in the chunk above it is “cluster = 6” and in the following it is cluster 6 that is studied.\n\nIs the rationale for developing the new method (or application) clearly explained? Partly\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1447
|
https://f1000research.com/articles/9-110/v1
|
12 Feb 20
|
{
"type": "Study Protocol",
"title": "Methods used to meta-analyse results from interrupted time series studies: A methodological systematic review protocol",
"authors": [
"Elizabeth Korevaar",
"Amalia Karahalios",
"Andrew B. Forbes",
"Simon L. Turner",
"Steve McDonald",
"Monica Taljaard",
"Jeremy M. Grimshaw",
"Allen C. Cheng",
"Lisa Bero",
"Joanne E. McKenzie",
"Elizabeth Korevaar",
"Andrew B. Forbes",
"Simon L. Turner",
"Steve McDonald",
"Monica Taljaard",
"Jeremy M. Grimshaw",
"Allen C. Cheng",
"Lisa Bero",
"Joanne E. McKenzie"
],
"abstract": "Background: Systematic reviews are used to inform healthcare decision making. In reviews that aim to examine the effects of organisational, policy change or public health interventions, or exposures, evidence from interrupted time series (ITS) studies may be included. A core component of many systematic reviews is meta-analysis, which is the statistical synthesis of results across studies. There is currently a lack of guidance informing the choice of meta-analysis methods for combining results from ITS studies, and there have been no studies examining the meta-analysis methods used in practice. This study therefore aims to describe current meta-analysis methods used in a cohort of reviews of ITS studies. Methods: We will identify 100 reviews that include meta-analyses of ITS studies from a search of eight electronic databases covering several disciplines (public health, psychology, education, economics). Study selection will be undertaken independently by two authors. Data extraction will be undertaken by one author, and for a random sample of the reviews, two authors. From eligible reviews we will extract details at the review level including discipline and type of interruption; at the meta-analytic level we will extract type of outcome, effect measure(s), meta-analytic methods, and any methods used to re-analyse the individual ITS studies. Descriptive statistics will be used to summarise the data. Conclusions: This review will describe the methods used to meta-analyse results from ITS studies. Results from this review will inform future methods research examining how different meta-analysis methods perform, and ultimately, the development of guidance.",
"keywords": [
"Interrupted time series",
"meta-analysis",
"systematic review"
],
"content": "Introduction\n\nSystematic reviews aim to collate and synthesise all available evidence on a particular topic. They are used to inform healthcare decision making, either directly, or through their inclusion in knowledge tools such as clinical practice guidelines. Many reviews examining the effects of clinical interventions are appropriately limited in scope to inclusion of randomised trials. However, in reviews where the aim is to examine the effects of organisational, policy change or public health interventions or exposures (e.g. chemical exposures), evidence from non-randomised studies may offer the only evidence, or provide important additional evidence to that gained from randomised trials1.\n\nInterrupted time series (ITS) studies are a type of non-randomised design in which measurements on a group of individuals (e.g. a community) are taken repeatedly both before and after an ‘interruption’2. The interruption may be intended (e.g. a government-implemented policy), although will not necessarily be initiated or designed by the ITS investigators (e.g. by researchers within a university)3, or may be unintended (e.g. an exposure such as a natural disaster). The key benefit of the ITS design is that the period before the interruption can be used to estimate the underlying time trend. If modelled correctly, this before-trend can be projected into the post-interruption period to provide a counterfactual for what would have occurred in the absence of the interruption4. ITS studies with controls (e.g. an internal or external control series, control outcome) may provide more certainty in causally attributing any observed effects to the interruption5. Several effect estimates can be obtained from an ITS study to characterise both short and long-terms effects of the interruption (e.g. level change and slope change).\n\nMeta-analysis is the statistical synthesis of results across studies leading to combined effect estimates6. Meta-analysis (and its extensions) is a core component of many systematic reviews. The benefits of meta-analysis have long been established, including the ability to more precisely estimate effects, examine and quantify inconsistency of the effects across studies, and identify factors that may potentially modify the size of the effects7–10. Two approaches for meta-analysing results from ITS studies include the two-stage or one-stage approach11. In the two-stage approach, effect estimates from each series are first computed, and these are then combined across series using a meta-analysis method (e.g. DerSimonian and Laird12). In the one-stage approach, a single model including all series is fitted to simultaneously obtain the combined effect estimates11. The one-stage approach requires the raw time series data to be available for all series, but has the proposed advantage of being more efficient since the data across all the series are used in estimating the effects11.\n\nFor two-stage meta-analysis, a notable challenge is that many primary ITS studies are analysed incorrectly. For example, ITS studies may be analysed as though the study was a before-after design13, or analysed as an ITS design, but without taking account of the correlation between observations over time (known as autocorrelation)14–16. The former is likely to result in estimates of the effect of the interruption that are biased, while the latter is likely to result in estimates of standard errors that are too small17. Both have important implications for a two-stage meta-analysis in terms of bias, the weights that studies receive, and in turn, the precision of the combined estimate. A further challenge is that the effect measures chosen and reported by the primary study authors (e.g. level change) may not match those of interest to the systematic reviewer (e.g. slope change).\n\nIn some studies, the raw time series data may be available through extraction of data from graphs or their availability in tables6,8,11,18. In this circumstance, it may be possible to overcome some of the above challenges through re-analysis of the ITS studies by appropriately accounting for the design and autocorrelation, or re-analysing the raw data to obtain the desired effect measure for the meta-analysis when it differs from that reported in the primary study. These computed effects may then be combined using two-stage meta-analysis. Alternatively, each study’s raw data may be analysed in a single model using a one-stage meta-analysis approach6,8.\n\nTo our knowledge, there have been no reviews examining the approaches and methods used to meta-analyse effect estimates from ITS studies. In this review we therefore aim to: 1) investigate whether reviewers re-analyse primary ITS studies included in reviews, and if so, what re-analysis methods are used; 2) what meta-analysis methods are used; 3) what effect measures are used, and how completely the estimated combined effects are reported; and 4) what tools and domains are used to assess the risks of bias or methodological quality of the included ITS studies. Here, we report the planned design of our review, including the criteria that we will use to identify eligible studies, as well as the information we will extract and describe.\n\n\nMethods\n\nThis study aims to identify and describe reviews that include meta-analyses of ITS studies. The reviews will be identified by searching several electronic databases including MEDLINE (Ovid), EMBASE (Ovid), Campbell Systematic Reviews, EconLit (EBSCOhost), 3ie, PsycINFO (Ovid), ERIC (ProQuest) and the Cochrane Database of Systematic Reviews (CDSR). Study selection will be undertaken independently by two authors; data extraction will be undertaken by one author, and for a minimum of 20% of randomly selected reviews, two authors. We will extract details at the systematic review level, including: discipline (public health, psychology, education, economics), type of interruption, assessment of risk of bias and methodological quality; and at the meta-analytic level: type of outcome, effect measure(s), meta-analytic methods, and any methods used to re-analyse the individual ITS studies. These aspects will be analysed and described using summary statistics, tables and figures.\n\nStudies which meet our eligibility criteria (described below) will be included. We will not restrict inclusion of reviews based on discipline or any of the PICO elements (i.e. participants/populations, interventions/interruptions, comparators, or outcomes).\n\nInclusion criteria. Studies meeting the following criteria will be included:\n\n1. the study is a review that includes at least two ITS studies which meet the review authors’ definition of an ITS design; and\n\n2. the review includes at least one meta-analysis of ITS studies.\n\nOur definition of a ‘review’ is very broad. It includes systematic reviews, reviews of selected studies (i.e. between-study meta-analysis), and studies that combine multiple ITS across sites within the same study (i.e. within-study meta-analysis). We have opted for broad inclusion since our primary interest is in the meta-analysis methods, which apply regardless of the particular study design. We will not restrict the meta-analysis by approach, that is, we will include both one-stage and two-stage meta-analyses. We will only include meta-analyses that combine estimates of model parameters, or combinations of these (e.g. pre-intervention fitted trend, slope change, level change).\n\nExclusion criteria. Studies will be excluded if they meet one or more of the following criteria. The study is:\n\n1. written in a language other than English;\n\n2. a methodological review that describes or evaluates methods to synthesise results from ITS studies;\n\n3. a review of ITS studies reported in a conference abstract, letter, book, or dissertation;\n\n4. a protocol for a review of ITS studies; or\n\n5. a stepped-wedge randomised trial.\n\nCriterion 1 is included because we are not able to translate studies written in a language other than English due to resource constraints. Criterion 2 excludes methodological reviews that describe or evaluate methods to synthesise data from ITS studies, as our aim is to describe current statistical methods applied in practice.\n\nSeveral databases will be searched to capture the broad range of disciplines that use the ITS study design. To capture reviews in health, we will search MEDLINE (Ovid), EMBASE (Ovid), Campbell Systematic Reviews, the CDSR and 3ie. For CDSR, we will directly search the ‘Characteristics of included studies’ table included in each systematic review for ITS studies. This will allow more specific identification of eligible reviews. The search of MEDLINE (Ovid) will also capture systematic reviews from the Joanna Briggs Institute Database of Systematic Reviews and Implementation Reports. To capture reviews in economics, we will search EconLit (EBSCOhost), and for psychology and education disciplines, we will search PsycINFO (Ovid) and ERIC (ProQuest).\n\nOur search strategy has been informed by previous publications that have reviewed ITS studies14,15,19. Reviews of ITS studies will be identified using terms adapted from the search strategies of these publications and then combined with terms to identify meta-analyses and systematic reviews. As there is little consistency in the terminology used to describe ITS studies15,16, our search terms are intentionally broad to achieve greater search sensitivity. Terms will be searched both as free text in the titles, abstracts and keywords fields, and as MeSH terms (or equivalent) where applicable. The MEDLINE (Ovid) strategy is presented in Table 1, and the search strategies for the remaining databases are presented in Appendix 1 (see Extended data)20. The search is limited to the period 1 Jan 2000 to 11 Oct 2019 for all databases except CDSR which is limited to the period 1 Jan 2000 to 9 Aug 2019.\n\nCitations identified from the searches will be imported into Endnote X8 (Clarivate Analytics, Philadelphia) to remove duplicates. Titles and abstracts will be sorted by year in descending order and will be screened against the eligibility criteria, with each abstract assessed as: 1) ‘Yes/Maybe includes two or more ITS studies’ and ‘Yes/Maybe a meta-analysis of ITS studies has been undertaken’, 2) ‘Yes/Maybe includes two or more ITS studies’ and ‘No meta-analysis of ITS studies has been undertaken’, or 3) ‘No, does not include two or more ITS studies’. This process will be piloted on 20 studies by EK, SLT, AK and JEM. The remaining abstracts will be screened independently by at least two members of the review team (EK, and any of SLT, AK and JEM). The full-text articles of the titles and abstracts assessed as potentially meeting the eligibility criteria (i.e. group 1 above) will be retrieved, sorted by most recent first and screened against the eligibility criteria until all reviews (if less than 100), or the 100 most recently published reviews are identified. Conflicts in screening decisions at the abstract and full-text stages will be resolved via discussion between the screeners or through consultation with the broader team.\n\nOur sample size of 100 reviews was primarily selected for reasons of feasibility. A sample of this size will allow estimation of the percentage of reviews with a particular element (e.g. reviews that re-analyse the primary study data) to within a 10% margin of error (assuming a prevalence of 50%). This margin of error will decrease if the prevalence varies from 50%.\n\nReviews may include several meta-analyses of ITS studies for different outcomes. We plan to examine the meta-analysis methods for only one outcome per review. The following set of rules will be applied hierarchically until a unique outcome is identified (for which there could be multiple meta-analyses of different effect estimates):\n\n1) The outcome that has the largest number of effect measures (e.g. the outcome that has meta-analyses of level change and slope change estimates would be selected ahead of an outcome with only a meta-analysis of level change estimates);\n\n2) The outcome with the largest number of ITS studies; or\n\n3) The outcome that is first reported in the abstract, then the methods section, then the results section of the manuscript.\n\nA single outcome is chosen as it is likely that the meta-analysis methods are consistent across outcomes within a review. Criterion 1 has been included so that we can capture the range of effect measures used. Uncertainty in the selection of the outcome will be resolved through discussion with the review team.\n\nThe data extraction form will be designed using the Research Electronic Data Capture (REDCap) online designer21,22. The review team (EK, AK, SLT, ABF, and JEM) will pilot the data extraction form by independently extracting data from 10 reviews. This pilot testing will be used to revise the form if we uncover ambiguity or a lack of clarity in any items, identify missing items and test the logic of the form. Following piloting, data extraction will be undertaken by EK for all eligible studies and independently by at least two members of the review team (one of AK, SLT, ABF and JEM) for a further 20% of randomly selected reviews. Any inconsistencies in data extraction will be resolved via discussion between the data extractors or through consultation with the broader team. For any items where a large percentage of inconsistency is found, the percentage of studies with double data extraction will be increased.\n\nA summary of the data extraction items is presented in Table 2. In brief, we will extract details of the review’s aims, meta-analysis methods (including the reviewer’s rationale) and methods used to assess the methodological quality and/or risk of bias of the included ITS studies. For the selected outcome, we will extract the type of effect measure(s), methods of synthesis, adjustment for autocorrelation and/or seasonality.\n\nAbbreviations: ITS, interrupted time series\n\nWe will summarise the characteristics of included systematic reviews with descriptive statistics. For categorical data (e.g. the meta-analysis approach used, the risk of bias tool used) we will present frequency and percentage, and for numerical data (e.g. the number of meta-analysed ITS studies, the number of pooled estimates) we will present means (with standard deviations) or medians (with interquartile range). Statistical analyses will be undertaken using Stata version 15.023.\n\n\nDiscussion\n\nTo our knowledge, this will be the first review to examine methods for meta-analysis of ITS studies that are used in practice. Specifically, the choice of meta-analysis approach, effect measures, completeness of reporting, and tools for assessing quality or risk of bias (if undertaken). The results of this review will inform our broader research program which aims to examine how different meta-analysis methods of ITS studies perform, using statistical simulation and empirical evaluation, and provide guidance on the methods.\n\nThere are several strengths to this study. The search, screening and data extraction methods have been prespecified, and the study has been registered with PROSPERO (submitted 4 Oct 2019). Further, we will search a broad range of databases, encompassing the areas of health, economics, psychology and education. This will allow us to identify a broader range of meta-analysis methods in use, not restricted to a particular discipline.\n\nWhile the study will be limited by our ability to identify all potentially eligible reviews and meta-analyses of ITS studies, our search strategy attempts to capture the various ways these studies are described. However, given ITS studies are often not identified as such16, it is likely that we will not capture all reviews and meta-analyses that include ITS studies. Conversely, we may end up including reviews where no information regarding the definition of the included ITS studies is provided, or where an inappropriate label of ITS has been applied to included studies. While we will not exclude these reviews, we will record the reviewers’ definition of an ITS study.\n\n\nConclusions\n\nThe ITS design is often used to examine the effects of organisational, policy change or public health interventions or exposures. Meta-analysis of results from these studies provides the opportunity to estimate the interruption’s impact more precisely, and investigate factors that may modify the size of the impact. However, there is a paucity of guidance available for meta-analysing results from ITS studies. Results from this review will provide the first examination of meta-analysis methods used in practice to combine results from ITS studies. This will be used to inform future research that investigates how different methods perform, from which guidance will be developed.\n\n\nData availability\n\nNo underlying data are associated with this article.\n\nFigshare: Methods used to meta-analyse results from interrupted time series studies: A methodological systematic review protocol - Appendix 1 Search strategy. https://doi.org/10.26180/5e3b5cc4acf3020.\n\nFigshare: PRISMA-P checklist for ‘Methods used to meta-analyse results from interrupted time series studies: A methodological systematic review protocol’. https://doi.org/10.26180/5e3b5d75c00a624.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "References\n\nHiggins JP, Ramsay C, Reeves BC, et al.: Issues relating to study design and risk of bias when including non-randomized studies in systematic reviews on the effects of interventions. Res Synth Methods. 2013; 4(1): 12–25. PubMed Abstract | Publisher Full Text\n\nWagner AK, Soumerai SB, Zhang F, et al.: Segmented regression analysis of interrupted time series studies in medication use research. J Clin Pharm Ther. 2002; 27(4): 299–309. PubMed Abstract | Publisher Full Text\n\nWatson SI, Dixon-Woods M, Taylor CA, et al.: Revising ethical guidance for the evaluation of programmes and interventions not initiated by researchers. J Med Ethics. 2020; 46(1): 26–30. PubMed Abstract | Publisher Full Text\n\nShadish WR, Cook TD, Campbell DT: Experimental and quasi-experimental designs for generalized causal inference. 2002. Reference Source\n\nBernal JL, Cummins S, Gasparrini A: The use of controls in interrupted time series studies of public health interventions. Int J Epidemiol. 2018; 47(6): 2082–93. PubMed Abstract | Publisher Full Text\n\nMcKenzie JE, Beller EM, Forbes AB: Introduction to systematic reviews and meta-analysis. Respirology. 2016; 21(4): 626–37. PubMed Abstract | Publisher Full Text\n\nDeeks JJ: Systematic reviews of published evidence: miracles or minefields? Ann Oncol. 1998; 9(7): 703–9. PubMed Abstract | Publisher Full Text\n\nDeeks J, Higgins J, Altman D: Chapter 10: Analysing data and undertaking meta-analyses. In: Cochrane Handbook for Systematic Reviews of Interventions. Cochrane. 2019. Publisher Full Text\n\nCaldwell DM, Dias S, Welton NJ: Extending Treatment Networks in Health Technology Assessment: How Far Should We Go? Value Health. 2015; 18(5): 673–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nThompson SG, Higgins JP: How should meta-regression analyses be undertaken and interpreted? Stat Med. 2002; 21(11): 1559–73. PubMed Abstract | Publisher Full Text\n\nGebski V, Ellingson K, Edwards J, et al.: Modelling interrupted time series to evaluate prevention and control of infection in healthcare. Epidemiol Infect. 2012; 140(12): 2131–41. PubMed Abstract | Publisher Full Text\n\nDerSimonian R, Laird N: Meta-analysis in clinical trials. Control Clin Trials. 1986; 7(3): 177–88. PubMed Abstract | Publisher Full Text\n\nRamsay CR, Matowe L, Grilli R, et al.: Interrupted time series designs in health technology assessment: lessons from two systematic reviews of behavior change strategies. Int J Technol Assess Health Care. 2003; 19(4): 613–23. PubMed Abstract | Publisher Full Text\n\nJandoc R, Burden AM, Mamdani M, et al.: Interrupted time series analysis in drug utilization research is increasing: systematic review and recommendations. J Clin Epidemiol. 2015; 68(8): 950–6. PubMed Abstract | Publisher Full Text\n\nHudson J, Fielding S, Ramsay CR: Methodology and reporting characteristics of studies using interrupted time series design in healthcare. BMC Med Res Methodol. 2019; 19(1): 137. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPolus S, Pieper D, Burns J, et al.: Heterogeneity in application, design, and analysis characteristics was found for controlled before-after and interrupted time series studies included in Cochrane reviews. J Clin Epidemiol. 2017; 91: 56–69. PubMed Abstract | Publisher Full Text\n\nHuitema BE, McKean JW: Identifying autocorrelation generated by various error processes in interrupted time-series regression designs - A comparison of AR1 and portmanteau tests. Educ Psychol Meas. 2007; 67(3): 447–59. Publisher Full Text\n\nTurner SL, Karahalios A, Forbes AB, et al.: Design characteristics and statistical methods used in interrupted time series studies evaluating public health interventions: A review. J Clin Epi. unpublished report.\n\nTurner SL, Karahalios A, Forbes AB, et al.: Design characteristics and statistical methods used in interrupted time series studies evaluating public health interventions: protocol for a review. BMJ Open. 2019; 9(1): e024096. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKorevaar E, Karahalios E, Forbes A, et al.: Methods used to meta-analyse results from interrupted time series studies: A methodological systematic review protocol - Appendix 1 Search strategy. figshare. Journal contribution. 2020. http://www.doi.org/10.26180/5e3b5cc4acf30\n\nHarris PA, Taylor R, Minor BL, et al.: The REDCap consortium: Building an international community of software platform partners. J Biomed Inform. 2019; 95: 103208. PubMed Abstract | Publisher Full Text\n\nHarris PA, Taylor R, Thielke R, et al.: Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009; 42(2): 377–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStataCorp: Stata statistical software: release 15. Tx: College Station StataCorp LLC; 2017. Reference Source\n\nKorevaar E, Karahalios E, Forbes A, et al.: Methods used to meta-analyse results from interrupted time series studies: A methodological systematic review protocol - Completed PRISMA-P checklist. figshare. Journal contribution. 2020. http://www.doi.org/10.26180/5e3b5d75c00a6"
}
|
[
{
"id": "60103",
"date": "25 Feb 2020",
"name": "Christopher James Rose",
"expertise": [
"Reviewer Expertise Statistical modelling",
"biomedical research"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this protocol, the authors propose to systematically survey methods used to meta-analyze results from interrupted time series (ITS) studies. ITS studies are particularly useful in systematic reviews (SRs) of interventions that cannot be studied using randomized trials (e.g., due to practical, ethical, or legal reasons). Briefly, the protocol plans to identify 100 recent systematic reviews that meta-analyze results from ITS studies and then extract and summarize characteristics of the methods used.\nI agree with the authors that we currently know relatively little about the characteristics of methods used in meta-analyses of ITS results. From my point of view, I would like to have evidence that can be used to help design SRs that will meta-analyze ITS results, and to help understand potential weaknesses of such SRs. For example, I would like to have evidence to inform the number of pre- and post-interruption time points that should be required of ITS studies for inclusion in a meta-analysis, and information about whether certain ITS designs and analysis methods lead to excessive bias or imprecision. The protocol says that the resulting work will inform subsequent research that addresses these kinds of questions, using simulation and empirical evaluation. My understanding is that knowing about the landscape of methods that have been used in recent SRs will allow this subsequent work to address relevant questions (e.g., it could allow the authors to design simulation studies that usefully model what is being done in practice).\nMy review focuses mainly on the conceptual and statistical aspects of the protocol. I cannot comment on other aspects such as the literature search.\nI have two \"major\" criticisms of the protocol:\nBecause I am more interested in the evidence that the subsequent research will hopefully deliver, I would like to see more detailed thought in this protocol about how the subsequent work will be performed. This should then inform what the product of the present protocol needs to deliver to ensure the success of the subsequent work. Perhaps this has already been thought through in detail and not presented here. However, if this work has not been done, I encourage the authors to do it and update the protocol.\n\nThe analysis is not planned in enough detail that it could be implemented without having to make important choices after having seen the data. I think this potentially leads to at two problems. First, being able to choose from among several possible analyses and means of presentation risks introducing bias. Second, more detailed planning at the protocol stage may prevent problems that would otherwise only become apparent while the work is being done. I suggest that the authors substantially revise this section to specify in detail what they will do and how they will report their results, including preparing skeleton tables and/or figures. I think that thinking through these issues at the protocol stage will likely make this and subsequent work (see point 1 above) more efficient, and lead to higher quality papers.\nI have the following \"minor\" comments and suggestions:\nI suggest that the abstract is clarified to say that the authors will study the 100 most recent systematic reviews that include ITS analyses (and include the date range), rather than simply saying 100 (or whatever the final sample size is determined to be). My concern when reading the abstract was that the 100 SRs could be chosen arbitrarily, giving rise to potential bias.\n\nIt would be useful for the authors to clarify that, with respect to estimating a binomial proportion, their proposed sample size of 100 would give them a worst-case margin of error of plus or minus approximately ten percentage points (i.e., 40.2% to 59.8%), if the population parameter is 50%.\n\nThis margin of error is actually quite wide. I wonder if the authors have considered how plausible it is that the population parameter will often be close to the worst-case of 50%, and if so, whether the relatively wide confidence intervals will be informative enough for their subsequent work that will build on this paper?\n\nThe sample size of 100 seems to have been chosen under the assumption that a binomial proportion will be estimated for each factor studied (i.e., that each factor will have two levels). However, many of the criteria specified are factors with more than two levels (e.g., the protocol gives the example of three types of outcome that included reviews may study: continuous, count, and rate). Given that, I would encourage the authors to think about the more general case of estimating multinomial proportions. This would require a larger sample size for a worst-case scenario equivalent to that of a binomial distribution. A quick search identified Thompson 19871, which provides a table for estimating sample size for estimating multinomial proportions. Briefly, if the authors want to estimate multinomial proportions with 95% CIs that give a margin or error of plus or minus 10%, that paper shows that the authors should include at least 128 studies (irrespective of the number of levels of the factor studied). However, I encourage the authors to double-check this informal power analysis. The authors would then also need to use an appropriate method to estimate the proportions and their confidence intervals. Given the authors plan to use Stata, I think the correct way to do this is to estimate the proportions using syntax like \"mlogit myvar\" and then obtain the estimates and their confidence intervals via \"margins, predict()\", but again I suggest the authors double-check this.\n\nWith respect to the \"Selection of outcomes\" section, it may be useful to rewrite item 1 in terms of \"number of model parameters\" rather than \"number of effect measures\". The meaning of the current text was not immediately clear to me.\n\nI suggest the authors plan to extract a little more detail about the methods used to analyze ITS studies, so that this can be used to inform subsequent work. For example, I am interested to know how often \"incorrect\" model assumptions are assumed, and under what conditions meaningfully incorrect conclusions result from meta-analyses that include ITS results from misspecified models. I'm thinking about the case where an outcome is bounded, where the observed pre- and/or post-interruption levels are close to the bounds, and the analysis assumes an error distribution whose domain includes values outside the bounds. For example, imagine the outcome \"percentage of prescriptions with dosage errors\", which is bounded to [0%, 100%]. If a normal error distribution is assumed in the ITS analysis, that distribution would incorrectly permit outcomes <0% and >100%. This may not be a problem if the residual standard deviation is small relative to the distance from the mean to the nearest bound, but if the observed data were close to 0% and/or 100% and the residual standard deviation was sufficiently large, the estimate of relative treatment effect estimate might be biased. I would then be interested in knowing, via empirical or simulation work from the planned subsequent research, whether these kinds of errors lead to meaningfully incorrect conclusions when such ITS results are brought into meta-analyses. If this kind of misspecification is uncommon, then perhaps that subsequent work is unnecessary, but my suspicion is that model misspecification (and this misspecification in particular) is actually very common.\n\nI'm also interested in knowing what proportion of ITSs included in meta-analyses assume ITS models that are misspecified in terms of being overly simplistic (e.g., piecewise constant time series with an instantaneous post-intervention step-change). The protocol plans to survey which ITS models are used, but it's not clear whether it aims to judge whether the use is appropriate or not. This is to some extent a judgement about whether particular models are appropriate in particular contexts (all models are approximations of reality), but it would be useful to have a solid evidence base on which to make recommendations for what people should consider when they undertake ITS-based SRs.\n\nThe protocol talks about ITS analyses that include versus exclude controls (i.e., where the interruption does not occur), but it's not clear that the protocol will extract data on this. It would be interesting to know what proportion of SRs of ITS studies permit or include uncontrolled ITS results, and ultimately to have some information about whether inclusion of such studies leads to meaningfully incorrect conclusions (I assume it does, though I could imagine that it may be possible to include controlled and uncontrolled ITS results in a meta-regression and still reasonably estimate treatment effect). Similarly, it would be interesting to know how many controls are used (i.e., I assume that ITSs with one control are common but that 10 or 20 controls are quite rare), and then from the subsequent research, the number of controls that SR protocols should specify.\n\nI suggest including the restriction to English as a possible limitation.\nI wish the authors success with their research!\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Partly\n\nAre the datasets clearly presented in a useable and accessible format? Not applicable",
"responses": [
{
"c_id": "5701",
"date": "23 Jul 2020",
"name": "Elizabeth Korevaar",
"role": "Author Response",
"response": "We would like to thank Dr Rose for their feedback on our protocol and the suggestions for its improvement. Below, we have addressed each of the items raised. Major comments:My review focuses mainly on the conceptual and statistical aspects of the protocol. I cannot comment on other aspects such as the literature search.1. Because I am more interested in the evidence that the subsequent research will hopefully deliver, I would like to see more detailed thought in this protocol about how the subsequent work will be performed. This should then inform what the product of the present protocol needs to deliver to ensure the success of the subsequent work. Perhaps this has already been thought through in detail and not presented here. However, if this work has not been done, I encourage the authors to do it and update the protocol.Thank you for your interest in our study and our future planned research. In this manuscript, we present the protocol for a systematic review of meta-analysis methods used to combine results from interrupted time series studies. While the results of our systematic review will inform subsequent research, outlining details of the future projects is beyond the scope of the current manuscript. However, we have provided more detail about the planned statistical simulation and empirical evaluation in the first paragraph of the ‘Discussion’: “The results of this review will inform our broader research program which aims to examine how different meta-analysis methods of ITS studies perform and provide guidance on the methods. Specifically, the review will identify the range of statistical methods that are used in practice, characteristics of the times series studies included in the meta-analyses (e.g. number of series, length of series) and the types of effect measures used (e.g. level change, slope change). These characteristics will inform a statistical simulation study that will examine the performance of different methods for meta-analysis of ITS studies. In addition, we will identify reviews for which the raw time series data of the included ITS studies are reported. These reviews will be used in an empirical evaluation to examine the impact of using different meta-analysis methods of real ITS studies.”. In addition, we have provided the current version of our data extraction form as an additional file (Appendix 2) in the ‘Extended data’ section. The data extraction form provides details of the data to be extracted and summarised. 2. The analysis is not planned in enough detail that it could be implemented without having to make important choices after having seen the data. I think this potentially leads to at two problems. First, being able to choose from among several possible analyses and means of presentation risks introducing bias. Second, more detailed planning at the protocol stage may prevent problems that would otherwise only become apparent while the work is being done. I suggest that the authors substantially revise this section to specify in detail what they will do and how they will report their results, including preparing skeleton tables and/or figures. I think that thinking through these issues at the protocol stage will likely make this and subsequent work (see point 1 above) more efficient, and lead to higher quality papers.We have added further detail in the ‘Analysis’ section about the general table structure we plan to use to present the results (see following). In addition, we provide the data extraction form which includes all the items we are collecting.“We will summarise the characteristics of included systematic reviews with descriptive statistics. For categorical data (e.g. the meta-analysis approach used, the risk of bias tool used) we will present frequencies (with percentages), and for numerical data (e.g. the number of meta-analysed ITS studies, the number of pooled estimates) we will present means (with standard deviations) or medians (with interquartile range).” The items (their response options – see the data extraction form in Appendix 2) and summary statistics will be grouped into tables using a structure such as the following: General study characteristics (e.g. research disciplines; types of interventions; definition used to classify interrupted time series; risk of bias or methodological quality assessment of included studies undertaken) Included study characteristics (e.g. number of included ITS studies; average number of points pre- and post-interruption) Meta-analysis methods (e.g. justification given for chosen meta-analysis method; re-analysis of primary study data undertaken, one-stage or two-stage meta-analysis used, chosen meta-analysis model, heterogeneity estimator used) Effect measures used (e.g. number of effect measure, which effect measures, standardisation of the effect measure) Additional analyses (e.g. sensitivity analyses, subgroup analyses) Completeness of reporting (e.g. reporting of combined effect, reporting of measure of precision) Minor comments:1. I suggest that the abstract is clarified to say that the authors will study the 100 most recent systematic reviews that include ITS analyses (and include the date range), rather than simply saying 100 (or whatever the final sample size is determined to be). My concern when reading the abstract was that the 100 SRs could be chosen arbitrarily, giving rise to potential bias.We have revised the text in the ‘Abstract’ as follows:“We will identify the 100 most recent reviews (published between 1 January 2000 and 11 October 2019) that include meta-analyses of ITS studies from a search of eight electronic databases covering several disciplines (public health, psychology, education, economics).”2. It would be useful for the authors to clarify that, with respect to estimating a binomial proportion, their proposed sample size of 100 would give them a worst-case margin of error of plus or minus approximately ten percentage points (i.e., 40.2% to 59.8%), if the population parameter is 50%.We have revised the text in the ‘Sample size’ section to:“Our sample size of 100 reviews was primarily selected for reasons of feasibility. A sample of this size will allow estimation of the percentage of reviews with a particular element (e.g. the prevalence of the reviews that re-analyse the primary study data) to within a maximum margin of error of 10% (assuming a prevalence of 50%). This margin of error represents the worst-case scenario and will decrease if the prevalence varies from 50%.”3. This margin of error is actually quite wide. I wonder if the authors have considered how plausible it is that the population parameter will often be close to the worst-case of 50%, and if so, whether the relatively wide confidence intervals will be informative enough for their subsequent work that will build on this paper?We believe that our sample size will generally be sufficient, such that our interpretation of the confidence interval limits will be consistent. For example, if we found that the percentage of reviews that use a particular method was 10% (95%CI: 4% to 16%), our interpretation of the limits of the confidence interval would lead to the same conclusion that the method was not commonly used. As another example, if the outcome was complete reporting of effect estimates (i.e., reporting the effect estimate and a measure of precision), and 50% of the reviews were found to completely report the effect, our interpretation would not differ at the limits of the confidence interval; if the true percentage was 40% or 60%, we would be concerned.4. The sample size of 100 seems to have been chosen under the assumption that a binomial proportion will be estimated for each factor studied (i.e., that each factor will have two levels). However, many of the criteria specified are factors with more than two levels (e.g., the protocol gives the example of three types of outcome that included reviews may study: continuous, count, and rate). Given that, I would encourage the authors to think about the more general case of estimating multinomial proportions. This would require a larger sample size for a worst-case scenario equivalent to that of a binomial distribution. A quick search identified Thompson 19871, which provides a table for estimating sample size for estimating multinomial proportions. Briefly, if the authors want to estimate multinomial proportions with 95% CIs that give a margin or error of plus or minus 10%, that paper shows that the authors should include at least 128 studies (irrespective of the number of levels of the factor studied). However, I encourage the authors to double-check this informal power analysis. The authors would then also need to use an appropriate method to estimate the proportions and their confidence intervals. Given the authors plan to use Stata, I think the correct way to do this is to estimate the proportions using syntax like \"mlogit myvar\" and then obtain the estimates and their confidence intervals via \"margins, predict()\", but again I suggest the authors double-check this.As noted in the ‘Sample size’ section, our sample size of 100 reviews was primarily selected for reasons of feasibility, so we do not have the ability to increase the sample size. Importantly, however, most items are binary and not multinomial. We acknowledge that for those few multinomial items, we will have less precision. 5. With respect to the \"Selection of outcomes\" section, it may be useful to rewrite item 1 in terms of \"number of model parameters\" rather than \"number of effect measures\". The meaning of the current text was not immediately clear to me.We have chosen the term “effect measures” since this is the terminology used in the Cochrane Handbook (Chapter 10: Analysing data and undertaking meta-analyses). The “number of model parameters” is not interchangeable with the “number of effect measures” because they have different meanings. Effect measures in ITS studies are generally computed as a combination of the model parameter estimates (e.g. area under the curve, long-term change in level).Deeks JJ, Higgins JPT, Altman DG (editors). Chapter 10: Analysing data and undertaking meta-analyses. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA (editors). Cochrane Handbook for Systematic Reviews of Interventions version 6.0 (updated July 2019). Cochrane, 2019. Available from www.training.cochrane.org/handbook.6. I suggest the authors plan to extract a little more detail about the methods used to analyze ITS studies, so that this can be used to inform subsequent work. For example, I am interested to know how often \"incorrect\" model assumptions are assumed, and under what conditions meaningfully incorrect conclusions result from meta-analyses that include ITS results from misspecified models. I'm thinking about the case where an outcome is bounded, where the observed pre- and/or post-interruption levels are close to the bounds, and the analysis assumes an error distribution whose domain includes values outside the bounds. For example, imagine the outcome \"percentage of prescriptions with dosage errors\", which is bounded to [0%, 100%]. If a normal error distribution is assumed in the ITS analysis, that distribution would incorrectly permit outcomes <0% and >100%. This may not be a problem if the residual standard deviation is small relative to the distance from the mean to the nearest bound, but if the observed data were close to 0% and/or 100% and the residual standard deviation was sufficiently large, the estimate of relative treatment effect estimate might be biased. I would then be interested in knowing, via empirical or simulation work from the planned subsequent research, whether these kinds of errors lead to meaningfully incorrect conclusions when such ITS results are brought into meta-analyses. If this kind of misspecification is uncommon, then perhaps that subsequent work is unnecessary, but my suspicion is that model misspecification (and this misspecification in particular) is actually very common.The issue outlined is interesting and it would be useful to examine under what conditions meaningfully incorrect conclusions result from meta-analyses that include ITS results from misspecified models in a simulation study. In our study we will document whether the reviewers re-analysed the primary study data (see row 90 of the data extraction form, Appendix 2); and which analysis methods were used for the primary studies, including whether the reviewers made a judgement about the appropriateness of the analysis methods used in the primary studies (see rows 93-104 of the data extraction form).7. I'm also interested in knowing what proportion of ITSs included in meta-analyses assume ITS models that are misspecified in terms of being overly simplistic (e.g., piecewise constant time series with an instantaneous post-intervention step-change). The protocol plans to survey which ITS models are used, but it's not clear whether it aims to judge whether the use is appropriate or not. This is to some extent a judgement about whether particular models are appropriate in particular contexts (all models are approximations of reality), but it would be useful to have a solid evidence base on which to make recommendations for what people should consider when they undertake ITS-based SRs.We will extract information on the model structure when the reviewers re-analyse the included ITS studies (see row 107 of the data extraction form, Appendix 2). From this information, we will summarise the different model structures (e.g. level change only, slope change only). We do not plan to judge the appropriateness of the model structures given, as the reviewer pointed out, this would require content knowledge of the studies. In addition, where reviewers directly use results from the primary studies, we will collect the review authors’ judgements of the appropriateness of the methods, which could include model structure misspecification (see rows 93-98).8. The protocol talks about ITS analyses that include versus exclude controls (i.e., where the interruption does not occur), but it's not clear that the protocol will extract data on this. It would be interesting to know what proportion of SRs of ITS studies permit or include uncontrolled ITS results, and ultimately to have some information about whether inclusion of such studies leads to meaningfully incorrect conclusions (I assume it does, though I could imagine that it may be possible to include controlled and uncontrolled ITS results in a meta-regression and still reasonably estimate treatment effect). Similarly, it would be interesting to know how many controls are used (i.e., I assume that ITSs with one control are common but that 10 or 20 controls are quite rare), and then from the subsequent research, the number of controls that SR protocols should specify.We will collect details on whether and how control series were incorporated into the meta-analysis (see rows 69-71 of the data extraction form, Appendix 2). Additionally, we will capture information regarding sensitivity analysis (e.g. comparing inclusion versus exclusion of control series, if performed). However, in this study, we will not assess whether or not the conclusion of the meta-analysis would change based on the inclusion of the control series.9. I suggest including the restriction to English as a possible limitation.We have amended the following paragraph in the ‘Discussion’ section to include the English language limitation:“While the study will be limited by our ability to identify all potentially eligible reviews and meta-analyses of ITS studies, our search strategy attempts to capture the various ways these studies are described. However, given ITS studies are often not identified as such16, and that our search is restricted to articles written in English, it is likely that we will not capture all reviews and meta-analyses that include ITS studies. Conversely, we may end up including reviews where no information regarding the definition of the included ITS studies is provided, or where an inappropriate label of ITS has been applied to included studies. While we will not exclude these reviews, we will record the reviewers’ definition of an ITS study.” Kind regards,Elizabeth Korevaar"
}
]
},
{
"id": "62669",
"date": "20 May 2020",
"name": "Alexandra McAleenan",
"expertise": [
"Reviewer Expertise Systematic reviewing."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article is a protocol for a systematic review that aims to describe the meta-analysis methods (and some of the systematic review methods) used in systematic reviews that include interrupted time series (ITS) studies.\nThe protocol is thorough and well written. A comprehensive search of the literature will be undertaken to identify reviews of health, public health, psychology, education and economic interventions. There are clear eligibility criteria for reviews, and the data to be extracted is well described, as is the planned analysis.\nI have a couple of minor suggestions. In the abstract it would be preferable to state how the 100 reviews will be selected (for e.g. \"We will identify up to the 100 most recently published reviews\"). You could also add in that you will be looking at the tools used to assess bias/methodological quality of ITS studies (which one could argue is more looking at systematic review methods that meta-analysis methods).\nYou could also clarify, by adding to the summary of data extraction items, that you will be looking at the review authors' definition of ITS studies. I presume you will also be extracting whether any of the ITS studies included a control group, and perhaps how the control group was used (especially if any re-analysis was undertaken)?\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Not applicable",
"responses": [
{
"c_id": "5702",
"date": "23 Jul 2020",
"name": "Elizabeth Korevaar",
"role": "Author Response",
"response": "We would like to thank Dr McAleenan for their feedback on our protocol and the suggestions for its improvement. Below, we have addressed each of the suggestions raised. Minor suggestions: In the abstract it would be preferable to state how the 100 reviews will be selected (for e.g. \"We will identify up to the 100 most recently published reviews\"). You could also add in that you will be looking at the tools used to assess bias/methodological quality of ITS studies (which one could argue is more looking at systematic review methods that meta-analysis methods). We have revised the ‘Abstract’ to make the selection process clearer: “We will identify the 100 most recent reviews (published between 1 January 2000 and 11 October 2019) that include meta-analyses of ITS studies from a search of eight electronic databases covering several disciplines (public health, psychology, education, economics).” In addition, we have added text to note we will be extracting information about the tools used to assess risk of bias/methodological quality: “From eligible reviews we will extract details at the review level including discipline, type of interruption and any tools used to assess the risk of bias / methodological quality of included ITS studies; at the meta-analytic level we will extract type of outcome, effect measure(s), meta-analytic methods, and any methods used to re-analyse the individual ITS studies. Descriptive statistics will be used to summarise the data.” You could also clarify, by adding to the summary of data extraction items, that you will be looking at the review authors' definition of ITS studies. I presume you will also be extracting whether any of the ITS studies included a control group, and perhaps how the control group was used (especially if any re-analysis was undertaken)? We have added “reviewers’ definition of ITS studies” to the summary of data extraction items table (Table 2). In addition, we have included the current version of our data extraction form as an additional file in the ‘Extended data’ section. We will collect data on whether and how control series were incorporated into the meta-analysis (see rows 69-71 of the data extraction form, Appendix 2). Kind regards, Elizabeth Korevaar"
}
]
}
] | 1
|
https://f1000research.com/articles/9-110
|
https://f1000research.com/articles/9-1444/v1
|
10 Dec 20
|
{
"type": "Method Article",
"title": "A guide to creating design matrices for gene expression experiments",
"authors": [
"Charity W. Law",
"Kathleen Zeglinski",
"Xueyi Dong",
"Monther Alhamdoosh",
"Gordon K. Smyth",
"Matthew E. Ritchie",
"Kathleen Zeglinski",
"Xueyi Dong",
"Monther Alhamdoosh"
],
"abstract": "Differential expression analysis of genomic data types, such as RNA-sequencing experiments, use linear models to determine the size and direction of the changes in gene expression. For RNA-sequencing, there are several established software packages for this purpose accompanied with analysis pipelines that are well described. However, there are two crucial steps in the analysis process that can be a stumbling block for many -- the set up an appropriate model via design matrices and the set up of comparisons of interest via contrast matrices. These steps are particularly troublesome because an extensive catalogue for design and contrast matrices does not currently exist. One would usually search for example case studies across different platforms and mix and match the advice from those sources to suit the dataset they have at hand. This article guides the reader through the basics of how to set up design and contrast matrices. We take a practical approach by providing code and graphical representation of each case study, starting with simpler examples (e.g. models with a single explanatory variable) and move onto more complex ones (e.g. interaction models, mixed effects models, higher order time series and cyclical models). Although our work has been written specifically with a limma-style pipeline in mind, most of it is also applicable to other software packages for differential expression analysis, and the ideas covered can be adapted to data analysis of other high-throughput technologies. Where appropriate, we explain the interpretation and differences between models to aid readers in their own model choices. Unnecessary jargon and theory is omitted where possible so that our work is accessible to a wide audience of readers, from beginners to those with experience in genomics data analysis.",
"keywords": [
"Design matrix",
"model matrix",
"contrast matrix",
"statistical models",
"gene expression analysis"
],
"content": "Introduction\n\nGene expression technologies are useful for the study of transcriptomics and their associated profiles amongst biological samples of interest. The technology is used worldwide to examine complex relationships between gene expression (which we will refer to as the response variable when performing statistical modelling) and the variables that influence the expression (referred to as explanatory variables). From the resulting datasets, careful statistical analysis can be used to find relationships that are of biological interest through the choice of appropriate statistical models applied to the data. The modelling process requires the use of a design matrix (or model matrix) that has two roles: 1) it defines the form of the model, or structure of the relationship between genes and explanatory variables, and 2) it is used to store values of the explanatory variable(s)1,2,3. Although design matrices are fundamental concepts that are covered in many undergraduate mathematics and statistics courses, their specific and multi-disciplinary application to the analysis of genomic data types through the use of the R programming language adds several layers of complexity, both theoretically and in practice.\n\nThis article describes the appropriate design matrix set up for differential expression analyses specific to using the limma4 software package, one of the most popular open-source software packages for such analysis worldwide. Our examples have been written for gene expression data, specifically with the assumption that the expression values are genewise log-count per million (log-CPM) measurements from an RNA-sequencing (RNA-seq) experiment. However, most of the concepts and R code covered in this article can also be applied to differential analyses of other genomic data types, including microarrays, ChIP-seq, ATAC-seq, BS-seq, Hi-C and proteomics. The main requirements are that the response data represents abundance on a log-scale and that each row corresponds to an appropriate genomic feature. Typically, the data table from an RNA-seq experiment contains the gene expression measurements for tens of thousands of genes and a small number of samples (usually no more than 10 or 20, although much larger sample sizes are possible). In the modelling process, a single design matrix is defined and then simultaneously applied to each and every gene in the dataset. Rather than demonstrating the application of design matrices across multiple genes, where the modelling concepts are consistent between genes, we simply describe the process for a single gene in our examples. This allows us to illustrate clearly differences between varying models and the implications of adding or removing model parameters.\n\nThe article begins by introducing the basic concepts associated with design and contrast matrices. We cover common experimental designs used in genomics research, and move onto more complex study designs as we progress through the sections. We have approached our work from a practical stand-point, with a focus on the R programming inputs and outputs, accompanied by associated plots to illustrate the observed data that we begin with and the fitted models that are produced from a graphical perspective. By omitting most of the theory associated with design matrices, our article allows readers from various backgrounds to gain a better understanding of design matrices, without having statistics as a prerequisite. To enable readers to select the most appropriate design matrix set up for their study design, we also discuss the interpretation of the models and the differences between them.\n\nIn each of our examples, we will explicitly display the observed data and include the R code for associated design and contrast matrices that are used in the modelling process. This allows readers to quickly grasp modelling concepts and to apply the R code in their own datasets. Each example is also accompanied by a figure displaying the design matrix and both a written and graphical representation of the statistical model. Whilst the complete data analysis process, from pre-processing data to variance modelling and parameter estimation is not discussed in this article, the design matrices we describe can be implemented in conjunction with the “RNA-seq analysis is easy as 1-2-3 with limma, Glimma and edgeR” differential expression workflow article5 for an RNA-seq analysis beginning with a table of counts.\n\nOther complementary work focusing on design matrices includes that of the CRAN codingMatrices package vignette6, which describes the theoretical aspects of design matrices, and the ExploreModelMatrix software package7, which allows interactive exploration of design matrices for specified explanatory variables. Although not focusing purely on design matrices, the user’s guides for the limma and edgeR8,9 software packages also contain many example case studies for different experimental designs.\n\n\nBasic models\n\nIn this section, we outline the general form of some basic models and introduce terminology that will be used in the remainder of the article. The concept of model equations and associated graphical illustrations for fitted models are also introduced here.\n\nTo begin with, let us consider two types of explanatory variables: covariates and factors. Covariates contain numerical values that are quantitative measurements associated with samples in the experiment. These can be the age or weight of an individual, or other molecular or cellular phenotypes on a sample, such as measurements obtained from a polymerase chain reaction (PCR) experiment or fluorescence activated cell sorting (FACS). For covariates, it is generally of interest to know the rate of change between the response and the covariate, such as “how much does the expression of a particular gene increase/decrease per unit increase in age?”. We can use a straight line to model, or describe, this relationship, which takes the form of\n\nLEFT: The basic model for covariates is referred to as a regression model, which is a line defined by the model parameters β0 the y-intercept, and β1 the slope. CENTER: One of two basic models for factors is referred to as a means model, where model parameters are calculated as the mean gene expression of each level of the factor e.g. β1 represents the mean expression for wildtype and β2 represents the mean of mutant. RIGHT: The other basic model we refer to for factors is a mean-reference model, where the first model parameter is calculated as the mean gene expression of the reference level, and subsequent parameters are calculated relative to the reference level e.g. β1 represents the mean expression for wildtype and β2 represents the difference between mutant and wildtype. In each plot, the points represent the original data; coloured lines are used to represent expected gene expression, where dashed lines are specifically used to represent expected gene expression for non-reference levels in the mean-reference model e.g. mutant.\n\nFactors are categorical variables or classifiers associated with samples in the experiment. They are often separated into those that are of a biological nature (e.g. disease status, genotype, treatment, cell-type) and those that are of a technical nature (e.g. experiment time, sample batch, handling technician, sequencing lane). Unique values within a factor are referred to as levels. For example, genotype as a factor may contain two levels, “wildtype” and “mutant”. Here, it is generally of interest to determine the expected or mean gene expression for each state or level of the factor. The relationship between gene expression and the factor can be described, or modelled, in the form of\n\nEach of the horizontal lines in Figure 1 are defined by their y-intercept (and a slope of 0), and are themselves regression models. We, however, will refer specifically to models of this type as a means model since the model parameters represent the group means. This also allows us to differentiate these models from the general regression models applied to covariates where the y-intercept and slope can both be non-zero. As noted for covariates, the true values for the model parameters are unknown but estimable. Whilst the expected expression of each factor level is informative, it is often the difference in expression between levels that are of key interest, e.g. “what is the difference in expression between wildtype and mutant?”. These differences are calculated using linear combinations of the parameters (a fancy way to say that we multiply each parameter by a constant) which we refer to as contrasts. For example, a contrast of (1,−1) calculates β1 − β2, the difference in means between wildtype and mutant.\n\nAn alternative parameterisation of the means model directly calculates the gene expression difference between mutant and wildtype. It does so by using one of the levels as a reference. Such a model is parameterised for the mean expression of the reference level (e.g. wildtype), and the rest of the levels are parameterised relative to the reference (e.g. the difference between mutant and wildtype). The relationship between gene expression and genotype is modelled in the form of\n\nThe terminology and concepts covered in this section are summarised in the table below, in the context of modelling gene expression data. The table also extends to some definitions and descriptions covered later in the article, and is a useful resource to refer to from time-to-time.\n\n\nOverview of models fitted\n\nIn the sections to follow, we explore various models for explanatory variables that are factors, starting from the most basic study designs to those that are more complex. We then cover some models for explanatory variables that are covariates. The tables below summarise the data examples, R input for the associated design matrices, and the sections from which they can be found. In the tables, factors are distinguished from covariates by the presence of subscripts listing their levels e.g. “factorLEVEL1,LEVEL2”. Associated sections are marked with an asterisk if the design matrix cannot be sufficiently summarised within the table.\n\nThis section describes and compares models that are coded with and without an intercept term for covariates and factors. It also shows the fundamental elements for computing differences between model parameters using contrasts and contrast matrices.\n\nThis section examines a study on four treatment groups (CTL, I, II, and III). The example here represents a study design that is very common in practice, where there are several treatments (or conditions or groups) including a control. Comparisons between the levels are computed using two alternative design matrices. In this section, we also look at more complex set ups for contrasts to compute comparisons that may be of interest.\n\nIn this section we consider the effect of combining two separate treatments. Our first example looks at the interactivity of the treatments using a model from the previous section, which has a single treatment factor. We then show an alternative method to calculate the same estimates using a two factor model.\n\nThis section looks at studies with multiple factors. It includes study designs that are more complex in nature and describes the approaches one can take to examine the differences of interest. The section covers studies with nested factors and the fitting of mixed effects models.\n\nIn this section, we look at explanatory variables that are covariates rather than factors. We begin with fitting some simple models and work up towards more complex ones such as the fitting of cyclical models.\n\n\nDesign and contrast matrices\n\nIn this section, we demonstrate how design and contrast matrices can be created for the most basic experimental designs with covariates and factors. Specifically, we discuss the similarities and differences between design matrices that include and exclude an intercept term.\n\nDesign matrices are used in the estimation process of model parameters. The design matrix has columns associated with the parameters and rows associated with samples (Figure 2). If the estimated parameters are not of direct interest, a contrast matrix can be used to calculate contrasts of the parameters. Combining multiple contrasts, each column in the contrast matrix represents a single contrast, and has rows associated with columns in the corresponding design matrix (Figure 2). Using the R programming language, we code for design matrices using the model.matrix function from the stats package, and contrast matrices using the makeContrasts function from the limma package. To learn more about these functions, one can bring up associated help pages by typing ?stats::model.matrix and ?limma::makeContrasts in the R console.\n\nFor a single explanatory variable, which we simply call variable, a design matrix can be coded by model.matrix(~variable) to include an intercept term, or by model.matrix(~0+variable) to exclude the intercept term. One of the most fundamental concepts in the coding of design matrices is to understand when one should include an intercept term, when not to, and how it affects the underlying model. If variable is a factor, then the two models with and without the intercept term are equivalent, but if variable is a covariate the then two models are fundamentally different.\n\nUsing age as an example, let’s look at the gene expression of mice where their age in weeks from birth were recorded. The expression of a single gene is recorded as a numerical vector called expression. The age of mice is also recorded as a numerical vector in age (as weeks), and we use an additional mouse character vector to show that these are independent measurements. The three vectors, expression, mouse and age, that represent our example data are displayed below as a data frame as follows:\n\n\n\nA design matrix with an intercept term can be coded as model.matrix(~age). The resultant design matrix, which is displayed in Figure 3, contains a column of 1s (the intercept term) and a column with values taken from age. This design matrix is associated with a regression model (see Figure 1), where the intercept term in the first column is parameterised for the y-intercept, and “age” in the second column is parameterised for the slope of the regression line. In other words, the second column is used to estimate the rate of change in gene expression per unit increase in age.\n\nThis particular model includes an intercept term so that the model (line) has flexibility in intersecting the y-axis at any point. MODEL: The fitted model in written form, where y represents gene expression and E(y) expected gene expression. Estimated model parameters are highlighted in colour. MATRIX: R input and output for the associated design matrix, with the colour of column names (model parameters) matching that of the estimated parameters above. For simplicity, ”assign” and ”contrasts” attributes of the design matrix are not displayed in our figures (see ‘?stats::model.matrix’ for more). PLOT: Observed data points are drawn together with the fitted model representing expected gene expression. Where appropriate, aspects of the fitted model are drawn in a colour that matches associated parameter estimates.\n\nThe parameters can be estimated as 1.85 for the y-intercept and 0.54 for the slope. This means that for every 1 unit (or week) increase in age, gene expression increases by a value of 0.54 on average. We can write our statistical model using the estimated model parameters to give us our fitted model. The fitted model can be written as E(y) = 1.85 + 0.54x, with y representing expression of the gene, E(y) representing the expected gene expression and x representing age.\n\nAlternatively, a design matrix without an intercept term can be coded as model.matrix(~0+age). This design matrix contains a single column that represents age, as shown in Figure 4, which is parameterised for the slope of a regression line. By adding a 0 to the formula in model.matrix, the intercept term has been removed. This means that the regression line is forced to intercept the y-axis at 0. The slope of the line can be estimated to be 0.97, such that gene expression is expected to increase by 0.97 for every 1 unit increase in age. The fitted model is written as E(y) = 0.97x.\n\nThis restriction is due to the design matrix set up which excludes the intercept term.\n\nWhen comparing between the models with and without an intercept term, we observe that the model without an intercept term (Figure 4) does not fit as closely to the observed data as the model with an intercept term (Figure 3). It is not surprising that the model with an intercept term provides a better fit to the data since it is less restrictive (allows the y-intercept to be at any point) than the model without an intercept term. The extra parameter in the model allows it to be more flexible. In general, we suggest the inclusion of an intercept term for modelling explanatory variables that are covariates since it provides a more flexible fit to the data points. A model without an intercept term would only be recommended in cases where there is a strong biological reason why a zero covariate should be associated with a zero expression value, and such contexts are rare in gene expression modelling.\n\nNow we consider an example of gene expression on healthy and sick mice, each in triplicate. Healthy and sick mice are classified using a group factor which contains two levels, HEALTHY and SICK. The level names are written in all capitals so that design matrices have column names that are easier to read by default e.g. “groupSICK” (Figure 5 and 6) rather than “groupsick” or “groupSick”. The data is displayed by combining vectors for expression, mouse and group as follows:\n\n\n\nThe associated design matrix includes an intercept term, where healthy mice acts as the reference level. The expected gene expression of non-reference levels, e.g. that of sick mice, are represented by dashed lines in the plot.\n\nThe associated design matrix excludes an intercept term.\n\nA design matrix with an intercept term can be coded as model.matrix(~group) to obtain a two column matrix (Figure 5). In general, the resulting design matrix will have the same number of columns as the factor group has levels. The design matrix contains a column of 1s (the intercept term), and a column with values as 0s or 1s (a value of 1 when the associated sample is in the sick group, and 0 otherwise). This design matrix is parameterised for a mean-reference model (Figure 1), where the intercept term in the first column is parameterised for the mean expression of the healthy group, and the second column is parameterised for the mean expression of the sick group relative to healthy (difference between sick and healthy). The healthy group is selected as the reference level since HEALTHY is the first level in group. Levels in a factor are ordered alphanumerically by default, but re-specification of the reference can be carried out using the relevel function in the stats package. Using this design matrix, the parameters can be estimated to be 2.95 and 1.62, such that the mean gene expression of the healthy group is 2.95, and the mean gene expression of the sick group relative to healthy is 1.62. We can calculate the mean expression of the sick group by summing both parameter estimates, which in this case is equal to 4.57. Using the estimated values, the fitted model for expected gene expression can be written as E(y) = 2.95 + 1.62x, where x is an indicator variable for mice that are sick. The indicator variable, or x, takes the value of 1 when calculating the expected expression of sick mice, and takes the value of 0 when calculating the expected expression of healthy mice. In other words, E(y) = 2.95 for healthy mice and E(y) = 4.57 for sick mice.\n\nA design matrix without an intercept term can be coded as model.matrix(~0+group), which gives an equivalent model to that of the previous model. The design matrix here also contains two columns (Figure 6), but is instead parameterised for a means model (Figure 1). This means that the first column of the design matrix is parameterised for the mean expression of the healthy group, where a value of 1 is present when the associated sample belongs to the healthy group, and 0 otherwise. The second column is parameterised for the mean expression of the sick group, and has a value of 1 when the associated sample belongs to the sick group, and 0 otherwise. The parameters in this model can be estimated as 2.95 and 4.57 for the mean gene expression of healthy and sick mice, respectively. Thus, the fitted model for expected gene expression can be written as E(y) = 2.95x1 + 4.57x2, where x1 and x2 are indicator variables for healthy mice and sick mice respectively. In other words, E(y) = 2.95 for healthy mice and E(y) = 4.57 for sick mice.\n\nAs mentioned in our earlier description of basic models (Figure 1), models with and without an intercept term are equivalent for factor explanatory variables, but differ in parameterisation. This means that the expected expression values for healthy and sick mice are the same regardless of whether a means model (without an intercept term in the design matrix) or a mean-reference model (with an intercept term in the design matrix) is fitted. The only difference is that the expected gene expression for sick mice is calculated by summing both parameter estimates in a mean-reference model, whereas it is estimated directly as the second parameter in a means model. For this reason, it ultimately does not matter which design matrix is used. We recommend the use of whichever design matrix that is better understood by the reader, which is often the design matrix without the intercept term since the interpretation of parameters is more straightforward.\n\nWhen fitting a means model, the parameter estimates themselves are usually not of direct interest. It is the difference between the parameter estimates, or difference between mean expression of groups, that is of interest. The difference in parameter estimates can be calculated using a contrast matrix via the makeContrast function. To specify the comparison of interest, column names from the design matrix, “groupHEALTHY” and “groupSICK”, are inserted into the function. The design matrix and associated contrast matrix is coded as follows:\n\n\n\nThe makeContrast function simply creates the contrast of (-1, 1) which subtracts the first parameter estimate (mean expression of healthy) from the second parameter estimate (mean expression of sick). Using the parameter estimates estimated earlier (Figure 6), the contrast calculates -2.95 plus +4.57 which equals 1.62. In other words, we expect gene expression of sick mice to be upregulated by 1.62 units relative to healthy mice. Notice how this is the same value as the second parameter estimate in the mean-reference model (Figure 5), since that model is directly parameterised for the difference between sick and healthy mice. It is also reasonable to compute the difference in the opposite direction, by having groupHEALTHY-groupSICK as the first argument of the makeContrasts function. This will result in the value of -1.62 instead. The two options only differ in their interpretation, “gene expression of sick mice is greater than healthy mice by a value of 1.62” versus “gene expression of healthy mice is greater than sick mice by a value of -1.62”.\n\n\nStudy of treatments and control\n\nIn this section, we focus on a single factor as an explanatory variable to modelling gene expression. The factor we use contains several levels, which allows us to discuss some common comparisons of interest, and show different methods of calculating those differences.\n\nA very common study design examines several conditions of interest, where one condition represents the control. Considering such an experimental design, we want to model the relationship between gene expression and four possible conditions: three treatments and a control. The explanatory variable is set up as a factor vector, which we have named treatment, and the factor is used to classify samples into the control group (CTL), treatment I, treatment II, and treatment III. The factor has a total of 4 levels. Gene expression is recorded as a numeric vector called expression, and mouse is a character vector showing that the observations are independent measurements. The data combines the vectors as follows:\n\n\n\nWe know from the previous section that the treatment factor can be represented in a means model or a mean-reference model using the design matrices coded as model.matrix(~0+treatment) or model.matrix(~treatment) respectively. Either representation would give equivalent models, and so it would be unnecessary to describe both models for the same exercise. Based on the comparison of interest at hand, we demonstrate the use of one of the models using the most direct approach.\n\nFor a comparison of each treatment group versus the control group, we model gene expression using a mean-reference model. This is ideal since the differences can be estimated directly from the model parameters, and without the use of an additional contrast matrix. To do this, the control group would act as the reference and must be the first level of the treatment factor vector. We can view the order of levels by levels(treatment). If the level associated with the control group is not listed first, it can be changed to the first or reference level with the code treatment <- relevel(treatment, ref=\"CTL\").\n\nWe can now create a design matrix that represents a mean-reference model by model.matrix(~treatment) (Figure 7). The columns of the design matrix represent the mean expression of the control group, and the difference in mean expression between treatment I and control, treatment II and control, and treatment III and control. Using the design matrix, the model parameters are estimated as 1.03, 1.09, 1.97 and 3.87. This means that the difference in expected gene expression between treatments and control are 1.09 for treatment I and control, 1.97 for treatment II and control, and 3.87 for treatment III and control. Treatment III has the greatest expected gene expression difference from the control. The fitted model for expected gene expression can then be written as E(y) = 1.03 + 1.09x1 + 1.97x2 + 3.87x3, where the x’s are indicator variables for treatment I, treatment II and treatment III, respectively. In other words, x1 = 1 for treatment I, x2 = 1 for treatment II, and x3 = 1 for treatment III. The x’s are equal to 0 elsewhere.\n\nThe design matrix that is used includes an intercept term which represents the mean gene expression of the control group, or the reference level in the treatment factor. Other levels in the factor have mean gene expression represented relative to the control group. This means that the second to fourth parameters in the mean-reference model represent gene expression differences between treatment groups and the control group. The x’s in the model are indicator variables for treatment groups, with x1 = 1 for treatment I, x2 = 1 for treatment II, and x3 = 1 for treatment III.\n\nIn order to make all possible pairwise comparisons between the treatments, we model gene expression using a means model. Due to its parameterisation, the means model is simple to work with when specifying the comparisons of interest in the contrast matrix.\n\nThe associated design matrix is coded as design <- model.matrix(~0+treatment), with columns or parameters representing the mean gene expression of each control and treatment group (Figure 8). The mean expression values can then be estimated as 1.03, 2.12, 3 and 4.9; where the fitted model for expected gene expression is written as E(y) = 1.03x0 + 2.12x1 + 3x2 + 4.9x3. The x’s are indicator variables for control, treatment I, treatment II and treatment III, respectively. Specifically, x0 = 1 for control, x1 = 1 for treatment I, x2 = 1 for treatment II, and x3 = 1 for treatment III, and 0 elsewhere.\n\nThe design matrix that is used excludes the intercept term so that the associated model is a means model. In other words, the mean gene expression of each level in ‘treatment’ is represented by a parameter in the model. The x’s in the model are indicator variables for control and treatment groups, with x0 = 1 for control, x1 = 1 for treatment I, x2 = 1 for treatment II, and x3 = 1 for treatment III.\n\nTaking these parameter estimates, we compute all pairwise differences between treatments using the makeContrasts function as follows:\n\n\n\nNote that there are six possible pairwise comparisons between the four treatments. Note also that default column names in the contrast matrix have been abbreviated here using the abbreviate function from the base package so that the contrast matrix can display neatly above (although this step is not usually necessary). The contrast matrix contains six columns, each representing one comparison: “tI-C” for treatment I versus control, “tII-C” for treatment II versus control, “tIII-C” for treatment III versus control, “tII-I” treatment II versus I, “trIII-I” for treatment III versus I, and “tIII-II” treatment III versus II. The 1s and -1s in each column of the contrast matrix mark the parameters from the design matrix from which comparisons are made. For example, the first contrast subtracts the mean expression of the control group (parameter 1) from the mean expression of treatment II (parameter 2), which is calculated as 1.09. In such a way, differences between treatments or between treatments and control are estimated. The difference in mean gene expression is estimated as 1.97 for treatment II versus control, 3.87 for treatment III versus control, 0.88 for treatment II versus I, 2.78 for treatment III versus I, and 1.9 for treatment III versus II.\n\nRather than considering each treatment-control comparison separately, suppose that it is of interest to compare the control group to all of the treatment groups simultaneously. The idea of this is to find the genes that may define the control relative to the treatments. The same can also be carried out for individual treatment groups. For example, we could also consider the genes that define treatment I relative to the rest of the groups.\n\nWhen comparing the control group to the rest of the groups, it is not advisable to merge treatments I, II and III into one big treatment group, and to simply fit a separate model for the combined treatment group and control. The combined treatment group does not account for group-specific variability, and the combined group would be biased towards larger treatment groups in an unbalanced study design. Instead, we demonstrate two methods to approach this. Both methods can use either of the fitted models from the previous sections (mean-reference or means model), where individual group means and variability are accounted for. The first method uses a contrast matrix to compare the control group to the treatment average, and the second looks at the overlap between treatment-control comparisons.\n\nUsing the means model defined earlier, we calculate the average of the mean gene expression of treatment groups. We then subtract the mean gene expression of the control group from the average treatment value. To do this, a contrast matrix is coded as follows:\n\n\n\nwhich calculates (2.12+3+4.9)/3 - 1.03 and is equal to 2.31. Notice how the parameter estimates for treatment groups are divided by 3, the number of treatment groups under consideration. This is important as it ensures the correct calculation of averages. What this method says is that the average gene expression of the treatment groups is greater than the control group by 2.31. In our case, the gene expression of each treatment group is also greater than the control. It is worth noting, however, that the average gene expression of the treatment groups can be greater than that of the control group when individual treatment groups are not necessarily all greater than the control.\n\nFor a more stringent approach that ensures that gene expression in each of the treatment groups are greater (or lower) than the control, we use a method of overlaps. Taking results from three treatment-control comparisons, we overlap or take the intersection of the genes that are significantly up-regulated (or down-regulated). Significance is usually defined by an adjusted p-value cut-off of 5%, but it can also be defined at varying thresholds or by using other summary statistics such as log-fold-changes. Notice that we take the direction of change into consideration so that genes are consistently up- or down-regulated in the control group. The direction of change can be determined by log-fold-change values, t-statistics or similar statistics. In the case where there are only a small number of significant genes in each of the treatment-control comparisons, the method described here can be overly stringent and result in no overlapping genes in the set. If this is the case, it would be reasonable to relax the threshold for defining significance.\n\nLet us suppose it is of interest to compare the gene expression of two groups against another two other groups. This may be of interest if there are prior expectations that two groups are more similar to each other than the other two. In this example, we compare control and treatment III against treatment I and II by applying the contrast coded as\n\n\n\nto the means model. In defining the contrast, parameter estimates are divided out by the number of groups used to calculate the average. Using the parameter estimates, the difference in the 2 versus 2 group comparison is calculated as (1.03 + 4.9)/2 - (2.12 + 3)/2, which equals 0.41.\n\n\nStudy of interactions and additivity of treatments\n\nIn this section, we reconsider the same experimental data as in the previous section, but we now suppose the treatment III is a combination of treatments I and II. Here we are interested in examining the effect of combining treatments I and II relative to their individual effects. We approach this using two methods. The first simply uses the parameter estimates that we have already calculated from the previous section, meaning that we use the single treatment factor to allocate sample information on the treatment and control types. The second approach uses two separate factors, which we will call treat1 and treat2, to allocate sample information on whether treatment I and/or treatment II were administered.\n\nUsing the first approach, we model the relationship between gene expression and the treatment factor with a mean-reference model. Taking the corresponding parameter estimates from the mean-reference model, such that we use the design matrix coded as model.matrix(~treatment), we find that the effect of treatment I relative to control is 1.09, such that the difference in means between treatment I and control is 1.09. The relative effect of treatment II is 1.97, and the relative effect of the combined treatment (previously referred to as treatment III) is 3.87. For simplicity, let us refer to these relative effects as A, B and C.\n\nWe consider the combined treatment to have an additive effect if the combined treatment effect is equal to the sum of the two individual effects, such that C − A − B = 0, which we simplify to δ = 0. On the other hand, we consider the combined treatment to have an interaction effect if the combined treatment effect is not equal to the sum of the two individual effects, such that δ≠0. An interaction effect is considered to be synergistic if the combined effect is greater than the sum of the individual effects (δ > 0), and is considered repressive if the combined effect is less than the sum of the individual effects (δ < 0). As you can see, it is of interest to determine the value of δ, which we call the interaction term. Using a design matrix with an intercept term, we define the interaction term, or δ = C − A − B, as a contrast in the makeContrast function, as follows:\n\n\n\nTaking the parameter estimates from the mean-reference model, this simply calculates δ as 3.87-1.09-1.97, which equals 0.82. Since the interaction term is a positive value, we conclude that combined treatment effect is interactive and synergistic.\n\nNote that in running the makeContrasts function above, the function automatically converted the “(Intercept)” column in the design matrix to “Intercept” since the brackets are syntactically invalid. To avoid distracting from the results, we suppressed the display of its warning message referring to this in our output above.\n\nAnother way to approach the same problem is by reassign the explanatory variable into two factors representing the presence and absence of the treatments. The factors treat1 and treat2 are defined as follows:\n\n\n\nHere treat1 indicates the presence or absence of treatment I and treat2 indicates the presence or absence of treatment II. The two factors allow us to create a model that directly includes an interaction term. The associated design matrix is coded as model.matrix(~treat1*treat2), where an asterisk is placed between the two factors (Figure 9). The design matrix is parameterised for the mean gene expression of the control group (first column), the difference in mean expression between treatment I and control (second column), treatment II and control (third column), and the interaction term (last column). Using this design matrix, the parameters can be estimated as 1.03, 1.09, 1.97 and 0.82. In other words, the interaction term is estimated as 0.82, and has the same value as calculated previously.\n\nThe design matrix that is used includes an interaction term in the last column, and we refer to the associated model as an interaction model. The interaction term can be used to indicates whether the combined administration of treatments I and II have an additive effect (interaction term equal to zero), have a synergistic effect (interaction term has a positive value), or have a repressive effect (interaction term has a negative value). In this example, the interactive effect is estimated as 0.82. The x’s in the model are indicator variables for treatment I and treatment II, where x1x2 is only equal to 1 if both treatments are present.\n\nMoreover, whether we use a single treatment factor or the two factors here, the two models are equivalent, differing only in parameterisation. The interaction model fitted here can be written as E(y) = 1.03 + 1.09x1 + 1.97x2 + 0.82x1x2, where the x1 and x2 are indicator variables for treatment I and treatment II respectively. Specifically, the fourth term in the model, the interaction term, is only included in the presence of both treatments, such that x1x2 = 1 but is 0 elsewhere.\n\nWhilst the interaction model is useful in identifying the effect of the combined treatment via the interaction term, such a model may not always be of interest. One may simply want to quantify the individual effects of treatment I and treatment II, and prefer the assumption that a combined treatment results in the additivity of the two effects. This means that we use all of the samples associated with treatment I (treatment I only and in combination with treatment II) to estimate the effect of treatment I. The same goes for treatment II.\n\nUsing the two factors treat1 and treat2, we create an additive model that excludes the interaction term. The associated design matrix is coded as model.matrix(~treat1+treat2), where a plus sign is placed between the two factors (Figure 10). The design matrix contains 3 columns that are identical to the first three columns of the design matrix from the interaction model (Figure 9). The interpretation of those parameters remain the same. The first parameter represents the mean expression of the control group, the second represents the difference in mean expression between treatment I and control, and the third parameter represents the difference in mean expression between treatment II and control. The parameter estimates for this model can be calculated as 0.83, 1.5 and 2.37. The mean expression of the combined treatment can be calculated by combining all parameter estimates (gene expression from the control group, and the relative change when treatments I and II are added), such that it is equal to 0.83+1.5+2.37=4.7. We can write the fitted additive model as E(y) = 0.83 + 1.5x1 + 2.37x2, where the x1 and x2 are indicator variables for treatment I and treatment II. Relative to the interaction model, the fit of the additive model results in expected gene expression values that are further from the observed values. This is not unexpected since fewer parameters are used to model the relationship between gene expression and groups, and we know from the interaction model that the interaction term is non-zero. Even so, the additive model may be preferred for its simple interpretation and thus may be more applicable to some studies.\n\nThe design matrix that is used excludes the interaction term, and we refer to the associated model as an additive model. This means that a combined treatment is assumed to have the additive effects of individual treatment I and treatment II effects. The x’s in the model are indicator variables for treatment I and treatment II.\n\n\nStudies with multiple factors\n\nIn this section, we examine several study designs that contain two or more factors as explanatory variables. We begin with an example where we convert two factors of interest into one, and then consider cases where there are factors that are not of interest. In the second half of this section, more complex study designs are introduced, such as scenarios where there are nested factors and repeated measurements. We finish off the section by fitting a mixed effects model using functions from the limma package, where we treat a factor that is not of interest to the study as a random effect.\n\nExperimental studies often include multiple factors of interest. This could involve different treatments, cell types, tissue types, sex, and so on. Let us consider an experiment on lung and brain samples that are enriched for B-cells and T-cells. The data is as follows:\n\n\n\nFor this experiment, there are several comparisons of interest: 1) overall differences between cell types, 2) overall differences between tissues, 3) differences between cell types within each tissue type, and 4) differences between tissues within each cell type. The simplest method is to merge tissue and cells factors into a single group factor, as follows:\n\n\n\nThis allows us to fit a means model to the data, using a design matrix coded as design <- model.matrix(~0+group) (Figure 11), and to define contrasts for comparisons of interest using the makeContrasts function. The contrasts are coded as\n\n\n\nwith columns of the matrix representing 1) overall differences between cells, B-cells versus T-cells; 2) overall differences between tissues, lung versus brain; 3) differences between cells within lung, and 4) differences between cells within brain; 5) differences between tissues within B-cells, and 6) differences between tissues within T-cells. Notice that we specified our own contrast names in the code above. The row names were also shortened so that the contrast matrix could display neatly.\n\nThe group factor is converted from two factors representing tissue samples and cell types.\n\nUsing the design matrix, the parameters are estimated as 1.03 for the mean gene expression of B-cells in the lung, 2.12 for B-cells in the brain, 3 for T-cells in the lung, and 4.9 for T-cells in the brain. By applying the contrast matrix to the estimated parameters, we calculate that overall gene expression difference between B-cells versus T-cells is -2.37, and -1.5 for lung versus brain. B-cells and T-cells differ by -1.97 in the lung, and -2.78 in the brain. Lung samples and brain samples differ by -1.09 in B-cells, and by -1.9 in T-cells.\n\nSome factors within an experiment may not be of biological interest. Often they are technical factors such as handling technician, experimental time if samples were processed in separate batches, or the sequencing lane on which the samples were processed on. There are also biological factors that may not be of direct interest; such as ethnicity of patients in a human drug trial or the sex of individuals from which samples were taken. Let us consider an experiment with mice belonging to groups A, B, C, or D, each in triplicate. It is of interest to compare gene expression between the groups. In the process of the experiment, two sequencing lanes (L1 and L2) were used for sequencing and samples were processed by different technicians (I and II). To ensure that differences detected between groups are not influenced by these factors, we can account for any differences between the sequencing lanes and handling technician in our modelling process. The data is as follows:\n\n\n\nA means model can be fitted to the data, with a design matrix coded as design <- model.matrix(~0+group+lane+technician) to model gene expression in groups, while accounting for effects resulting from differences in lane and technician. The first 4 columns of the design matrix are associated with parameters for the mean expression of group A, B, C and D (Figure 12). Specifically, the group means are parameterised for when the samples are in lane L1 and processed by technician I.\n\nThe design matrix excludes the intercept term for the first factor added to the function. Only lines reflecting the first 4 parameters are drawn in the plot, representing the mean gene expression of groups A, B, C and D in lane L1 and with handling technician I. Samples are labelled by their sequencing lane (L1 or L2), and coloured black if they are processed by technician I, yellow if they are processed by technician II.\n\nThe fifth column in the design matrix is parameterised for difference between lane L2 and lane L1 (for group A samples processed by technician I), and the sixth column is parameterised for the difference between technician II and I (for group A samples in lane L1). Although an intercept-free design matrix has been coded using the 0+ notation, the intercept is only excluded from the first factor that is listed within the model.matrix function. In other words, the second and third factors added to the model.matrix function are parameterised as though there is an intercept term. This is why we place the factor of interest first as it simplifies the subsequent code for the comparisons of interest, even though a different order of factors added give equivalent models with variations in parameterisation.\n\nBy estimating model parameters, the fitted model can be written as E(y) = 0.92x1 + 2.04x2 + 2.87x3 + 4.74x4 + 0.1x5 + 0.15x6, where x1 to x4 are indicator variables for groups A to D, respectively. Additionally, x5 is an indicator variable for lane L2, and x6 is an indicator variable for technician II. In other words, a group A sample processed in lane L1 and by technician I has expected gene expression E(y)=0.92. Whereas, the expected gene expression is E(y)=0.92+0.1=1.02 if it were processed in lane L2, E(y)=0.92+0.15=1.06 for technician II, and E(y)=0.92+0.1+0.15=1.16 for a group A sample processed in lane L2 and by technician II.\n\nFor comparisons between groups, we form contrasts using only the first 4 parameter estimates, and keep lane and technician consistent. For example, a contrast comparing group A to group B can be coded as makeContrasts(groupA-groupB, levels=colnames(design)). All other pairwise comparisons can also be included into the contrast matrix.\n\nWhen modelling multiple factors of interest, the factors may be converted into a single factor for modelling, as shown in the previous section. We also note that it may not be sensible to add all known factors associated with the experiment. This could well exceed the number of degrees of freedom available for modelling (too many parameters when compared to the number of data points). A reasonable way to check the factors that should be accounted for include the use of unsupervised clustering plots, such as principal components analysis (PCA) or multi-dimensional scaling (MDS). Factors associated with separation between sample clusters should be included in the model. An alternative method is to fit a model to the biological groups of interest with one addition factor to observe whether the factor has substantial influence on gene expression (such that many genes are detected as differentially expressed for that factor). Repeat this for subsequent factors to determine the factors that should be included into the final model.\n\nNow consider a study design that includes two of the factors, group and batch, representing biological groups of interest and experimental batches. The samples in group A and group B are processed in batch B1, whilst samples in group C and group D are processed in batch B2. We say that the groups are nested within batches. The data is as follows:\n\n\n\nIt is of interest to compare the gene expression between groups. Naturally, one may include both factors into a design matrix coded as design <- model.matrix(~0+group+batch) or design <- model.matrix(~0+batch+group). This, however, produces a design matrix that is not of full rank, meaning that there are more columns in the design matrix (5 columns in this case) than what is needed (4 columns). The resultant design matrix has some columns that are linearly dependent, which is due to batch information being redundant once all group means are defined. This is because batch B1 is uniquely defined by group A and B, and batch B2 is uniquely defined by group C and D. Similarly, two of the groups are redundant if batch means are defined first. One would usually notice that their design matrix is not of full rank when the parameter estimation process returns NA or non-estimable results for some parameters.\n\nTo check for redundancy of model parameters, one can compare between the number of columns in the design matrix with ncol(design) to the rank of the matrix with qr(design)$rank. This would show that there are 5 columns in the design matrix but only a rank of 4, meaning that one of the parameters defined in the design matrix is linearly dependent. This should prompt us to consider how to set up the model properly, figuring out which factors are dependent on others, and ultimately redefining the design matrix. For example, the design matrix can be set to model.matrix(~0+group) instead, although we should keep in mind that some pairwise group comparisons would be confounded by batch effects, such as when comparing group A to group C.\n\nIn a study of treatment effects, gene expression measurements were taken from mice at multiple time points. Three of the mice were administered treatment X, and another three were administered treatment Y. Measurements were taken for the mice at two timepoints, T1 (baseline) and T2. The data is as follows:\n\n\n\nIt is of interest to compare timepoint T1 with T2 within treatment X, while accounting for how the samples are paired. What this means is that it is important to account for the relative change from timepoint T1 to T2 of each mice, when estimating the overall change between the timepoints. Similarly, a comparison between timepoint T1 and T2 within treatment Y is of interest. Additionally, we want to examine the overall differences between treatment X and Y. Since the mice are nested within treatment types, we create a custom design matrix to avoid a matrix that has linearly dependent columns or that is not of full rank. The custom design matrix is created using model.matrix and cbind functions in the following way:\n\n\n\nIn the first step, model.matrix(~0+id), we account for each individual mouse and the pairing of samples. Since mice are nested within treatments, the treatment effects are encompassed within the mouse effects. In the second step, cbind(design, X= treatment==\"X\" & timepoint==\"T2\"), an extra “X” column is added to the design matrix to represent treatment X at timepoint T2. Similarly, the third step, cbind(design, Y= treatment==\"Y\" & timepoint==\"T2\"), appends column “Y” to the design matrix to represent treatment Y at timepoint T2. The two additional columns differentiate between samples at timepoint T1 (first 6 columns) from those at timepoint T2 (last 2 columns), such that the first 6 columns in the design matrix now represent the effect of each mouse at timepoint T1, and the last 2 columns represent the overall difference between timepoint T2 and T1 for treatments X and Y respectively (Figure 13).\n\nRepeated measurements are taken from mice, as indicated by the numbers in the plot, such that label ”1” represents MOUSE1. A custom design matrix is created to model mouse IDs, treatment and timepoint (complete R code shown in the main article). Fitted lines are drawn in pink for treatment X, and in aqua for treatment Y. Solid lines represent expected gene expression at timepoint T1, and dashed lines for timepoint T2.\n\nThe first 6 parameters, which represent the expected expression of each mouse at timepoint T1, can be estimated as 0.96, 1.15, 0.98, 2.99, 3.07 and 2.93. Using the first three estimates (0.96, and 1.15, 0.98), we can calculate the expected expression of treatment X at timepoint T1 by taking the mean value, which equals 1.03 (and is marked by the thick, dark red line in the plot of Figure 13). Similarly, the next three estimates (2.99, 3.07 and 2.93) can be used to calculate the expected expression of treatment Y at timepoint T1, which equals 3 (and is marked by the thick, dark green line in the plot of Figure 13). The seventh parameter is estimated at 1.09, and the eighth is 1.9, such that gene expression is greater by 1.09 and 1.9 at timepoint T2 relative to timepoint T1 for treatments X and Y respectively. The overall difference between treatments X and Y can be coded as makeContrasts(X-Y, levels=colnames(design)), which calculates the difference between the seventh and eighth parameters, and has a value of -0.82.\n\nIn the previous section, repeated measurements taken from mice receiving treatment X or Y were accounted for within the design matrix. By including the mouse IDs into the design matrix, we say that mouse IDs were treated as fixed effects in the modelling process. An alternative method treats the mouse IDs instead as random effects, and does not include the IDs into the design matrix. We refer to this type of model as a mixed effects model, such that treatment and timepoint are included into the design matrix as fixed effects in the model, whilst mouse IDs are included as random effects. One important advantage to the limma package is that it has the ability to fit a mixed effects model, unlike edgeR or DESeq2, which can only fit fixed effects.\n\nWhy do we fit mouse IDs as a random effect rather than a fixed effect? The specific differences between mice are not of direct interest to the study, so removing them from the design matrix reduces the number of model parameters, conserves the number of degrees of freedom in modelling, and likely increases statistical power for testing. The effects, however, cannot be omitted completely because they are integral to the study design; individual mouse effects should still be accounted for when calculating relative difference between timepoints T1 and T2.\n\nTo fit a mixed effects model, let us first define our fixed effects in a design matrix. To simplify the two factors of interest, treatment and timepoint, we merge them into a single factor called group. The data which includes the new group factor are as follows:\n\n\n\nA means model is fitted to the groups by coding the design matrix as design <- model.matrix(~0+group), which gives a 4 column matrix representing the mean gene expression in each group (Figure 14). Mouse IDs are set as random effects by assigning id as the blocking variable in the lmFit function in limma. Before doing this, we first estimate the correlation between repeated mice measurements using the duplicateCorrelation function in limma as follows:\n\n\n\nA means model is fitted to the data using a design matrix that excludes the intercept term. In the plot, data points are labelled by mouse ID.\n\nThe correlation between measurements taken from the same mouse is estimated as -0.05, which is considered to be quite a small correlation value. This is expected in our example since we did not program specific mouse effects into the dataset. In the case of a negative estimated correlation, the blocking variable should be removed and we can resume the usual modelling approach of accounting for the group fixed effect only in the design matrix. In fact, the blocking variable can be removed if the estimated correlation is very, very small (say less than 0.01) as its contribution to the overall model fit would also be very minor, however, keeping it in the model would not adversely affect the modelling process either. For real use cases, correlation estimates of 0.7 to 0.9 are considered high but not uncommon. Despite our recommendation above, let us continue with the fitting of our mixed effects model for the sake of demonstrating how it can be carried out.\n\nUsing the lmFit function, we fit our random effects by setting the correlation argument to the estimated correlation and block argument to mouse id. The fixed effects modelled within the design matrix are given to the design argument, along with the expression data as follows:\n\n\n\nThe mixed effects model estimates the mean gene expression of mouse receiving treatment X at timepoint T1 to be 1.03, treatment X at timepoint T2 to be 2.12, treatment Y at timepoint T1 to be 3, and treatment Y at timepoint T1 to be 4.9. To obtain estimates for the comparisons of interest, we use a contrast matrix coded as:\n\n\n\nThe first contrast calculates the difference between timepoint T2 and T1 in treatment X, the second contrast calculates the difference between timepoint T2 and T1 in treatment Y, and the last contrast the overall difference between treatment X and Y. The overall difference between treatment X and Y is calculated after adjusting the mean gene expression at timepoint T2 by that of the baseline (timepoint T1). The contrast matrix is incorporated into the limma pipeline using the contrasts.fit function as follows:\n\n\n\nUsing the contrast matrix, the difference between timepoint T2 and T1 in treatment X is calculated as 1.09, and for treatment Y as 1.9. The overall difference between treatment X and Y is estimated at -0.82. Since the correlation of repeated mouse measurements is small, individual mouse effects have little influence on the calculation of expected gene expression values for groups. This means that the inclusion of mouse IDs as fixed or random effects do not have much practical influence in this particular example. For this reason, we observe similar results between this section and that of the previous section.\n\n\nStudies with covariates\n\nIn the remaining sections, we switch from looking at explanatory variables that are factors, and instead consider studies where the explanatory variable of interest is a covariate. Let us recall the basic models outlined in earlier sections, where a simple regression model for a covariate can be represented as a straight line defined by its y-intercept and slope. In the following section, we cover models that are more complex in their design, starting with a mix of covariates and factors. We also discuss options for non-linear fitted models that extend beyond the simple framework of y-intercept and slope. Whilst in practice the vast majority of study designs involve only factor variables, which we have covered extensively over multiple sections, this section is useful for the occasional study where the relationship between gene expression and a given covariate is of interest.\n\nIn the previous section we looked at models for the factors treatment and timepoint. We now consider a similar example, where there are two explanatory variables, the treatment factor and time as a covariate. The timepoints from the previous example is now treated as a numerical variable. Let us suppose that the timepoints T1 and T2 represent 1 hour post treatment and 2 hours post treatment, by either treatment X or Y. For simplicity, we do not consider having repeated mouse measurements here and assume that the measurements are taken from different mice each time. The data is as follows:\n\n\n\nAn appropriate design matrix that considers both the treatment factor and time covariate can be coded as model.matrix(~0+treatment+treatment:time), which gives a fitted line for each of the treatments (Figure 15). For two fitted lines with the same slope, we could use the design matrix coded as model.matrix(~0+treatment+time). Let us recall that each regression line is defined by a y-intercept and slope. In our design matrix, the first and third columns are parameterised for the y-intercept and slope of the line modelling treatment X. The second and fourth columns are parameterised for the y-intercept and slope of treatment Y.\n\nIn the plot, data points are labelled by treatment, and a fitted line is drawn for each of the treatments.\n\nThe y-intercepts for treatment X and treatment Y are estimated as -0.06 and 1.09 respectively. The y-intercepts, though, are generally not of interest. It is the slopes that are usually of interest since this quantifies the rate of increase or decrease in gene expression over time. The slope for treatment X is estimated as 1.09, and the slope for treatment Y is estimated as 1.9. In the previous section with the factors treatment and timepoint, time is consider as distinct changes in state from timepoint T1 to T2. Here, time as a covariate allows us to quantify the expected change in gene expression over an interval of time, such that over 0.5 units of time we can expect an increase of 0.54 in gene expression for treatment X (which is calculated by halving the third parameter estimate).\n\nThe fitted model can be written as E(y) = -0.06x1 + 1.09x2 + 1.09x1t + 1.9x2t, where the x1 and x2 are indicator variables for treatment X and treatment Y respectively, and t is a numerical variable representing time. Specifically, the model for treatment X can be written as E(y) = -0.06 + 1.09t. The model for treatment Y can be written as E(y) = 1.09 + 1.9t.\n\nWe now consider a mouse study over several time points. This could represent a study examining gene expression changes in a developmental stage of interest, such as early embryonic development. There are gene expression measurements from mice at times 1, 2, 3, 4, 5 and 6, each in duplicate. The times could represent hours, days or weeks, with the data as follows:\n\n\n\nNaturally, we use a design matrix coded as design <- model.matrix(~time) (Figure 16). The design matrix contains two columns and is parameterised for the y-intercept (estimated here as 4.22) and the slope of the line (estimated as -0.6). Using this model, the gene expression is expected to decrease by 0.6 units for every unit increase in time.\n\nLooking more closely at the plot of expression versus time in Figure 16, we notice that expression seems to increase between time points 1 and 2, peaks between 2 and 3, and decreases between time points 3 and 6. So we also consider another model that describes the data using a curved line, such that the curved line takes the form of y = a + bt + ct2 where y represents gene expression, t represents time and t2 is calculated as the square of time (or time to the power of 2), and a, b, and c are model parameters which we estimate. Since the model includes the time covariate to the power of 2, the curve is referred to as a second degree polynomial or quadratic model.\n\nTo fit such a model, the poly function from the stats package is used to specify the polynomial, with the specification that degree=2 for a second degree polynomial. The design matrix is coded as design <- model.matrix(~poly(time, degree=2, raw=TRUE)) (see Figure 17 where column names have been simplified). Using the design matrix, the parameters can be estimated as a = 1.2, b = 1.67, and c = -0.32. The fitted model can be written as y = 1.2 + 1.67t + -0.32t2. The quadratic model has a the peak (or trough) occurring at t = −b/(2c), which in our case is at time 2.57, and this translates to a maximum gene expression of 3.33. Using this model, we can say that gene expression is increasing up until time 2.57, and decreases after time 2.57. The quadratic model can help differentiate between the genes that are increasing in expression over time (where b > 0 and c = 0), genes that decreasing in expression over time (b < 0 and c = 0), genes that increase in expression then decrease in expression (c < 0), and genes that decrease then increase in expression (c > 0).\n\nIn our example, we specify the use of raw polynomials via the raw argument in the poly function. This allows us to easily demonstrate what the linear component of the model (t or time in the design matrix) and the quadratic component (t2 or time2 in the design matrix) looks like based on the data at hand. In practice, however, we would recommend using orthogonal polynomials rather than raw polynomials, by specifying raw=FALSE in the poly function. In fact, the default setting in poly computes orthogonal polynomials so the raw argument does not need to be specified. Raw polynomials can result in covariates that are correlated, which means that the importance of each individual effect is indistinguishable from each other. We can observe this correlation quite easily in our own example, by checking cor(poly(time, degree=2, raw=TRUE)). Orthogonal polynomials, on the other hand, ensure that the covariates are not correlated, which we can check with cor(poly(time, degree=2)), but give a messy demonstration of the parameters and model for the purpose of this article. Since the covariates are not correlated, orthogonal polynomials allows us to determine the genes where the linear term is important but the quadratic term is not, and vice versa. There can also be genes where both or neither linear and quadratic terms are important. This means that we recommend a design matrix coded as model.matrix(~poly(time, degree=2)) when fitting a quadratic model in practice, rather than the design matrix that is displayed in Figure 17. To compare the polynomial model against that of a constant term (i.e. gene expression does not change over time), an F-test can be used to test multiple parameters together (e.g. the linear and the quadratic term). In limma, we calculate F-statistics and their associated raw and adjusted p-values using the topTable function by specifying multiple columns in the coef argument, e.g. topTable(fit, coef=c(2,3), number=Inf) for such values across all genes as ranked by significance.\n\nWe extend our example with two additional time points, time 7 and 8, with the data as follows:\n\n\n\nWe know from the previous section that between times 1 and 6, a linear fit to the data shows a decreasing trend (Figure 16), and a quadratic fit to the same data shows an increasing trend before time 2.57 and a decreasing trend after time 2.57 (Figure 17). The new data points, however, do not continue with the decreasing trend at time 7 and 8. For this reason, we fit a cubic model, or a third degree polynomial, to the data so that it can capture the two changes in gene expression trends. Like before, we use the poly function from the stats package to specify the polynomial we want to fit. The design matrix is coded as design <- model.matrix(~poly(time, degree=3, raw=TRUE)) (see Figure 18 where column names have been simplified). Note that we are again using a raw polynomial in this example for simplicity of illustration of the model parameters, although we would use an orthogonal polynomial in practice.\n\nThe four parameters in the model represent the intercept term, the coefficient for time (represented as t), the coefficient for time to the power of 2 (represented as t2), and the coefficient for time to the power of 3 (represented as t3). The third degree polynomial takes the form of y = a + bt + ct2 + dt3, where y represents gene expression and t represents time, and a, b, c, and d are model parameters which we estimate. The parameters can be estimated as a = -1.85, b = 5.56, c = -1.64, and d = 0.13. The fitted model can be written as y = -1.85 + 5.56t + -1.64t2 + 0.13t3.\n\nOther than polynomial models, another choice for fitting smooth curves to the data is via regression splines using the ns function for natural splines in the splines package. Examples of this can be found in the limma and edgeR user’s guides. In general, one would fit the most complex curve that one wishes to interpret and for which the number of time points can support. Usually the complexity of curve would never exceed that of the fifth order. For example, if there is no replication at time points, one might choose a second order polynomial (degree=2) or a spline with 2 parameters (df=2) for a study on 5 time points, this would leave 2 residual degrees of freedom for the model. If there are 10 time points, then a fifth degree polynomial or a spline with 5 parameters may be appropriate, resulting in a model with 4 residual degrees of freedom. Keep in mind that in gene expression analyses, the same model is applied to every gene in the dataset, even though a simpler model may be sufficient for some genes, whilst a more complex one is needed for others.\n\nIn the case where one wants to fit smooth curves to multiple groups where samples for the groups are taken at different time points, using regression splines allow the fitted trends to be compared between the groups whilst a polynomial fit does not. Furthermore, splines tend to give more sensible and stable curves with better behaviour at the boundaries of the fit than polynomials. On the other hand, polynomials such as the quadratic fit are handy for when one wants to determine the peak or trough of the fitted curve. Note that both the fitting of a spline or polynomial is particularly useful for time course experiments with lots of time points but no replication at each time point. The time series examples in the previous sections have replication at the time points. In that case, one could treat time as a factor rather than a covariate to avoid the interpretation of curves, and to obtain differences between distinct time points explicitly. This approach could be preferred by many.\n\nWe extend our example further to include an additional two time points, time 9 and 10. It turns out that the biology under study involves genes that are turned on and off cyclically over time. This example can represent studies on circadian rhythm where certain genes are turned on in the day and turned off at night, whilst others are on in the night and off in the day. The data is as follows:\n\n\n\nA cyclical model that models the rhythmic pattern in the data is considered for this example. Although the model itself may be complex, the interpretability of the model is greater than that of higher order polynomials since the fitted cyclical pattern is repeated over time. To fit a cyclical model, we use sin and cos functions from the base package to obtain the sine and cosine trigonometric functions on the time covariate. Now, let us consider an approximate cycle length. Our cycle length appears to be roughly 6 units - the first peak occurs at time 2.57 (Figure 17), and the second peak occurs just after time 8 (Figure 18). Note that a cycle length of 24 units would be appropriate for studies on circadian rhythm if time were measured in hours. The sine and cosine functions, with cycle length of 6 units, are coded as follows:\n\n\n\nThe design matrix is then coded as:\n\n\n\nThe first column in the design matrix (Figure 19), or first parameter in the model, represents the horizontal shift of the cycling pattern from 0. The second column represents the amplitude (or height) of the sine function over time. Similarly, the third column represents the amplitude of the cosine function over time. The parameters can be estimated as 2.1, 0.53 and -1.87. The fitted model can be written as E(y)=2.1+0.53 sin(π3t)+−1.87 cos(π3t), where t represents the time covariate.\n\nThe cyclical pattern can be integrated with other models in the data. If there is an upwards trend to the overall cyclical pattern in the data (data not shown), then we can include a linear component associated with time into the design matrix. The design matrix is coded as design <- model.matrix(~time + sinphase + cosphase), and the linear component is represented in the second column of the design matrix (Figure 20). One can consider including a natural spline (using the ns function from the splines package) rather than a linear component, if the upward trend is more “curvy” and the linear trend is inadequate. For example, the natural spline can be included into the design matrix by coding as model.matrix(~ns(time, df=3) + sinphase + cosphase) (design matrix and model not shown).\n\nAn upwards trend in gene expression over time is also accounted for by including the time covariate in the design matrix.\n\n\nFurther code notes\n\nIn this article, we have shown the coding of design matrices with an intercept term in the form of model.matrix(~variable) and those without an intercept term in the form of model.matrix(~0+variable) for an explanatory variable variable. There are other ways to code for the same design matrix, such as model.matrix(~1+variable) for a design matrix with an intercept term, and model.matrix(~-1+variable) or model.matrix(~variable-1) for a design matrix without an intercept term.\n\nThe makeContrasts function for creating contrast matrices ensures that the contrast matrix is ordered correctly according to model parameters in an associated design matrix. It also returns an error message if there is a mismatch with the associated design matrix which is helpful. However, the makeContrast functions requires one to type full column names from the design matrix which can be tedious. The function also complains (in the form of a warning message) about column names from the design matrix that are syntactically invalid.\n\nAlternatively, one can create a contrast matrix manually by using the cbind function as follows:\n\n\n\nThe above contrast matrix contains three contrasts, which are linear combinations of four model parameters (A, B, C and D). The first contrast compares A to C, the second compares B to C, and the last contrast compares the average of A and B to the average of C and D. When coding for contrast matrices manually, one should carefully check that the rows in the contrast matrix match the columns of the design matrix.\n\nStarting with a counts table, a complete workflow for differential gene expression analysis of RNA-seq data using the limma package can be found in the “RNA-seq analysis is easy as 1-2-3 with limma, Glimma and edgeR” workflow article5. A summary of the main steps for fitting a linear model to each gene and obtaining parameter estimates are as follows:\n\n\n\nThe above code is non-runnable as the counts object is missing. The counts object here is assumed to be a table of counts, with rows as genes and columns as samples. In this example, there are six samples, three of which belong to group 1 and the other three to group 2. The design matrix is parameterised for a means model, and the contrast matrix is used to calculate the difference in mean expression between group 1 and group 2.\n\n\nTheory of linear regression models, design matrices and contrast matrices\n\nThis section briefly summarises the mathematics of design and contrasts matrices. This section is optional and understanding it is not required to undertake any of the analyses described earlier in the article. It is provided as a reference for those comfortable with the mathematical notation.\n\nA regression model, in the general sense, can be used to describe the relationship between explanatory variables and gene expression, the response variable. There are many different types of regression models, where each assume different characteristics for the relationship between explanatory and response variables, as well as the properties associated with variability in the data. Consider one such model, a simple linear regression model\n\nThrough an estimation process, we obtain parameter estimates which we denote as β^. Using the parameter estimates, we can calculate fitted values for gene expression by multiplying the design matrix X by the parameter estimates, such that the fitted values are calculated as y^=Xβ^. The least squares estimation strategy obtains estimates of regression parameters by minimising the sum of squared residuals, where residuals are calculated as the difference between observed y and fitted gene expression values y^.\n\nThe simple linear regression model, E(y) = β0 + β1x, can be generalised and extended to contain more explanatory variables and written as\n\nContrasts are set up to examine relationships of interest, such that a contrast matrix C contains K column vectors of length p (number of model parameters), and can be written as\n\n\nDiscussion\n\nIn this article, we described the set up and interpretation of design matrices and their associated models for a variety of case studies involving factors and/or covariates. The examples in this article are completely reproducible via our Rmarkdown document that can be downloaded from the RNAseq123 workflow package available from https://doi.org/doi:10.18129/B9.bioc.RNAseq123. The document can be used to recreate design matrices and plots found in this article, as well as to obtain estimated values for model parameters.\n\nThe estimation process was not described explicitly in our work since it is not of direct interest here. Parameter estimates were merely used to illustrate aspects of the model relating to the design matrix. For simplicity, parameter estimation in our single gene examples was carried out using the lm function from the stats package, with the exception of the mixed effects model that uses lmFit from limma due to its complexity. In practice, limma’s lmFit function would be used to obtain parameter estimates for multiple genes simultaneously in RNA-seq datasets and other genomic data types. The estimation process performed by lm and lmFit can be different (lm carries out ordinary least squares estimation, whereas lmFit usually carries out weighted least squares estimation), so their parameter estimates may differ also. The two functions would produce the same parameter estimates if lmFit was run in its simplest form without intrablock correlations, precision weights or robustification.\n\nOur article describes case studies that are common to the field of genomics research, where the choice of language used throughout the article makes it easily adaptable to studies and datasets for various applications. We have taken special care to explain, where appropriate, the reasoning behind specific design matrix set ups and describe how one would go about interpreting the associated model and its parameters. This allows readers to build their knowledge and understanding of simpler concepts, and work their way up to more advanced concepts, such as mixed effects or cyclical models, that are also described. Although we have covered design matrices in many common experimental settings, there will certainly be more complex scenarios that have been missed. We do not describe, for example, a study with a covariate and multiple factors, however, a reader should be able to tackle such an example quite easily with their understanding from the sections on multiple factors, combined with their understanding from the studies on covariates. For more complicated experimental designs, we would advise readers to consult with a statistician or bioinformatician who is experienced in linear modelling.\n\n\nSoftware availability\n\nThis article was written using Bioconductor10 version 3.12, running on R version 4.0.3 (2020-10-10). The examples in this article made use of the software packages limma version 3.45.19 and TeachingDemos11 version 2.12. This article was written as an Rmarkdown document that was compiled using knitr, and converted from an Rmarkdown document to LaTex with the help of BiocWorkflowTools version 1.15.0. All packages and their version numbers are shown below. Reproducible code for this article is available at https://doi.org/doi:10.18129/B9.bioc.RNAseq123 (Artistic License 2.0), which also stores the code for the ”RNA-seq analysis is easy as 1-2-3 with limma, Glimma and edgeR” workflow article5.\n\n",
"appendix": "Acknowledgments\n\nThe authors would like to thank Tim Thomas, Quentin Gouil, Shani Amarasinghe and Mengbo Li for contributing suggestions that have improved the quality and clarity of the work presented in this article.\n\n\nReferences\n\nSmyth GK: Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Stat. Appl. Genet. Mol. Biol. 2004; 3(1): Article 3. PubMed Abstract | Publisher Full Text\n\nSmyth GK: Limma: linear models for microarray data. In: Gentleman R, Carey V, Dudoit S, editors. Bioinformatics and computational biology solutions using R and Bioconductor. New York: Springer; 2005. p. 397–420.\n\nGlonek GFV, Solomon PJ: Factorial and time course designs for cDNA microarray experiments. Biostatistics. 2004; 5(1): 89–111. Publisher Full Text\n\nRitchie ME, Phipson B, Wu D, et al.: limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015; 43(7): e47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLaw CW, Alhamdoosh M, Su S, et al.: RNA-seq analysis is easy as 1-2-3 with limma, glimma and edgeR. F1000 Research. 2016; 5(1408). PubMed Abstract | Publisher Full Text | Free Full Text\n\nVenables B: Coding matrices, contrast matrices and linear models. 2018; CRAN codingMatrices package vignette.\n\nSoneson C, Marini F, Geier F, et al.: ExploreModelMatrix: interactive exploration for improved understanding of design matrices and linear models in R. F1000 Research. 2020; 9(512). PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobinson MD, McCarthy DJ, Smyth GK: edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010; 26: 139–140. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcCarthy DJ, Chen Y, Smyth GK: Differential expression analysis of multifactor RNA-Seq experiments with respect to biological variation. Nucleic Acids Res. 2012; 40: 4288–4297. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuber W, Carey VJ, Gentleman R, et al.: Orchestrating high-throughput genomic analysis with Bioconductor. Nat. Methods. 2015; 12(2): 115–21Reference Source. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSnow G: TeachingDemos: Demonstrations for Teaching and Learning. 2020. Reference Source R package version 2.12."
}
|
[
{
"id": "76094",
"date": "05 Feb 2021",
"name": "Federico Marini",
"expertise": [
"Reviewer Expertise Bioinformatics",
"RNA-seq data analysis"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript presents a detailed description of an essential aspect of the differential expression analysis workflows, applicable to a variety of high-throughput assays, namely the creation, specification, and comprehension of proper design matrices, followed by the correct setup for the relevant comparisons, to be done via contrast matrices.\nWhile technologies such as RNA-seq are widespread in the current literature, and an extensive collection of software packages for such datasets are nowadays well established, the appropriate choices in statistical modeling might remain a more elusive topic to many researchers with experience in genomic data analysis, but not formally trained in biostatistics/bioinformatics.\nSeeing over the years the amount of questions on specialized support forums (the Bioconductor Support Site, Stack Overflow, Biostars) about the topics addressed in this work, I can only say it is a very welcome and valuable contribution to the scientific literature, and additionally become an essential resource to be used for teaching (e.g. in lectures but also in more focused workshops, as it is written in a way that it integrates seamlessly with the widely used framework of limma).\nOverall, the manuscript is clearly written, avoiding overcomplicated jargon to make it understandable to a large readership, and well structured in a way that builds up from the fundamental elements of commonly encountered experimental designs, followed by more complex yet plausible scenarios, where more refined approaches might be required to properly address the underlying scientific questions.\nI have a few points that a revised version of the manuscript might address to improve the overall quality of this work:\nThe text could refer to Figure 1 subpanels explicitly, for more clarity. Moreover, in the center and right panel the description and legend both refer to Beta1 and Beta2, while the figure shows Beta0 and Beta1.\n\nSome formal definitions in the Terminology table can be improved with more complete wordings to provide the clearest, non-ambiguous, yet formally correct descriptions of the fundamental concepts. Response variable: add something like \"often specified/referred to as Y in this manuscript\". Explanatory variable: add \"often noted as X_i\" or similar. Nested factors: does this definition need e.g. the addition of \"mutually exclusively\" for better clarity? Mixed effect models: \"where random effects are usually not of interest to the study at hand.\", probably add something like \"yet need to be specified for proper modelling of the data at hand\"? Related to this, a similar formulation could be adopted at page 19 when defining the random effect as \"not of interest to the study\" - while this is truly not of interest, the omission of accounting for that could be detrimental in the modeling. Overall, I think that if the definitions in here would be a little more complete, it would increase a lot the value of this table as a go-to reference for many occasions - of course, without overcomplicating too much by overloading it with theoretical details.\n\nAs a general comment - for the table and the main text: I am not so sure that the \"covariate\" refers only to the continuous variables; I often encountered the nomenclature of a \"categorical covariate\". I suggest it is worth revisiting this aspect and if needed clarify it further in the scope of this manuscript.\n\nIn the subsection \"Studies with covariates\", the cyclical models are not included in the table just below the text.\n\nIt could be helpful to have a visual representation for the \"Overlap of treatment-control results\" with something like a Venn Diagram or an Upset plot to provide an immediate summary of the textual description.\n\nWhile I understand that it could be out of scope for this manuscript, focused on the limma framework, it might be worth mentioning that some conventions might differ in other frameworks. For example, DESeq2 specifies the main variable of interest as the last term in the design, while on page 22 it is stated \"This is why we place the factor of interest first as it simplifies the subsequent code for the comparisons of interest, even though a different order of factors added give equivalent models with variations in parameterisation.\". The authors could mention that different software might follow different default implementation and recommend to accurately read the software documentation.\n\nOn page 36, the term delta_k was not previously defined (but was previously used for the interaction term). Does this need to be clarified explicitly?\n\nIt might concern the pagination of the article, but sometimes it is not entirely clear where the output of the code is close to a figure which actually relates to a section above (e.g. on page 10 with Figure 4). Another example where the sequence of text-output-figures is not so clear is on page 33 (Model belongs to above section, Matrix + Plot to the one below).\n\nThe authors could provide a snapshot of the RNAseq123 repository (e.g. on Zenodo) to guarantee the complete reproducibility of the manuscript - often times vignettes are edited over the development cycles. Even if not entirely a \"Software\" article, it is inserted into a workflow software package in Bioconductor.\n\nSome sentences are missing some prepositions or are not fully correct from the grammatical point of view. I am reporting here some other sentences as well that would need to be fixed:\n\nAbstract: \"set up *of* an appropriate model via design matrices\", \"Differential expression analysis [...] use*s* linear models\".\n\npage 4, \"where the line is defined by Beta0 the y-intercept and Beta1 the slope\" should have some commas inserted.\n\n\"classifiers associated with samples\", the wording is not so clear.\n\npage 8, \"if variable is a covariate the then\", invert the and then.\n\nLegend of Figure 3, the backticks for rendering the code monospace font are still visible.\n\npage 12: refers to the `makeContrast` function, but should read `makeContrasts` (also page 17, please check all occurrences).\n\npage 23: \"the use of unsupervised clustering plots\" could be rephrased in a more general way as \"the use of exploratory data analysis techniques\". In the sentences after that: \"with one addition*al* factor to observe\", please correct this.\n\npage 28: \"time is consider*ed* as distinct changes in state\".\n\npage 35: Section \"Theory of linear regression models, design matrices and contrast matrices\" might be missing an oxford comma.\n\npage 36, \"A model with an intercept term, simply sets\", please remove the comma.\n\npage 37, \"R markdown\" should read \"R Markdown\" (at least it does so in most of RStudio's documentation)\n\nMost of these points are easily addressable, and I believe they would contribute in making even clearer the concepts already nicely described throughout this work, which I envision to be extremely useful for many practitioners that have to deal with (and properly perform) omics data analysis.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "76092",
"date": "09 Feb 2021",
"name": "Jenny Drnevich",
"expertise": [
"Reviewer Expertise Gene expression analysis",
"bioinformatics",
"RNA-Seq data"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nLaw et al.’s Method Article focuses on two crucial aspects of differential expression analysis: how to properly form the statistical model and how to pull appropriate contrasts from the model. Unlike other aspects involved with a full analysis that are generalizable to all experiments (e.g., sample quality control, normalization, filtering, visualization of results and down-stream data mining), the statistical modeling must be tailored for each experimental design and hence why it often is the most difficult part for beginners and experienced analysts alike. Technical mathematical notation is confined to one “optional” section, so the rest of the manuscript is more understandable to people from varied backgrounds that are looking for help with making correct design and contrast matrices. Although the manuscript is more understandable, it still contains such necessary depth of information spanning simple to very complex models that it will require many readings to fully digest all of it. I envision this to become a welcome guide to be referenced again and again. This manuscript distinguishes itself from other guides on selecting design and contrast matrices like the limmaUsersGuide() and edgeRUsersGuide(), which are excellent but also contain a tome of information on other aspects of gene expression analysis such that the design/contrast matrix is not as easy to find as this stand-alone article. I also really like the combination in each figure of the fitted model statements (e.g., E(y) = 1.85 + 0.54x), the actual design matrix and a plot visualizing the data points and estimated parameter values. The fitted model statements are an addition that I have not seen much in other references on making design matrices. I do have a few minor points below that should be addressed:\nWhy are all the makeContrasts() functions called with levels=colnames(design) instead of just levels=design? It seems the second way is not only simpler but also reinforces the idea that the contrast matrix must be defined from the design matrix. I am not sure I have seen the first way in any other documentation in making contrast matrices, so it seems like an explicit change but without a compelling rationale.\n\nAt the end of the “Contrast matrix for computing differences “ section, the sentence “gene expression of healthy mice is greater than sick mice by a value of −1.62” required several readings to understand what it was trying to say, probably because of the conflict between “greater than” and “-1.62”. I would suggest replacing “greater than” in this sentence and the one above with “different than”.\n\nIn the “2 versus 2 group comparisons” section, I question the choice to compare control and treatment III against treatment I and II given the rationale “This may be of interest if there are prior expectations that two groups are more similar to each other than the other two”. Given that we have already seen that treatment III has the biggest change from the control, I feel like readers will be confused as to why control and treatment III were paired together instead of control and treatment I.\n\nI am somewhat disappointed not to see the “sum to zero” parametrization for a two-factor model that is in the limmaUsersGuide(). Upon reflection, I agree with the decision to not add yet another model type to this already complex guide which does contain example contrasts to get the traditional main effects contrasts.\n\nWould unequal sample sizes in groups affect the results of the modeling in any way? Perhaps this only applies to the estimate of the grand mean in the “sum to zero” parametrization, but I have occasionally struggled with getting the same answers from different model parametrizations and on-line help has suggested it was due to uneven sample sizes in the groups. If uneven sample sizes would not affect the choice of parametrization or formation of contrasts, it would be nice to have a simple sentence of that somewhere. If there are situations where uneven sample sizes would be an issue but it is beyond the scope of this paper, that should be addressed briefly and outside references given.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1444
|
https://f1000research.com/articles/9-1441/v1
|
10 Dec 20
|
{
"type": "Method Article",
"title": "A simple adeno-associated virus-based approach for the generation of cardiac genetic models in rats",
"authors": [
"Michal Schlesinger-Laufer",
"Guy Douvdevany",
"Lilac Haimovich-Caspi",
"Yaniv Zohar",
"Rona Shofty",
"Izhak Kehat",
"Michal Schlesinger-Laufer",
"Guy Douvdevany",
"Lilac Haimovich-Caspi",
"Yaniv Zohar",
"Rona Shofty"
],
"abstract": "Background: Heart failure is a major health problem and progress in this field relies on better understanding of the mechanisms and development of novel therapeutics using animal models. The rat may be preferable to the mouse as a cardiovascular disease model due to its closer physiology to humans and due to its large size that facilitates surgical and monitoring procedures. However, unlike the mouse, genetic manipulation of the rat genome is challenging. Methods: Here we developed a simple, refined, and robust cardiac-specific rat transgenic model based on an adeno-associated virus (AAV) 9 containing a cardiac troponin T promoter. This model uses a single intraperitoneal injection of AAV and does not require special expertise or equipment. Results: We characterize the AAV dose required to achieve a high cardiac specific level of expression of a transgene in the rat heart using a single intraperitoneal injection to neonates. We show that at this AAV dose GFP expression does not result in hypertrophy, a change in cardiac function or other evidence for toxicity. Conclusions: The model shown here allows easy and fast transgenic based disease modeling of cardiovascular disease in the rat heart, and can also potentially be expanded to deliver Cas9 and gRNAs or to deliver small hairpin (sh)RNAs to also achieve gene knockouts and knockdown in the rat heart.",
"keywords": [
"Cardiac research",
"animal models",
"Adeno-associated virus"
],
"content": "Abbreviations\n\nAdeno-Associated Virus (AAV); Viral genomes (vg); Short hairpin RNAs (shRNA); Fractional shortening (FS%); left ventricular end diastolic dimeter (LVIDd); enhanced green fluorescent protein (eGFP); cardiac troponin T (cTnT)\n\n\nIntroduction\n\nCardiovascular disease is the leading global cause of mortality and is expected to account for >22.2 million deaths by 20301. Despite some progress in the medical treatment of heart failure, the 5-year mortality is still dismal, at about 50%, a worse prognosis than in most cancers. With the continued aging of the population, an increase in the number of new patients is estimated. Unless there is substantial progress in prevention and treatment, heart failure will remain a major health problem2.\n\nContinuous progress in therapy relies on a better understanding of the pathobiology of heart failure, and studies to discover new mechanisms and test new therapeutics are direly needed3. Animal modeling of cardiovascular disease is challenging, and most models aim to create the etiological factors leading to cardiac disease either by surgery, selective breeding, or genetic modifications. Small animal models, particularly mice and rats, are essential models in cardiac research allowing for relatively rapid, high through-put, and cost effective means of studying cardiac physiology, disease, and novel therapeutic targets4.\n\nThe rat has several advantages over the mouse for cardiovascular research. It offers some of the advantages of a larger animal but with reduced costs, and it is preferable for surgical procedures. Technically it is more feasible to create ischemia, pressure, or volume overload models by coronary artery ligation, aortic banding, or shunt procedures respectively in the rat than in the mouse5, and those models are well established in the literature6. It is easier to perform physiological monitoring in the rat, and in many cases, the physiology is more similar to humans7. Rats also have a greater ability to increase their heart rate during exercise, have a more positive FFR (force-frequency relationship) and have slightly slower kinetics of contraction and relaxation as compared to mice8. Indeed, the rat became the initial small animal model of choice for cardiovascular research9.\n\nFunctional genomics aimed at studying the effects of changes in gene expressions by targeted mutagenesis or transgenesis allowed many insights into signaling pathways involved in the pathogenesis of heart failure10. With the sequencing of the mouse genome and the development of many genetically modified mouse lines in the last 20 years, the mouse has become widely used as a tool in generating complex myocardial phenotypes11. However, functional genomic research in the rat has been limited due to difficulties in manipulation of the rats genome7. As a result, despite being the most remote rodent from humans in terms of contractile function, the mouse became the most used animal model for cardiovascular research.\n\nHere we aimed to develop a simplified cardiac-specific rat transgenic model based on a single adeno-associated virus (AAV) injection. We show that we can achieve a robust high level and specific cardiac expression of transgene in the rat heart. This model will allow easy and fast transgenic based disease modeling of cardiovascular disease in the rat heart.\n\n\nMethods\n\nA total of 15 HsdHan: Wistar rat pups were used in this study. Sample size of n=5 per group was calculated based on preliminary mice experiments with 80% power to detect 1.5-fold increase in mean GFP fluorescence. All pups were born at the SPF unit of the pre-clinical research authority at the Technion (Israel Institute of Technology; IIT). Health monitoring was carried out in accordance with FELASA recommendations12. Each litter was housed with the dam, weened at 3 weeks and separated by sex. Rats were group housed in IVC racks in Sealsafe Plus GR900 TECNIPLAST cages, on Sani chips bedding (Teklad ENVIGO) and Tek-fresh bedding (Teklad ENVIGO). Disposable play tunnels were added as environmental enrichment. Rats were maintained under climate-controlled conditions of 12:12 hrs light/dark cycle, temperature range 21±2°C, a relative humidity of 30–70% and fed ad libitum a commercial food – pellet diet (Altromin 1414 IRR). Ad libitum reverse osmosis acidified water (pH of 3.0±0.2) was accessible in polycarbonate water bottles covered with stainless steel lids. All efforts were made to ameliorate any suffering of animals, and treatment consisted of a single intraperitoneal injection, anesthesia during echocardiography, and euthanasia at the end of protocol. Daily monitoring by a veterinarian ensured animal well-being.\n\nStudies were conducted at the Technion (IIT), Faculty of Medicine, Haifa, Israel, after obtaining approval from the institute’s IACUC. All proceedings complied with the Animal Welfare Act of 1966 (P.L. 89-544), as amended by the Animal Welfare Act of 1970 (P.L.91-579) and 1976 (P.L. 94-279). Animals allocation to control and experimental groups was done randomly. Cages were arranged on the racks by a technician unaware of the experimental plan. Cages were then allocated to the control, ‘low dose’, and ‘high dose’ groups according to their order on the rack without prior examination of the rats in the cages to avoid bias. Viral injection, echocardiography, euthanasia, and sample analysis were each done on all the animals in the same day to minimize confounders. Echocardiography and histology (H&E) analyses were performed by an investigator unaware of the group allocation of the animals. Because the GFP fluorescence was obvious, the investigator performing the fluorescence microscopy could not be blinded to the treatment. To avoid bias in the GFP fluorescence analysis, we performed and show analysis of the entire heart sections. High power fields were picked at random without looking at the green fluorescence.\n\nThe following plasmids were used for AAV production: pENN.AAV.cTNT.PI.eGFP.WPRE.rBG was a gift from James M. Wilson (Addgene plasmid #105543; RRID:Addgene_105543); pAdDeltaF6 was a gift from James M. Wilson (Addgene plasmid #112867; RRID:Addgene_112867); pAAV2/9n was a gift from James M. Wilson (Addgene plasmid #112865; RRID:Addgene_112865).\n\nAAV was produced as previously described in detail13. In brief, ten 150 mm dishes of HEK293T cells (ATCC) were triple transfected using polyethylenimine (PEI). Media containing virus was collected after 72 hours and combined with the media and lysate that were collected after 120 hours. Both lysate and media were purified over iodixanol gradient (Optiprep, Sigma D1556-250ML) via ultracentrifugation. Buffer was exchanged to PBS via amicons (Millipore, UFC910008). AAV titration was conducted to determine viral particle load by qPCR with AAV transfer plasmid as a positive control to create a standard curve using WPRE primers.\n\nWe used two viral stock concentrations - a ‘low dose’ of 4×1012 viral genomes (vg)/ml, and a ‘high dose’ of 2.6×1013 viral genomes (vg)/ml concentrations. Five-day-old rat pups (n=5 pups per group) received a single 50 µl intra-peritoneal injection of saline (control group), 2×1011 viral genomes (low dose), or 1.3×1012 viral genomes (high dose) using a 30-gauge needle. Injections were performed in all the animals in the morning at the animal facility. Rats were maintained in house and analyzed at 12 weeks of age.\n\nAfter euthanasia the animal weight was measured on a laboratory scale (Precisa BJ 610C). The heart was harvested, washed with cold PBS, and blotted on Kimwipe tissue paper, and the weight measured on a laboratory scale (Precisa XT 220A).\n\nAt 12 weeks of age, rats were anesthetized with 2% Isoflurane; body temperature was maintained by placing the rats on a warm 40°C heating plate. Breathing and heart rate were monitored throughout the procedure. Echocardiography was performed using a High-Resolution Ultrasound Imaging system Vevo2100 (Visual Sonics, Fujifilm) using a MS 250 13–24 MHz linear array transducer. Measurements were performed on parasternal short axis view at the level of the papillary muscles using M-Mode. Fractional shortening (FS%) was calculated as follows: FS (%) = [(LVIDd − LVIDs) /LVIDd] × 100. All values were based on the median of 3 independent measurements for each rat.\n\nHearts were fixed in ice cold 4% formaldehyde in PBS for 2 hours, then placed in cryopreservation solution containing 30% sucrose in PBS at 4°C overnight. The next day hearts were washed in cold PBS, embedded in optimal cutting temperature compound and snap frozen in 2-methylbutane immersed in liquid nitrogen. Sectioning of the hearts was performed in a cryostat (Leica) at 5 µm intervals. Hematoxylin and eosin (H&E) staining was performed using standard protocols and imaged with 3DHistech Pannoramic 250 Flash III automatic slide scanner. For immunofluorescence, cardiac sections were permeabilized with 1% Triton and blocked with a solution containing 5% bovine serum. Primary antibody monoclonal Anti-sarcomeric-alpha-Actinin (catalog number A7811, clone EA-53, Sigma-Aldrich) was incubated overnight at 4°C. Secondary antibody (Jackson ImmunoResearch, catalog number 715-175-151) was incubated for 1 hour at room temperature. Nuclei were counterstained with DAPI for 10 min at room temperature. Slides were imaged with Axio Observer inverted fluorescent microscope (Zeiss) using an X-cite metal-halide light source and a high-resolution camera (Hamamatsu Orca R2) and with 3DHistech Pannoramic 250 Flash III slide scanner.\n\nRNA was purified from apical segments of hearts using TRI-Reagent (Sigma-Aldrich), according to the manufacturer's protocol. RNA was then reverse transcribed with 5x All-In-One Reverse Transcriptase MasterMix (Applied Biological Materials, Inc). Quantitative real‐time PCR was performed with iTaq universal SYBR green supermix (Bio‐Rad) using Bio‐Rad CFX96 real‐time system (model C100 Touch). Cycling conditions were: step 1 - 95°C for 3 minutes, step 2 - 95°C for 10 sec, step 3 - 55°C for 30 sec and read plate. Steps 2–3 were repeated 39 times. Expression data were normalized to the expression of Gapdh and ribosomal protein L4 (Rpl4). For each reaction we used technical duplicates, no-RT, and negative controls. The primers used for qRT-PCR are shown in Table 1.\n\nA two-tailed Student t-test was used to compare each experimental group with the control group.\n\n\nResults\n\nRecombinant AAVs are the leading platform for in vivo delivery of gene therapies. To test the ability to transduce the rat heart with a single intraperitoneal AAV injection, we generated AAV 2/9 vectors, encoding for the green fluorescent protein eGFP under the control of the cardiac troponin T (cTnT) promoter, referred to as AAV9-cTnT-eGFP. Five -day-old rats were allocated to receive a single intraperitoneal control, ‘low dose’, or ‘high dose’ virus injection.\n\nOne of the most promising feature of AAV as a gene therapy vector is its low toxicity14. To verify that the viral transduction did not result in cardiac toxicity, we performed a gravimetric analysis of the injected rats. Bodyweight analysis did not show any significant change in either the low or high dose virus injected groups, as compared to control saline injected rats (Figure 1A). Similarly, the heart weight, normalized to body weight, did not significantly change following viral transduction, indicating no significant cardiac atrophy or hypertrophy (Figure 1B). To assess structural damage to the heart, necrosis, or signs of inflammation, we performed a histological analysis of the hearts with H&E staining. As shown in the representative images of control and high doses injected rat hearts (Figure 1C), viral transduction did not result in area of necrosis or in inflammatory cell infiltrate in the heart.\n\nA. Gravimetric analysis show no difference in body weight between rats injected with ‘high dose’ or ‘low dose’ AAV9-cTnT-eGFP and control saline injected rats. Black bars show average and each dot represent a measurement from one rat. N.S = not statistically significant from control by Student t-test. B. Gravimetric analysis shows no difference in heart weight normalized to body weight between rats injected with ‘high dose’ or ‘low dose’ AAV9-cTnT-eGFP and control saline injected rats. Black bars show average and each dot represent a measurement from one rat. N.S = not statistically significant from control by Student t-test. C. Representative hematoxylin & eosin stained cardiac sections at low (top) and high (bottom) magnification, showing normal histology with no foci of necrosis or inflammatory cell infiltrates even in rats injected with ‘high dose’ AAV9-cTnT-eGFP (right). Control Saline injected rat heart is shown on the left. Bar = 100 µm.\n\nTo assess any function impairment, we performed two-dimensional echocardiography as well as M-mode measurements in all the rats. This echocardiographic analysis showed that neither the low nor the high dose injected rats had cardiac dimensions and contractile function that significantly differed from the control saline injected rats. Specifically, the left ventricular end diastolic dimeter (LVIDd) and the fractional shortening percent (FS%) were not-significantly changed in the low or high dose injected groups, as compared with the control saline injected rats (Figure 2).\n\nA. Representative echocardiographic images in two-dimensional short axis view (top) and M-mode (bottom) magnification, showing normal cardiac dimensions, wall thickness and contractile function even in rats injected with ‘high dose’ AAV9-cTnT-eGFP (right). Control saline injected rat heart is shown on the left. B. Echocardiographic measurements of left ventricular internal diameter in end diastole (LVIDd) showing no significant changes between rats injected with ‘high dose’ or ‘low dose’ AAV9-cTnT-eGFP and control saline injected rats. Black bars show average and each dot represent measurement from one rat. N.S = not statistically significant from control by Student t-test. C. Echocardiographic measurements of left ventricular fractional shortening (FS%) showing no decrement in cardiac function between rats injected with ‘high dose’ or ‘low dose’ AAV9-cTnT-eGFP and control saline injected rats. Black bars show average and each dot represent measurement from one rat. N.S = not statistically significant from control by Student t test.\n\nTogether these data show that cardiac transduction with a single intraperitoneal injection of AAV9 in a dose of up to 1.3×1012 viral genomes does not result in any significant cardiotoxicity, cardiac atrophy, cardiac hypertrophy, or functional impairment. This approach may, therefore, be useful for cardiac studies in the rat.\n\nAAVs can achieve high transduction efficiency in vivo. To assess the efficiency of our simplified approach, we analyzed cardiac sections from the control and transduced rats for eGFP green fluorescence. As can be seen in the images (Figure 3A–F) the transduction with the low dose of AAV9-cTnT-eGFP resulted in only a low number of GFP positive cells in the heart. In contrast, transduction with the high AAV9-cTnT-eGFP dose resulted in robust and high GFP expression in the entire heart, including both the right and left ventricles in all animals, as compared with control, saline injected rat hearts, that showed no GFP green fluorescent signal.\n\nA–C. Low magnification images of cardiac sections of rats with GFP fluorescence (green) and DAPI nuclear stain (blue) showing no green GFP fluorescence in control saline injected rats (A), few transduced GFP positive cardiomyocytes in the low dose AAV9-cTnT-eGFP injected rats (B), and robust high GFP expression in the high dose AAV9-cTnT-eGFP injected rats (C). Each image was taken from a different rat. Bar = 5 mm. D–F. High magnification images from the same rats shown in A–C, showing no green GFP fluorescence in control saline injected rats (D), few transduced GFP positive cardiomyocytes in the low dose AAV9-cTnT-eGFP injected rats (E), and robust high GFP expression in the high dose AAV9-cTnT-eGFP injected rats (F). Each image was taken from a different rat. Bar = 200 µm.\n\nTo ensure that the bright green fluorescent signal was originating from the transduced cardiomyocytes in the heart, we performed a higher magnification analysis coupled with sarcomeric α actinin fluorescent immunostaining, to label the cardiomyocytes. As shown in the representative images (Figure 4A–D), transduction with the high dose of AAV9-cTnT-eGFP resulted in green GFP fluorescence in almost all the cardiomyocytes. Importantly non-cardiomyocytes cells in the heart were not labeled by eGFP (Figure 4D).\n\nA–B. Representative images of cardiac sections of control rats (A – low magnification, Bar= 100 µm; B- High magnification Bar= 20 µm). C–D. Representative images of cardiac sections of ‘high dose’ AAV injected rats (C – low magnification, Bar= 100 µm; D- High magnification Bar= 20 µm). For all panels (A–D) staining with sarcomeric α actinin immunostaining (red), eGFP fluorescence (green) and DAPI nuclear stain (blue). Left column of panels shows a composite of actinin (red) and DAPI (blue) channels, middle column shows a composite of GFP (green) and DAPI (blue) channels, and right column of panels shows a composite of all three actinin (red), GFP (green) and DAPI (blue) channels. No green GFP fluorescence is seen in cardiomyocytes from control saline injected rats (A–B), and ~100% of cardiomyocytes show green GFP fluorescence in the high dose AAV9-cTnT-eGFP injected rats (C–D). Importantly, non-cardiomyocyte cells, likely cardiac fibroblasts, identified by the lack of sarcomeric α actinin immunostaining do not show GFP fluorescence even in the high dose AAV9-cTnT-eGFP injected rats, demonstrating the expression is restricted to cardiomyocytes (D, arrows). Bar = 20 µm.\n\nFinally, we quantified the transduction efficiency. We measured the mean GFP signal in individual cardiomyocytes chosen from random high power fields (Figure 5A). This analysis showed a cardiomyocyte mean ± standard deviation GFP intensity of 4.24±0.71, 6.14±5.48, and 34.53±16.15 in arbitrary units in the control, low dose, and high dose injected hearts respectively (N=3 rats, n=~900 cardiomyocytes, in each group). Using a value of mean + three standard deviations of the signal intensity in the control cardiomyocytes as the cutoff for GFP expression, this would be translated to transduction efficiency (mean ± standard deviation) of 0.25± 0.35%, 20.75±7.53%, and 99.66±0.28% of the cardiomyocytes in the control, low dose, and high dose injected hearts respectively. Next, we quantified GFP mRNA expression levels using quantitative reverse transcriptase polymerase chain reaction (qRT-PCR). This analysis showed that cardiac GFP expression achieved with the low dose AAV9-cTnT-eGFP injection was 1.47±0.68 fold higher than that of the background (saline injected animals), while the expression achieved with the high dose AAV9-cTnT-eGFP injection was 32.6±11.5 higher (Figure 5B). AAV9 is known to efficiently transduce the heart but has a general distribution of expression throughout the body, most notably the liver15. To direct the expression specifically to the cardiomyocytes in the heart, we used the cardiac troponin T promoter in our viral vectors, because of its well-documented ability to drive strong, cardiomyocyte-selective transgene expression. We therefore also quantified the expression level in two additional tissues, the kidney and liver. The qRT-PCR showed that GFP expression in the kidney or liver was undetected, even in the high dose AAV9-cTnT-eGFP injected rats (Figure 5B).\n\nA. A violin scatter plot of mean cardiomyocyte GFP intensity measured from random high power magnification of cardiac sections. Each point represents measurements from one cardiomyocytes in the control, low dose, and high dose AAV9-cTnT-eGFP injected rats, black bars show the mean intensity (N=3 rats, n=~900 cardiomyocytes, in each group). B. Quantitative reverse transcriptase PCR of normalized GFP expression showing very modest expression in the low dose AAV9-cTnT-eGFP rat hearts and high expression in the high dose AAV9-cTnT-eGFP rat hearts. No GFP expression was detectable in the liver or kidney even in the high dose AAV9-cTnT-eGFP rats. N=5 rats. Bars show the average ± standard error. * Student t-test vs. control p<0.05.\n\n\nDiscussion\n\nSome of its special features make AAV the preferred in vivo gene transfer vector. It is not associated with human or rat disease, it has a wide and promiscuous tropism, it is minimally immunogenic, and has a long-lived and efficient gene transfer ability15. There are several serotypes of AAV that show different tropism to target tissues. AAV9 is the most cardiotropic serotype in the mouse and rat, and provides high level and stable expression in the heart16. We showed here that combining the cardiotropic feature of AAV9 with the cardiac specific activity of the cTnT promoter resulted in a high but also specific expression in cardiomyocytes. We also showed that even in the high dose group we did not see expression in non-cardiomyocytes in the heart, there were no expression in the liver, and no signs of cardiotoxicity, inflammation, or functional impairment. Therefore, our approach is both safe and efficacious, and enables a scalable expression of a transgene in the adult rat heart.\n\nThere are several delivery methods for cardiac gene transfer: one method which has been described in mouse and rat is the direct intramyocardial injection, an invasive procedure that includes left thoracotomy surgery, and requires high skills16,17. Another common delivery method is intracoronary delivery via aortic root injection. This also requires invasive surgery and the use of potentially harmful vasodilators6. The intravenous injection was described in mouse models5,18,19, but in the rat, this approach resulted in low cardiac transgene expression20. The efficiency of intra-venous delivery was shown to be increased by using ultrasound-targeted microbubble destruction, but this approach required continuous viral infusion through a centrally placed venous catheter and appropriate ultrasound equipment20. Compared with these methods an intraperitoneal injection is simple, does not require special expertise or equipment, is safe for the animal, and does not elicit considerable stress. Intraperitoneal injection of AAV8 vectors was shown to be effective for cardiac transduction in the mouse21. Neonatal gene transfer has some advantages from an immunological point of view since neonates have an immature immune system and inoculation at this period has shown to induce tolerance to the transgene products22, and relatively lower vector doses are needed. Here we show for the first time, to the best of our knowledge, that a single intraperitoneal injection of AAV9 based vectors in neonatal rats is sufficient to achieve a near complete and long-lasting transduction of the adult rat heart. The ‘low dose’ used in our study of 2x1011 viral genomes is similar to the dose used in mice21, however this dose resulted in low percentage of transduced cardiomyocytes in rats. In contrast, the ‘high dose’ of 1.3×1012 viral genomes was sufficient for a near complete transduction. Defining this range will allow future researchers to titrate the AAV dose to the desired level of transgene expression.\n\nThe use of AAVs is not without disadvantages. A major limitation of using AAV vectors is the relatively small transgene size (~ 4.7 kilobases) that can be cloned to the virus backbone; therefore, our approach cannot be used effectively for the expression of large genes. Recombinant AAV constructs in which the transgene does not encode a potentially tumorigenic gene product or a toxin molecule and is produced in the absence of a helper virus can usually be handled in a Biosafety Level 1 facility, but otherwise, a Biosafety Level 2 or higher may be required.\n\nThe targeting of animal genomes to add, remove, or substitute coding or non-coding sequences has revolutionized cardiovascular research. The development of the CRISPR technology has further facilitated and expanded these tools23. This approach has already been utilized in the rat24, but has not gained a wide-spread use as in the mouse, and generation of gene modified rats remains a difficult, time- and resources- consuming endeavor. Here we showed that a single intraperitoneal injection of AAV9 vectors encoding a transgene under the control of the cTnT promoter to neonatal rats resulted in a highly robust and highly specific cardiomyocyte transgene expression in the adult rat heart, with no signs of cardiotoxicity. In the future, this approach could be expanded to deliver Cas9 and gRNAs25, or to deliver small hairpin (sh)RNAs or artificial microRNA (amiRNAs)26 by AAVs to also achieve gene knockouts and knockdown in the rat heart.\n\n\nData availability\n\nFigshare: A simple Adeno Associated Virus (AAV) -Based Approach for The Generation of Cardiac Genetic Models in Rats_ Fig 1, https://doi.org/10.6084/m9.figshare.13148633.v227\n\nThis project contains the following underlying data:\n\n- Raw histological images as shown in Figure 1C in TIF format\n\nFigshare: A simple Adeno Associated Virus (AAV) -Based Approach for The Generation of Cardiac Genetic Models in Rats_figure 2, https://doi.org/10.6084/m9.figshare.13259624.v128\n\nThis project contains the following underlying data:\n\n- Raw echocardiographic images as shown in Figure 2A in JPG format\n\nFigshare: A simple Adeno Associated Virus (AAV) -Based Approach for The Generation of Cardiac Genetic Models in Rats_Fig 3, https://doi.org/10.6084/m9.figshare.13148663.v129\n\nThis project contains the following underlying data:\n\n- Raw immunofluorescence images as shown in Figure 3 in TIF format\n\nFigshare: A simple Adeno Associated Virus (AAV) -Based Approach for The Generation of Cardiac Genetic Models in Rats_Fig 4, https://doi.org/10.6084/m9.figshare.13148678.v230\n\nThis project contains the following underlying data:\n\n- Raw GFP signal images as shown in Figure 4 in TIF format\n\nFigshare: A simple Adeno Associated Virus (AAV) -Based Approach for The Generation of Cardiac Genetic Models in Rats_gravimetry_echo_GFP fluorescence_qPCR, https://doi.org/10.6084/m9.figshare.1325936631\n\nThis project contains the following underlying data:\n\n- gravimetric analysis.xlsx\n\n- Echo data.xlsx\n\n- GFP fluorescence analysis.xlsx\n\n- qPCR raw data.xlsx\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "References\n\nVirani SS, Alonso A, Benjamin EJ, et al.: Heart Disease and Stroke Statistics-2020 Update: A Report From the American Heart Association. Circulation. 2020; 141(9): e139–e596. PubMed Abstract | Publisher Full Text\n\nBraunwald E: Heart failure. JACC Heart Fail. 2013; 1(1): 1–20. PubMed Abstract | Publisher Full Text\n\nBraunwald E: The war against heart failure: the Lancet lecture. Lancet. 2015; 385(9970): 812–824. PubMed Abstract | Publisher Full Text\n\nRecchia FA, Lionetti V: Animal models of dilated cardiomyopathy for translational research. Vet Res Commun. 2007; 31 Suppl 1: 35–41. PubMed Abstract | Publisher Full Text\n\nPawlowski WP, Grelon M, Armstrong S: METHODS IN MOLECULAR BIOLOGYTM Series Editor. 2013. Publisher Full Text\n\nSakata S, Lebeche D, Sakata N, et al.: Restoration of mechanical and energetic function in failing aortic-banded rat hearts by gene transfer of calcium cycling proteins. J Mol Cell Cardiol. 2007; 42(4): 852–861. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIannaccone PM, Jacob HJ: Rats! Dis Model Mech. 2009; 2(5–6): 206–210. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMilani-Nejad N, Janssen PML: Small and large animal models in cardiac contraction research: advantages and disadvantages. Pharmacol Ther. 2014; 141(3): 235–249. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHasenfuss G: Animal models of human cardiovascular disease, heart failure and hypertrophy. Cardiovasc Res. 1998; 39(1): 60–76. PubMed Abstract | Publisher Full Text\n\nBreckenridge RA: Animal Models of Myocardial Disease. Animal Models for the Study of Human Disease. (Elsevier). 2013; 145–171. Publisher Full Text\n\nMolkentin JD, Robbins J: With great power comes great responsibility: using mouse genetics to study cardiac hypertrophy and failure. J Mol Cell Cardiol. 2009; 46(2): 130–136. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFELASA working group on revision of guidelines for health monitoring of rodents and rabbits, Mähler Convenor M, Berard M, et al.: FELASA recommendations for the health monitoring of mouse, rat, hamster, guinea pig and rabbit colonies in breeding and experimental units. Lab Anim. 2014; 48(3): 178–192. PubMed Abstract | Publisher Full Text\n\nChallis RC, Kumar SR, Chan KY, et al.: Systemic AAV vectors for widespread and targeted gene delivery in rodents. Nat Protoc. 2019; 14(2): 379–414. PubMed Abstract | Publisher Full Text\n\nWang D, Tai PWL, Gao G: Adeno-associated virus vector as a platform for gene therapy delivery. Nat Rev Drug Discov. 2019; 18(5): 358–378. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZincarelli C, Soltys S, Rengo G, et al.: Analysis of AAV serotypes 1-9 mediated gene expression and tropism in mice after systemic injection. Mol Ther. 2008; 16(6): 1073–1080. PubMed Abstract | Publisher Full Text\n\nBish LT, Morine K, Sleeper MM, et al.: Adeno-associated virus (AAV) serotype 9 provides global cardiac gene transfer superior to AAV1, AAV6, AAV7, and AAV8 in the mouse and rat. Hum Gene Ther. 2008; 19(12): 1359–1368. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPalomeque J, Chemaly ER, Colosi P, et al.: Efficiency of eight different AAV serotypes in transducing rat myocardium in vivo. Gene Ther. 2007; 14(13): 989–997. PubMed Abstract | Publisher Full Text\n\nBostick B, Ghosh A, Yue Y, et al.: Systemic AAV-9 transduction in mice is influenced by animal age but not by the route of administration. Gene Ther. 2007; 14(22): 1605–1609. PubMed Abstract | Publisher Full Text\n\nBostick B, Yue Y, Lai Y, et al.: Adeno-associated virus serotype-9 microdystrophin gene therapy ameliorates electrocardiographic abnormalities in mdx mice. Hum Gene Ther. 2008; 19(8): 851–856. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMüller OJ, Schinkel S, Kleinschmidt JA, et al.: Augmentation of AAV-mediated cardiac gene transfer after systemic administration in adult rats. Gene Ther. 2008; 15(23): 1558–1565. PubMed Abstract | Publisher Full Text\n\nWang Z, Zhu T, Qiao C, et al.: Adeno-associated virus serotype 8 efficiently delivers genes to muscle and heart. Nat Biotechnol. 2005; 23(3): 321–328. PubMed Abstract | Publisher Full Text\n\nZhang J, Xu L, Haskins ME, et al.: Neonatal gene transfer with a retroviral vector results in tolerance to human factor IX in mice and dogs. Blood. 2004; 103(1): 143–151. PubMed Abstract | Publisher Full Text\n\nMiano JM, Zhu QM, Lowenstein CJ: A CRISPR Path to Engineering New Genetic Mouse Models for Cardiovascular Research. Arterioscler Thromb Vasc Biol. 2016; 36(6): 1058–1075. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi W, Teng F, Li T, et al.: Simultaneous generation and germline transmission of multiple gene mutations in rat using CRISPR-Cas systems. Nat Biotechnol. 2013; 31(8): 684–686. PubMed Abstract | Publisher Full Text\n\nZhao H, Li Y, He L, et al.: In Vivo AAV-CRISPR/Cas9-Mediated Gene Editing Ameliorates Atherosclerosis in Familial Hypercholesterolemia. Circulation. 2020; 141(1): 67–79. PubMed Abstract | Publisher Full Text\n\nFechner H, Vetter R, Kurreck J, et al.: Silencing Genes in the Heart. In: Methods Mol Biol. (Humana Press Inc,). 2017; 1521: 17–39. PubMed Abstract | Publisher Full Text\n\nKehat I: A simple Adeno Associated Virus (AAV) -Based Approach for The Generation of Cardiac Genetic Models in Rats_ Fig 1. figshare. Figure. 2020. http://www.doi.org/10.6084/m9.figshare.13148633.v2\n\nKehat I: A simple Adeno Associated Virus (AAV) -Based Approach for The Generation of Cardiac Genetic Models in Rats_figure 2. figshare. Figure. 2020. http://www.doi.org/10.6084/m9.figshare.13259624.v1\n\nKehat I: A simple Adeno Associated Virus (AAV) -Based Approach for The Generation of Cardiac Genetic Models in Rats_Fig 3. figshare. Media. 2020. http://www.doi.org/10.6084/m9.figshare.13148663.v1\n\nKehat I: A simple Adeno Associated Virus (AAV) -Based Approach for The Generation of Cardiac Genetic Models in Rats_Fig 4. figshare. Figure. 2020. http://www.doi.org/10.6084/m9.figshare.13148678.v2\n\nKehat I: A simple Adeno Associated Virus (AAV) -Based Approach for The Generation of Cardiac Genetic Models in Rats_gravimetry_echo_GFP fluorescence_qPCR. figshare. dataset. 2020. http://www.doi.org/10.6084/m9.figshare.13259366.v2"
}
|
[
{
"id": "76025",
"date": "04 Jan 2021",
"name": "John Elrod",
"expertise": [
"Reviewer Expertise Cardiac physiology",
"heart failure",
"genetic mouse models",
"metabolism",
"mitochondria",
"calcium signaling"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nOverview: The manuscript “A simple adeno-associated virus-based approach for the generation of cardiac genetic models in rats” by Kehat et al. details the development of a simplified method for the cardiac-restricted genetic expression of constructs by i.p. delivery of AAV. The rationale for the current study is the need in the scientific community to have the ability to manipulate the rat genome to better utilize the many positive characteristics =this animal model has to offer such as the ability to perform surgical procedures, and how the rat is more physiologically similar to the human and the development of newer HFpEF models etc. The study verifies the benefits of using AAVs as a gene therapy due to the low cardiac toxicity, as well as the ability to have high expression restricted to cardiomyocytes with no evidence of inflammation or functional impairment as is often associated with tamoxifen and Cre-dependent models. The authors propose that a single intraperitoneal injection of AAV9 with a cardiac troponin T promotor creates a robust cardiac specific transgenic rat model with minimal no adverse effects or extra-cardiac expression. The manuscript is well written and without any major flaws.\nMinor Concerns:\nThe manuscript would benefit by including one schematic at the beginning showing the timeline of AAV delivery and when physiological and experimental assessment was performed.\n\nWhat is the quantification of GFP transduction for all of the hearts in Fig3B and Fig3C? In Fig5, the authors quantify n=3 in each group, however in figure 3, n=4 and in the methods n=5. Please clarify.\n\nWhat are the authors' thoughts on why Fig3C has 1 of the 4 mice with more intense GFP transduction? Is this due to different imaging setting due to background or something?\n\nThe authors did not discuss direct cardiac injection at day 1 or day 2, which does not require open heart surgery. What are your thoughts on this method as a way to increase efficacy while decreasing the viral load required?\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
},
{
"id": "77191",
"date": "04 Feb 2021",
"name": "Jop H. van Berlo",
"expertise": [
"Reviewer Expertise cardiovascular physiology and pathology"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe method article by Schlesinger-Laufer et al describes a simplified method for obtaining cardiomyocyte-specific expression of a gene of interest in rats. The authors provide a good rationale for why this method is useful and needed. The manuscript provides sufficient details to replicate the results presented.\nThe simplified method of generating a transgenic rat involves a single intraperitoneal injection of adeno-associated viral particles that can be readily produced in cell culture or purchased from commercial sources. The authors inject the AAV particles into neonatal rats in their experiments and evaluate expression at the adult stage. The authors show that there is dose-dependent gene transduction efficiency.\nThis is an important and timely report that will help investigators to implement transgenesis in their rat-based cardiovascular research.\nI only have a couple of minor comments to further improve the manuscript.\nThe methodology for measuring transduction efficiency is not very clear. Why is mean GFP intensity used as a determinant to establish a threshold for determining GFP expression? The control rat hearts should not have any GFP expression, and it makes more sense to use this as the baseline to establish threshold levels to distinguish autofluorescence from GFP expression.\n\nThe numbers reported are 5 animals for each group, but the text only mentions quantification on 3 animals, while the figures show 4 replicates for each group. Why this discrepancy?\n\nWas any thresholding done to the images prior to measuring GFP expression?\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1441
|
https://f1000research.com/articles/9-1439/v1
|
10 Dec 20
|
{
"type": "Research Article",
"title": "Clinical characteristics and predictors of the duration of hospital stay in COVID-19 patients in Jordan",
"authors": [
"Rami S. Alqassieh",
"Isam K. Bsisu",
"Mohammed Qussay Al-Sabbagh",
"Naser M. El-Hammuri",
"Moh’d A. Yousef",
"Mohammad A. El Jarbeh",
"Ahmed A. Sharqawi",
"Heba Z. Smadi",
"Sami A. Abu-Halaweh",
"Mohammad M. Abufaraj",
"Rami S. Alqassieh",
"Mohammed Qussay Al-Sabbagh",
"Naser M. El-Hammuri",
"Moh’d A. Yousef",
"Mohammad A. El Jarbeh",
"Ahmed A. Sharqawi",
"Heba Z. Smadi",
"Sami A. Abu-Halaweh"
],
"abstract": "Background: On March 11th, 2020, the World Health Organization (WHO) declared coronavirus disease 2019 (COVID-19) as a global pandemic. Healthcare systems in low- and middle-income countries may face serious limitations during a pandemic, for which understanding the predictors of prolonged hospital stay are crucial in decreasing the mortality rate. The aim of this study was to investigate the predictors of increased length of hospitalization among COVID-19 patients. Methods: In this prospective study, we investigated the effect of presenting symptoms and laboratory investigations on the duration of hospitalization of 131 COVID-19 patients at a tertiary hospital in Jordan from March 17th to April 9th, 2020. Results: Patients median age was 24 years [interquartile range (IQR): 8-39], of which 67 (51.15%) were males and 64 (48.85%) were females. Smokers had shorter in-hospital stay (OR: -3.52; 95% CI: -6.73 to -0.32; P=0.03). Taste loss (OR: 5.1; 95% CI: 1.95 to 8.25; P<0.01) and chills or rigors (OR: 4.08; 95% CI: 0.73 to 7.43; P=0.02) were the symptoms significantly associated with increased in-hospital stay, while those who had malaise (OR: -4.98; 95% CI: -8.42 to -1.59; P<0.01) and high white blood cell (WBC) count (OR: -0.74; 95% CI: -1.31 to -0.17; P=0.01) had faster recovery. Conclusions: Our study found that the most common presenting symptoms of COVID-19 are cough, malaise, and headache. Smoking, presenting with malaise or elevated WBCs were associated with shorter hospital stay, while loss of taste and chills or rigors at presentation were associated with a longer in-hospital stay.",
"keywords": [
"COVID-19",
"SARS-CoV-2",
"symptoms",
"smoking",
"Hydroxychloroquine."
],
"content": "Introduction\n\nIn December, China reported a group of pneumonia cases of unknown etiology that were later identified to be caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)1,2. On March 11th, 2020, the World Health Organization (WHO) categorized coronavirus disease 2019 (COVID-19) as a global pandemic3. Patients with COVID-19 frequently present with a cluster of different respiratory symptoms, including fever, chills cough, shortness of breath, sore throat, and new loss of taste and/or smell within 2 to 14 days after exposure in most cases4.\n\nCoronaviruses belong to viral family Coronaviridae (order Nidovirales) and include viruses approximately 26–32 kilobases in size with a positive-sense single-stranded RNA genome (+ssRNA)5. The Coronaviridae family contains four genera, one of which is the Betacoronavirus genus, which SARS-CoV-2 belongs to 6.\n\nInitial assessments of the epidemiologic characteristics and transmission dynamics showed that the basic reproductive number (R0), which is defined as the expected number of additional cases that can directly be generated by one case in a population susceptible to infection on average over the course of its infectious, range from 2.2 to 3.587,8, with the potential for asymptomatic transmission being a major concern for most of previous investigations.\n\nThe first identified case of COVID-19 in Jordan was on March 2nd, 2020, a returning traveler two weeks prior to quarantine procedures9. On March 13th, a wedding ceremony led to a large outbreak of COVID-19 cases in northern Jordan10, after which a strict lockdown took place and five tertiary hospitals were selected to provide medical care for patients suspected or diagnosed COVID-19 cases9,10.\n\nHealthcare systems in low- and middle-income countries may face serious limitations in capacity and accessibility during a pandemic, leading to worse clinical outcomes and an increase in mortality rate11. Therefore, the aim of this study is to investigate the predictors of increased length of hospitalization among COVID-19 patients, in order to provide evidence-based public health outbreak response strategies for COVID-19 and future pandemics.\n\n\nMethods\n\nIn this prospective observational investigation, we reviewed COVID-19 patients who were admitted to the isolation center at Prince Hamza Hospital (PHH), which is a tertiary hospital in Amman, the capital of the Hashemite Kingdom of Jordan. We included Jordanian patients above the age of 18 years who were diagnosed with COVID-19 and admitted to the isolation center of PHH. The diagnosis of COVID-19 was made after the collection of nasopharyngeal swabs using Xpert® sample collection kit (catalog number XPRSARS-COV2-CE-10, Cepheid, Sunnyvale, CA, USA)12 at the emergency department of five tertiary hospitals in Amman, after which positive cases were transferred to the isolation center at PHH. The data collection took place between March 17th till April 9th, 2020, during which all of the 131 patients admitted to PHH with the primary diagnosis of COVID-19 were included in our investigation, representing 40.4% of the total 324 cases diagnosed with COVID-19 in Jordan during the study’s timeframe. None of the patients were not eligible, declined to be enrolled, or withdrew from the study, and the study was reported in accordance with the STROBE statement (https://www.strobe-statement.org/). It is noteworthy that, in Jordan, all patients diagnosed with COVID-19 were admitted to hospital during the study’s timeframe, regardless of the severity of their illness.\n\nBased on semi-structured interviews by anesthesia and intensive care resident physicians, the demographic data, smoking habit, and past medical history of patients were documented directly from the patients during medical history taking. Moreover, we documented the current presenting symptoms, including cough, shortness of breath, chest pain, fever, chills/rigors, sweating, malaise, myalgia, headache, diarrhea, abdominal pain, palpitations, loss of taste, loss of smelling, nasal congestion, and rhinorrhea. Furthermore, baseline vital signs and laboratory investigations were collected, and whether the physician started the patient on hydroxychloroquine as a treatment. All patients were followed-up daily from admission till discharge from the hospital after clinical resolution and having negative test results.\n\nThe study protocol was approved by the Institutional Review Board (IRB) committee of The Hashemite University (No. 1/10/2019-2020). Written informed consent was obtained from all patients prior to participation in the study. All patients were able to withdraw from the investigation at any time without affecting their care. No identifying information were obtained from the patients, and all collected data were used solely for statistical analysis.\n\nStatistical analysis was performed using STATA (Stata Statistical Software: Release 16. College Station, TX: StataCorp LLC). The total sample number comprised of 131 patients. The association between the categorical baseline variables and the length of hospitalization was studied using Mann Whitney-U test for the dichotomous variables and Kruskal- Wallis test for the rest of the poly chotomous variables. Data were reported as medians with interquartile range (IQR) as they were non-normally distributed (P<0.01 for the Shapiro-Wilk test).\n\nA Linear regression analysis was used to examine the factors predicting the length of hospitalization. Baseline characteristics, presenting symptoms, and laboratory results at day of presentation were first analyzed using univariable linear regression analysis. Then, only significant variables were fitted into the final multivariable linear regression analyses. We reported the regression coefficients with their level of significance; p-value and 95% confidence intervals (95% CI). The p-value of statistical tests was two-sided, and statistically significant results were defined as those with a p-value <0.05.\n\n\nResults and discussion\n\nThe median age of included patients was 24 years [interquartile range (IQR): 8-39], of which 67 (51.15%) were males and 64 (48.85%) were females. Of the 131 patients, 29 (22.14%) were smokers. No comorbid conditions at presentation was recorded for 95 patients (72.52%), while 9 (6.87%) had hypertension, 8 (6.11%) had chronic respiratory diseases, and 7 (5.34%) had diabetes mellitus [Table 113].\n\nOverall, 74 (57%) patients presented with cough, of which 47 (36%) had dry cough, while 27 (21%) had productive cough. Malaise (n= 62; 47%) and headache (n=59; 45%) were the second and third commonest presenting symptoms, followed by loss of smell (n=54; 41%), loss of taste (n=51; 29%), and diarrhea (n=51; 39%). Only 49 (37%) patients presented with nasal congestion, chills, or rigors, and 48 (37%) presented with myalgia. Of the 131 patients, 42 (32%) had fever at time of presentation [Table 213].\n\nOverall, smokers had shorter in-hospital stay (β: -3.52; 95% CI: -6.73 to -0.32; P=0.03). Moreover, taste loss (β: 5.1; 95% CI: 1.95 to 8.25; P<0.01) and chills or rigors (β: 4.08; 95% CI: 0.73 to 7.43; P=0.02) were the symptoms significantly associated with increased in-hospital stay, while those who had malaise (β: -4.98; 95% CI: -8.42 to -1.59; P<0.01) and high white blood cell (WBC) count (β: -0.74; 95% CI: -1.31 to -0.17; P=0.01) at presentation had faster recovery. Hydroxychloroquine was not associated with decreasing the duration of hospital stay (β: -2.55; 95% CI: -5.67 to 0.56; P=0.11) [Table 313].\n\n\nDiscussion\n\nIn spite of the great progress in understanding COVID-19, a therapeutic or preventive solution is yet to be achieved14. Consequently, better management of medical facilities during the next phase of this pandemic is crucial in improving the outcomes in these patients15. In the present study, smoker, as well as patients presenting with malaise and elevated WBCs at presentation had shorter hospital stay, while loss of taste and chills or rigors at presentation were associated with lengthier in-hospital stay.\n\nSeveral previous studies investigated the clinical manifestations of COVID-19. Most common presenting symptoms in most of these studies were fever, cough, dyspnea, malaise and myalgia16,17. Interestingly, it has been suggested that anosmia (loss of smelling sense) and ageusia (loss of taste function) can represent the first or only symptomatology18, while an investigation from Italy found that 64% mildly symptomatic patients had impaired olfaction19.\n\nRemarkably, even though multivariable linear regression analyses in the current study did not show significant correlation between demographic factors and the length of in-hospital stay except for smoking habit, an investigation from Germany revealed that those with preexisting respiratory diseases, obese patients, and those with persistently elevated inflammatory markers are at increased risk of developing acute respiratory distress syndrome (ARDS), which will prolong their hospitalization period20. Moreover, a study conducted in China found that severe cases more frequently had dyspnea, lymphopenia, and hypoalbuminemia, with higher levels of c-reactive protein, d-dimer, lactate dehydrogenase, alanine aminotransferase, ferritin, IL-2R, IL-6, IL-10, and TNF-α21. In the current study, patients with elevated WBCs count had shorter in-hospital stay. Immunocompetent WBCs play a significant role in systemic inflammatory response to infection, with neutrophil-lymphocyte count ratio being significantly higher in mortality cases of community-acquired pneumonia22.\n\nAn interesting finding in our study is that smokers had shorter in-hospital stay, moreover, only 22% of hospitalized COVID-19 patients were smokers. There is a controversy in the current literature about the role of smoking and nicotine in COVID-1923,24. It is postulated that smokers are at higher risk to get respiratory tract infections and develop more severe illness, due to the preexisting bronchopulmonary damage, reduced muco-ciliary clearance, and the local inflammatory status23. Some reports, however, showed lower COVID-19 related mortality and morbidity in smokers24. Nicotine, as a cholinergic agonist, inhibits pro-inflammatory cytokines such as TNF, IL-1, IL-6 by binding acetylcholine receptors (nAChR). These cytokines, among the others, might result in the notorious feature of this illness, the cytokine storm. Therefore, it was hypothesized that COVID-19 is a disease of nicotinic cholinergic system. Another supporting observation is that patients with ageusia in our sample had longer in-hospital stay; this sensory disturbance is mainly related to the cholinergic system of the brain, and it might indicate that such patients are having more extensive disease, or technically, nicotinic cholinergic dysfunction23,24. Although it is not wise or acceptable to advice people to smoke, this might shed the light on the promising role of nicotine in preventing and treating COVID-19 infection.\n\nThe rapid upsurge in the number of confirmed cases makes the control of the spread of COVID-19 and its treatment challenging25. Several ongoing clinical trials will soon confirm or disprove the usefulness of several candidate medications in treating COVID-1926. Studies investigating the use of hydroxychloroquine were unable to confirm its benefit on in-hospital outcomes of COVID-19 patients27. On the other hand, dexamethasone has shown a decrease in the 28-day mortality among those who were receiving respiratory support, with both of dexamethasone and methylprednisolone being equally effective in treating moderate to severe COVID-19 cases28,29. With no definitive method of prevention or treatment being available to date, precautions should be made in order to control the spread of COVID-19, including proper hand hygiene, isolation of infected or suspected persons in properly ventilated hospitals, social distancing, discouraging large gatherings, and avoiding direct contact with suspected animal reservoir hosts30.\n\nThe main limitation of this study is that it did not investigate the predictors of in-hospital mortality of COVID-19 patients. Previous studies suggested that older age, high Sequential Organ Failure Assessment (SOFA) score, and elevated d-dimer are associated with poor prognosis31,32. However, upon plotting mortality against the incidence of COVID-19, a significant positive correlation was found, suggesting that mortality is associated with heavier healthcare burden33. Hence, understanding the predictors of prolonged hospital stay and prognostic factors are crucial in decreasing the mortality of any pandemic. Moreover, we recommend future studies to take into consideration duration and severity of each symptom in order to develop a better understanding of the clinical course of COVID-19. The main strength of our study is that, in Jordan, all COVID-19 patients were hospitalized during the study’s timeframe, regardless of the severity of their illness. Therefore, our sample gives a good opportunity to observe the natural progression of COVID-19 under controlled conditions.\n\n\nConclusion\n\nOur study found that the most common presenting symptoms of COVID-19 are cough, malaise, and headache. Smoking, presenting with malaise or elevated WBCs were associated with shorter hospital stay, while loss of taste and chills or rigors at presentation were associated with a longer in-hospital stay. Such findings are important in risk stratifying COVID-19 patients according to their presenting symptoms and past medical history. Although it is not wise or acceptable to consider smoking behaviors, future research should shed light on the role of nicotinic receptors in mitigating this illness.\n\n\nData availability\n\nHarvard Dataverse: Clinical characteristics and predictors of the duration of hospital stay in COVID-19 patients in Jordan. https://doi.org/10.7910/DVN/SBV1K613\n\nThis project contains the following underlying data:\n\n- Phh data coded-final.tab (Clinical characteristics and predictors of the duration of hospital stay in COVID-19 patients in Jordan)\n\n- Codebook.docx (Data codebook)\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).",
"appendix": "Acknowledgements\n\nThe authors would like to thank healthcare workers and first responders across the globe for their efforts during this pandemic. Moreover, the authors would like to thank the WHO for the great effort they made in response to this pandemic.\n\n\nReferences\n\nZhu N, Zhang D, Wang W, et al.: A Novel Coronavirus from Patients with Pneumonia in China, 2019. N Engl J Med. 2020; 382(8): 727–33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMunster VJ, Koopmans M, van Doremalen N, et al.: A Novel Coronavirus Emerging in China - Key Questions for Impact Assessment. N Engl J Med. 2020; 382(8): 692–4. PubMed Abstract | Publisher Full Text\n\nCucinotta D, Vanelli M: WHO Declares COVID-19 a Pandemic. Acta Biomed. 2020; 91(1): 157–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcMichael TM, Currie DW, Clark S, et al.: Epidemiology of Covid-19 in a Long-Term Care Facility in King County, Washington. N Engl J Med. 2020; 382(21): 2005–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWeiss SR, Navas-Martin S: Coronavirus pathogenesis and the emerging pathogen severe acute respiratory syndrome coronavirus. Microbiol Mol Biol Rev. 2005; 69(4): 635–64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAbd El-Aziz TM, Stockand JD: Recent progress and challenges in drug development against COVID-19 coronavirus (SARS-CoV-2) - an update on the status. Infect Genet Evol. 2020; 83: 104327. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi Q, Guan X, Wu P, et al.: Early Transmission Dynamics in Wuhan, China, of Novel Coronavirus-Infected Pneumonia. N Engl J Med. 2020; 382(13): 1199–207. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLai CC, Shih TP, Ko WC, et al.: Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and coronavirus disease-2019 (COVID-19): The epidemic and the challenges. Int J Antimicrob Agents. 2020; 55(3): 105924. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSuleiman A, Bsisu I, Guzu H, et al.: Preparedness of Frontline Doctors in Jordan Healthcare Facilities to COVID-19 Outbreak. Int J Environ Res Public Health. 2020; 17(9): 3181. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYusef D, Hayajneh W, Awad S, et al.: Large Outbreak of Coronavirus Disease among Wedding Attendees, Jordan. Emerg Infect Dis. 2020; 26(9): 2165–2167. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSiow WT, Liew MF, Shrestha BR, et al.: Managing COVID-19 in resource-limited settings: critical care considerations. Crit Care. 2020; 24(1): 167. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSheridan C: Fast, portable tests come online to curb coronavirus pandemic. Nat Biotechnol. 2020; 38(5): 515–8. PubMed Abstract | Publisher Full Text\n\nBsisu I, Alqassieh R, Abu-Halaweh S, et al.: Clinical characteristics and predictors of the duration of hospital stay in COVID-19 patients in Jordan. Harvard Dataverse, V2. 2020. http://www.doi.org/10.7910/DVN/SBV1K6\n\nDel Rio C, Malani P: Translating Science on COVID-19 to Improve Clinical Care and Support the Public Health Response. JAMA. 2020; 323(24): 2464–2465. PubMed Abstract | Publisher Full Text\n\nHopman J, Allegranzi B, Mehtar S: Managing COVID-19 in Low- and Middle-Income Countries. JAMA. 2020; 323(16): 1549–1550. PubMed Abstract | Publisher Full Text\n\nChen N, Zhou M, Dong X, et al.: Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study. Lancet. 2020; 395(10223): 507–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuang C, Wang Y, Li X, et al.: Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. Lancet. 2020; 395(10223): 497–506. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVaira LA, Salzano G, Deiana G, et al.: Anosmia and Ageusia: Common Findings in COVID-19 Patients. Laryngoscope. 2020; 130(7): 1787. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSpinato G, Fabbris C, Polesel J, et al.: Alterations in Smell or Taste in Mildly Symptomatic Outpatients With SARS-CoV-2 Infection. JAMA. 2020; 323(20): 2089–2090. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDreher M, Kersten A, Bickenbach J, et al.: The Characteristics of 50 Hospitalized COVID-19 Patients With and Without ARDS. Dtsch Arztebl Int. 2020; 117(16): 271–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen G, Wu D, Guo W, et al.: Clinical and immunological features of severe and moderate coronavirus disease 2019. J Clin Invest. 2020; 130(5): 2620–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nde Jager CPC, Wever PC, Gemen EFA, et al.: The neutrophil-lymphocyte count ratio in patients with community-acquired pneumonia. PLoS One. 2012; 7(10): e46561. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvan Zyl-Smit RN, Richards G, Leone FT: Tobacco smoking and COVID-19 infection. Lancet Respir Med. 2020; 8(7): 664–665. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFarsalinos K, Niaura R, Le Houezec J, et al.: Editorial: Nicotine and SARS-CoV-2: COVID-19 may be a disease of the nicotinic cholinergic system. Toxicol Rep. 2020; 7: 658–663. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZheng YY, Ma YT, Zhang JY, et al.: COVID-19 and the cardiovascular system. Nat Rev Cardiol. 2020; 17(5): 259–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRome BN, Avorn J: Drug Evaluation during the Covid-19 Pandemic. N Engl J Med. 2020; 382(24): 2282–2284. PubMed Abstract | Publisher Full Text\n\nMehra MR, Desai SS, Ruschitzka F, et al.: RETRACTED: Hydroxychloroquine or chloroquine with or without a macrolide for treatment of COVID-19: a multinational registry analysis. The Lancet. 2020. Publisher Full Text\n\nRECOVERY Collaborative Group,Horby P, Lim WS, et al.: Dexamethasone in Hospitalized Patients with Covid-19 - Preliminary Report. N Engl J Med. 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFatima SA, Asif M, Khan KA, et al.: Comparison of efficacy of dexamethasone and methylprednisolone in moderate to severe covid 19 disease. Ann Med Surg (Lond). 2020; 60: 413–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPeeri NC, Shrestha N, Rahman MS, et al.: The SARS, MERS and novel coronavirus (COVID-19) epidemics, the newest and biggest global health threats: what lessons have we learned? Int J Epidemiol. 2020; 49(3): 717–726. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhou F, Yu T, Du R, et al.: Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study. Lancet. 2020; 395(10229): 1054–62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang L, Yan X, Fan Q, et al.: D-dimer levels on admission to predict in-hospital mortality in patients with Covid-19. J Thromb Haemost. 2020; 18(6): 1324–1329. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJi Y, Ma Z, Peppelenbosch MP, et al.: Potential association between COVID-19 mortality and health-care resource availability. Lancet Glob Health. 2020; 8(4): e480. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "85534",
"date": "02 Jun 2021",
"name": "Denise Battaglini",
"expertise": [
"Reviewer Expertise Intensive care medicine",
"mechanical ventilation",
"ARDS",
"COVID-19",
"infectious disease",
"neurocritical care"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study investigates the predictors of hospital length of stay in COVID-19 patients in Jordan.\n\nThe study is well written and interesting. However, it has a lack of novelty and should be improved. I would suggest to add more information: 1) Hospital length of stay is often made by different wards and eventually ICU. I think it is important to understand which patients were admitted to ICU, if some of them were endotracheally intubated, tracheostomize, if some patients had hemorrhage, thrombosis, infections, other complications, which PaO2/FiO2 on admission, if they were non-invasively ventilated (CPAP, NIPPV, High flow), if CPR, D-dimer, previous antibiotic therapy, SOFA on admission, Charlson comorbidity index, steroidal therapy, sedation, analgesia, myorelaxants, etc. and other factors that could have been predictors of hospital stay.\n\nThe study aims to investigate only predictors but I believe that there is a lack of some important factors which could have changed patients' clinical course. I suggest to extend the analysis to other important factors and, if possible to divide between those who survived and those who did not OR those who were admitted to ICU/those who did not.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "6752",
"date": "09 Jun 2021",
"name": "Isam Bsisu",
"role": "Author Response",
"response": "We would like to thank Dr. Denise Battaglini for his valuable review, which will enable us to improve the quality of our manuscript. This study included data collected at the beginning of the pandemic in Jordan, during which all patients diagnosed with COVID-19 were admitted to hospital during the study’s timeframe. We agree with the reviewer that the aforementioned laboratory investigations and interventions are important in determining the clinical course, and could have been predictors of hospital stay. Due to the limited number of cases at that time, we were unable to compare between cases in terms of outcomes and ICU admission. For instance, there were only 7 mortality cases related to COVID-19 in the whole country on the last day of patients recruitment, April 9th, 2020. Moreover, this prospective study only documented the laboratory investigations performed at that time, which were done based on their indication and patient's clinical scenario. The scientific knowledge and clinical practice is dynamically changing during COVID-19 pandemic, and so does the indicated investigations, the guidelines for medical management, and post discharge medical care for cured patients. We understand that this study was conducted and published prior to the two peaks in Jordan. However, it provided unique data in which we can understand the predictors of hospital stay based on one of the most important steps in the clinical practice of any physician, which is history taking and clinical examination. In addition, simple laboratory values can be of great benefit as a predictor. Future studies including patients from the whole country during the two peaks can be conducted (as we will recommend in our discussion section in the updated version 2), and it will surely be of great benefit in understanding the clinical course of the disease and its effect on the hospital stay and clinical outcome. "
}
]
},
{
"id": "86710",
"date": "11 Jun 2021",
"name": "Omar Soliman Mohamed El-Masry",
"expertise": [
"Reviewer Expertise Clinical laboratory sciences with a special focus on cancer biology and biomarkers"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article titled \"Clinical characteristics and predictors of the duration of hospital stay in COVID-19 patients in Jordan” represents an attempt to assess clinical factors that might be associated with COVID-19 patients' hospitalization in Jordan. The article rationale is good, however, it cannot reflect the figure in the whole country as data being taken from five centers and included a limited number of patients. I would suggest revising the title unless the data at this time represents the total number of patients in the whole country.\nIn the introduction, the authors started to recount history of the beginning of COVID-19 observation in China, but the year was not mentioned (December 2019). Please, add 2019.\nThe introduction should include a background section on factors reported in the study that might affect patients’ hospitalizations that were reported, at least, for similar diseases (MERS, for example). In addition, the authors should discuss the other factors that could affect this parameter; such as comorbidities.\nIn the material and methods’ section, the following sentence “It is noteworthy that, in Jordan, all patients diagnosed with COVID-19 were admitted to hospital during the study’s timeframe, regardless of the severity of their illness” needs further clarification; does this mean that those were all COVID-19 patients reported in the whole country? If yes, it would be very early to generalize the findings of the current study and this must be clearly indicated as a limitation.\nIn the results section, smoking status was not found as a predictor for the length of the hospital stay, which is odd knowing that COVID-19 patients suffer from serious lung problems. Did the author investigate confounding factors with smoking status; such as age, for example? Also, I think calculating odd ratio is not suitable for the study design.\nI think the findings were concluded from a premature study, which was conducted at the very beginning of the Corona crisis; therefore, the conclusions are premature and cannot reflect the logical and the expected conclusions regarding COVID-19. Thus, the study should be revised by including data of a larger sample size to be more representative and provide evidence-based conclusions. Having said that, the study rationale is good and interesting; but when supported with robust design, it will be of more interest to the scientific community and will better reflect the real situation.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1439
|
https://f1000research.com/articles/9-1438/v1
|
10 Dec 20
|
{
"type": "Research Article",
"title": "Measuring indicators of health system performance for palliative and end-of-life care using health administrative data: a scoping review",
"authors": [
"Suman Budhwani",
"Ashlinder Gill",
"Mary Scott",
"Walter P. Wodchis",
"JinHee Kim",
"Peter Tanuseputro",
"Ashlinder Gill",
"Mary Scott",
"Walter P. Wodchis",
"JinHee Kim",
"Peter Tanuseputro"
],
"abstract": "Background: A plethora of performance measurement indicators for palliative and end-of-life care currently exist in the literature. This often leads to confusion, inconsistency and redundancy in efforts by health systems to understand what should be measured and how. The objective of this study was to conduct a scoping review to provide an inventory of performance measurement indicators that can be measured using population-level health administrative data, and to summarize key concepts for measurement proposed in the literature. Methods: A scoping review using MEDLINE and EMBASE, as well as grey literature was conducted. Articles were included if they described performance or quality indicators of palliative and end-of-life care at the population-level using routinely-collected administrative data. Details on the indicator such as name, description, numerator, and denominator were charted. Results: A total of 339 indicators were extracted. These indicators were classified into nine health care sectors and one cross-sector category. Extracted indicators emphasized key measurement themes such as health utilization and cost and excessive, unnecessary, and aggressive care particularly close to the end-of-life. Many indicators were often measured using the same constructs, but with different specifications, such as varying time periods used to ascribe for end-of-life care, and varying patient populations. Conclusions: Future work is needed to achieve consensus ‘best’ definitions of these indicators as well as a universal performance measurement framework, similar to other ongoing efforts in population health. Efforts to monitor palliative and end-of-life care can use this inventory of indicators to select appropriate indicators to measure health system performance.",
"keywords": [
"palliative care",
"terminal care",
"end-of-life care",
"quality indicators",
"performance measurement",
"evaluation"
],
"content": "Introduction\n\nGrowing health care costs in an environment of tight financial constraints and an aging population are challenging many health care systems globally. Recent health care reform initiatives have underscored the importance of providing quality care to patients at all phases of the disease trajectory, including palliative care (PC) both specialist and generalist, and end-of-life care (EOLC), to improve patient and health system outcomes1,2. The Worldwide Palliative Care Alliance and the World Health Organization recommend provision of PC for all persons with chronic and life limiting/threatening conditions early in the illness trajectory3. However, universal access to quality, integrated and timely PC for patients remains intermittent at best, with PC largely being provided close to the end-of-life (EOL)4. Moreover, concerns about high costs associated with health care at the EOL are prevalent in the literature; studies attribute these costs to intensive, aggressive, and sometimes unnecessary utilization of institutionalized care, such as inpatient hospital services in the last few months of life5.\n\nMonitoring health system performance is one important component of evaluating achievement of quality care for improved patient and health system outcomes. This can be done through health system performance indicators, which summarize directional quantitative information on the quality of health, often evaluated through measurement of structure (inputs) or characteristics of the health care system, process (outputs) or services required to provide health care, and outcomes or measures of ultimate impact of the health care provided6. Health system performance indicators allow comparisons across jurisdictions, organizations, and/or administrative databases to track progress over time in efforts to improve health care quality7. Quantitative information for indicators can utilize health administrative data or data that is routinely collected in the process of delivering health care programs and services, often used due to its extensive reach (many databases are population-based), low cost, and low respondent burden8. Administrative data can also enable contextualization of health system performance indicators due to availability of data on patient sociodemographic and clinical characteristics.\n\nIn recent years, jurisdictions in Canada, the United States, and the United Kingdom have focused on defining standards and monitoring health care quality as efforts to expand PC programs intensify9–14, and as a result a plethora of health system performance indicators for PC and EOLC using health administrative data have been proposed13,15–18. These indicators have been suggested based on differing objectives (e.g. quality improvement, performance measurement) and/or for different patient populations (e.g. cancer, intensive care units, long-term care (nursing home) patients)13,19–24. Given the diversity of indicators in the literature, a scoping and cataloguing exercise becomes important in order to consolidate and summarize common measurement concepts (themes) and indicators that can be recommended for use in the performance measurement of PC and EOLC. Such a catalogue can serve as a reference guide for use by policy and decision makers, thereby reducing potential redundant efforts on collecting indicators, and enabling movement to the next step of decision making on what to measure and how best to measure it given each jurisdiction’s individual data systems. Furthermore, categorizing indicators by health care sector or setting of health care delivery can be a helpful demarcation that enables measurement of health system performance to align with funding and accountability structures25. Consideration of health care setting or health sector (e.g., acute care hospitals, long-term care (LTC), home care) becomes important given each sector’s potentially differing care quality aims, target populations, health care processes, and definitions of outcomes. Such organization does not preclude inclusion of indicators of transitional care or coordination of care between health care settings, but rather enables tailoring of indicator definitions that are more reflective of the relevant sector or health setting’s contribution to system-level performance.\n\nAs such, the primary objective of this study was to create a catalogue or inventory of health system performance measurement concepts and indicators for PC and EOLC utilizing routinely-collected population-level health administrative databases and as categorized by health sector by conducting a scoping literature review.\n\n\nMethods\n\nThis scoping literature review of health system performance indicators for PC and EOLC followed the Arksey and O’Malley (2005) methodological framework for scoping reviews26. Scoping reviews are a “type of knowledge synthesis, (that) follow a systematic approach to map evidence on a topic and identify main concepts, theories, sources and knowledge gaps”27. The primary objective of this scoping review was to collect and map health system performance indicators for PC and EOLC using routinely collected population-based health administrative data and to organize indicators by health care sector. Since this required consulting a broad array of literature to uniformly collect indicators, rather than appraisal of published indicators, this objective aligned well with the rationale for the use of scoping review methodology in comparison to systematic review methodology, and hence a scoping review methodology was chosen to complete research objectives28. A scoping review protocol was not published a priori for this study. Subsequent sections describe specific steps taken for this scoping review in accordance with the Arksey and O’Malley (2005) methodological framework.\n\nWe collected both peer-reviewed and grey literature published in English for this scoping review. For peer-reviewed literature, we conducted a search of MEDLINE and EMBASE databases between years of 2010 to July 2018. A University of Toronto librarian and comparable literature reviews were consulted in the development of the search strategy, which included a combination of the United States National Library of Medicine Medical Subject Headings (MeSH)29 and keywords (see Extended data: S1 Appendix Search Strategy30). For grey literature, we utilized Google search engine using a combination of keywords such as “palliative care”, “end-of-life-care”, “performance indicators”, and/or “quality indicators”, as well as incorporated reference documents made available to us through our knowledge user partners based on their previously conducted literature searches. No jurisdictional restrictions were placed for the grey literature search; however Canadian references were generally better known by study authors and knowledge users, and formed the bulk of included references.\n\nInclusion and exclusion criteria were developed in accordance with the study objectives and points of inquiry for the review, notably: a) focused on core concepts of quality or performance measurement in palliative and/or end-of-life care, b) measurement conducted using population-level, routinely-collected data, and c) studies with a health system focus. The full set of generated articles was reviewed based on developed inclusion and exclusion criteria (see Extended data: S2 Table Study Inclusion and Exclusion Criteria30).\n\nReviewers (SB and AG) independently screened all articles using titles and abstracts, with any conflicts resolved through discussion and inclusion and exclusion criteria accordingly updated. This generated a total of 285 articles for full text review (Figure 1). Full text review was also independently conducted by SB and AG, with a total of 54 peer-reviewed studies and 42 grey literature documents being identified for indicator extraction. Additionally, forward reference searching of reference lists was conducted to include any additional relevant articles, as well as any other relevant studies known to the all study authors. Studies outside of the study time period of 2010 to July 2018 were included. This resulted in a total of 32 additional articles included into the study for data extraction.\n\nA data extraction tool was developed to effectively chart details on presented and potential indicators (such as indicator definition, numerator and denominator if available). Indicators were extracted if they a) measured health system performance for PC and EOLC, and b) used population-level data using health administrative datasets. Once the full set of indicators were charted, SB, AG and PT removed duplicates and conducted health sector classification. We chose to focus on categorizing our collected indicators by health sectors specifically for the Canadian province of Ontario, with a population of over 14.5 million residents31 with universal health coverage for costs associated with acute care, hospitalizations, physician visits, emergency room visits, long-term care, home care, complex continuing care, and medications for those meeting select age-based and need-based criteria32. Health administrative data is also collected comprehensively and at the population-level for the majority of health sectors with public health coverage, thereby increasing potential for performance measurement with indicators based on routinely-collected administrative data. For the purposes of our study, indicators were classified based on the following Ontario health sectors based on how data from health administrative databases is currently obtained and organized in Ontario administrative databases: hospital care (including emergency department (ED)); home care or care services provided in the home and community;33. LTC (i.e., nursing homes), hospice care, physician services or care services captured through physician billing codes; medications covered through the public health insurance system;34 complex continuing care (CCC) or technology-based care provided to patients with chronic and complex health conditions;35 cancer care or care specifically targeted for cancer prevention and treatment; and other. A category entitled “Cross Sectors” was created to capture indicators transcending more than one health sector (e.g. place of death at various care locations). Further thematic analysis was conducted collectively by the study authors in discussion to group indicators measuring similar constructs by common themes, leading to the creation of measurement themes for indicator classification under each health sector category.\n\nFollowing this initial categorization, a working group with subject matter experts (both researchers and clinicians) was organized to review the final list of indicators. Indicators were cross-checked to ensure a) all relevant concepts related to PC and EOLC had been captured, and b) all indicators were conceivably measurable using (Ontario/Canadian) health administrative data based on the knowledge and expertise of the subject matter experts present.\n\n\nResults\n\nA total of 1111 indicators were extracted from 128 articles. 722 indicators were excluded (due to reasons of duplication and irrelevancy), resulting in a total of 339 indicators included for summarization (Figure 1). Studies and grey literature ranged from publication dates of 1992 to 2018, with only 5 included studies published prior to 2003, and the majority (n=82) being published after 2008 (Table 1).\n\nThe 42 grey literature documents included publications from Canadian provinces (e.g., Ontario, Alberta, Saskatchewan) (n=31), where routinely collected administrative data is readily available, and from the United States (n=6), United Kingdom (n=4), and Australia (n=1). They included documents generated by PC delivery organizations, such as local home care agencies, and also jurisdictional collaborative efforts to improve care.\n\nTen health care sectors were utilized to categorize indicators, with sector categories of Cross Sector, Cancer and Home Care having the greatest number of indicators (Table 2). Table 3 presents a summary of indicators and key measurement themes by health care sector, and a collective list of all indicators collected can be found in Extended data: S1 Table Detailed List and Information on Collected Indicators30. These findings are discussed next.\n\nWithin this review, a total of 58 collected indicators were categorized as transcending sectors or cross sector9,13–18,20–24,36–106 and four measurement themes were identified. Indicators classified under the cross sector category tended to focus on how well palliative and EOL patients were being cared for through examination of aggressiveness of care, place of care/death, availability of palliative care services and overall cost; this measurement was to better understand how the overall health care system was performing in the care of palliative and EOL patients, rather than on the performance on any specific health care sector by itself (e.g. the enrollment of patients in a palliative care program and how early this enrollment occurred).\n\nA total of 57 indicators were extracted for measuring performance in the delivery of cancer care13,17,20,21,37,39,48,50,51,54,56,60,63,64,67,70,72–74,77–80,91,94,98,103,106–119. These indicators centered around three key measurement themes which were tied to the overall provision of quality care to cancer patients, including the regular assessment and management of symptoms, wait times for specific cancer care services, and aggressive care near the end-of-life.\n\nA total of 48 indicators were categorized under the home care sector13,16–18,39,42,43,52,58,65,67,68,70,72,80,87,92,99,103,110,123,128,130–136 with four key measurement themes. Most indicators described patient utilization/cost of, or access to PC and/or home care services. Indicators on patient outcomes and quality of care emphasized specific clinical characteristics (e.g. pain, cognitive function), and involvement of specific health care providers (multidisciplinary care, social workers).\n\nA total of 40 indicators were extracted for the LTC or nursing home sector43,48,54,60,61,67,70,74,76,87,91,100,106,137–144 classified under four key measurement themes. Indicators focused on the quality of care being provided through adequate symptom management and burdensome transitions at the EOL; indicators also focused on length of stay and costs of these patients at the end-of-life.\n\nA total of 36 indicators were categorized under the hospice sector25–29,31,34,45–47,51,52,54,57,58,60,63,64,67,69,72,75,76,80,81,90,96,101,116,122,128,141,142 and five key measurement themes were identified. Overall, indicators heavily emphasized access to timely, adequate and good quality hospice care, with a large number of indicators measuring access to hospice care across as measured based on different time periods prior to death, by disease groups, etc.\n\nA total of 39 indicators were extracted for the hospital sector19,22,23,25–29,31–35,41–54,56–60,63,64,66–86,89–91,95–108,110–113,116–119,120–122,123,127–133,135–143,145,148 under seven key measurement themes including ED use, intensive care unit use, and cost. Hospital-specific indicators largely focused on utilization, costs related to utilization, and the appropriateness of hospital care at the EOL.\n\nA total of 29 indicators were extracted for the physician care sector23,26,29,32,35,40,41,42,60,66,68,70,71–74,76,77,79,80,83,87,89,91,92,95,96,97,99,100,103,106,108,111,116,119,120,130,132,133,135,137,138,141,142,145,148 under three key measurement themes. Indicators largely emphasized understanding the intersection between generalist and specialized PC provided by different physicians (general physicians, PC specialists, oncologists etc.) in outpatient settings.\n\nA total of 26 indicators were extracted on medication use9,18,21,40,44,46,61,69,74,84,87,98,107,115,118,126,139,144,153–156 under three measurement themes. Indicators were largely focused on the most appropriate medications for symptom management, including overmedication and aggressive treatment at the end-of-life.\n\nA total of three indicators were measured within the CCC sector that typically focuses on rehabilitation and/or PC. They focused on pain management108, access to PC58, and cost68. This sector was the most underdeveloped in the literature compared to other sectors with respect to performance measurement likely due to its uniqueness to the Ontario context.\n\nThree indicators did not fit the sectors above. They were ambulance use by EOL patients56,87, medical lab services and equipment expenses144, and diagnostic testing at the end-of-life133. These indicators were classified together under the “Other” sector category.\n\n\nDiscussion\n\nThe purpose of our scoping review was to collect, organize, and share a distilled inventory of sector-specific health system performance measurement indicators for PC and EOLC. Our scoping review revealed 339 indicators, organized across nine distinct and one cross-cutting health care sector categories. Indicators within each sector category were subgrouped by key measurement themes. Collectively, these indicators represent the field’s thinking on how best to measure high health system performance in the delivery of both PC and EOLC.\n\nOne of the most commonly occurring measurement themes across sectors focused on health system utilization (e.g., hospital length of stay, ICU admissions, chemotherapy or home care received). Indicators of utilization were collected both for comparison across palliative/end-of-life and non-palliative/non-end-of-life populations (to represent access), to examine cost of delivered care, and also to understand how well the system was performing in the delivery of care (e.g. on wait times). However, emphasis on utilization through measurement of process indicators, such as in the hospital sector, highlighted that measurement themes such as patient experience and effectiveness of palliative care were not captured. Measurement of such themes through indicators can serve as an indication of the achievement of positive patient and health system outcomes.\n\nAnother frequent measurement theme across sector categories was the aggressiveness of care at the EOL for palliative and EOL populations. This theme included indicators of inappropriate treatment, medications, and transitions at the end-of-life that would be avoided in a well-performing health system focused on quality patient care. As health system resources become finite, this measurement theme will likely gain prominence in not only helping to improve patient outcomes, but also in reducing costs from unnecessary and aggressive health services.\n\nDespite the wealth of indicators collected, the measurement of patient-centered or patient-reported outcomes was infrequent at the population-level using health administrative databases. Measurement of themes related to unmet care and self-management support needs, advance care planning, goals of care, consideration of patient preferences, and patient and caregiver burden were limited. However, absence of indicators from the review does not mean that measurement is not occurring, but rather, that it may not be occurring in population-level health administrative databases. Given the importance of these indicators, measurement through primary data collection may be required to obtain a comprehensive picture of health system performance.\n\nAdditionally, the literature review revealed that although there were a large number of indicators, many of these indicators measured similar constructs, but with different specifications. This includes variations in identifying the time period of measurement and the patient population. There were also differences in how debated concepts in the literature were operationalized such as what constitutes markers of quality care (example, how to define a burdensome transition). S1 Table (Extended data30) includes each of these collected indicator iterations across various data sources in more detail. The multiple variations of indicators – with little justification of why they were chosen – posed challenges in summarizing the data, and in recommending how these indicators should be measured. This difficulty was compounded by the absence of information on how indicators were operationally defined, including lack of information on the numerators and denominators in many of the extracted studies. As such, subsequent iterations of similar indicator themes resulted in issues of comparability across jurisdictions. While noting differences in data availability, efforts are needed to systematically define a set of standardized indicators for use across jurisdictions. There have been efforts by some stakeholders to implement accepted performance measurement frameworks98,151. Subsequent efforts can then be aimed at improving the construction and measurement of these indicators, and also in continuing efforts to evaluate system performance.\n\nLimitations specific to the methods and results of this scoping review exist. Indicators were collected based on their ability to be extracted from a population-based data source, dependent on the judgement of study authors and subject matter experts which may have introduced some bias (e.g. large focus on Canadian grey literature). As one of the reasons for this scoping review was to better understand which health system performance indicators currently exist for local efforts of policy planning, a bias towards indicators that can be readily collected in the Canadian and/or Ontario context may exist. Moreover, quality appraisal of indicators was not conducted. Next, the conflation of the terms, palliative, end-of-life, and terminal care was difficult to operationalize when searching and organizing indicators. For the purposes of this review, all terms were included as a part of the search strategy to ensure indicators reflected all references to care related to advanced disease. When reviewing administrative data sources and medical-based records, health service encounters can be coded either or as palliative or end-of-life depending on the care setting and health service provider157,158. Authors however recognize that palliative care reflects a philosophy of care and service that can be provided at any time during the advanced disease trajectory, whereas end-of-life care is not always palliative, and is provided during the final period of a patient’s life. Lastly, this research study answers the question on the types and varieties of indicators available for measurement of performance in PC and EOLC using administrative data. However, it does not answer the question on what these indicators should be based on quality and policy criteria. Next steps would include conducting a consultation with relevant stakeholders to create a universally accepted performance measurement framework of what health system performance measurement for PC and EOLC should look like, using the list of indicators provided in this study as a reference guide. This would include patient-reported outcome measures that have not been included in this study. While some jurisdictions may not be ready to measure many of these indicators, findings from this study can provide insight on what can be potentially measured.\n\nOverall, this study reviews – by health care sector – population-level health system performance indicators for PC and EOLC that can be measured through administrative databases. Although a large number of indicators have been reported for each sector, these indicators are often variations on the same theme, reflecting a lack of consensus on key debated concepts within the PC and EOLC literature. Future work is needed to achieve consensus ‘best’ definitions of these indicators as well as a universal performance measurement framework, similar to other ongoing efforts in population health121. Our scoping review will reduce duplication of the extensive amount of work that is required when a jurisdiction wishes to make a concerted effort to improve care in palliative and EOL populations through adoption of a performance measurement framework; one of the first steps in such efforts is typically to collect performance indicators from literature. This review can instantaneously inform indicator selection and development for other local, national and international efforts currently underway to improve PC and EOLC. Such performance measurement through indicators can help identify gaps in the access and quality of care and evaluate the impact of PC interventions that aim to bridge these gaps.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\nDataverse: Supplementary files for Measuring indicators of health system performance for palliative and end-of-life care using health administrative data: a scoping review, https://doi.org/10.7910/DVN/R3S0RT30.\n\nThis project contains the following extended data:\n\nS1 Appendix: Search Strategy\n\nS1 Table: Detailed List and Information on Collected Indicators\n\nS2 Table: Study Inclusion and Exclusion Criteria\n\nDataverse: PRISMA-ScR checklist for ‘Measuring indicators of health system performance for palliative and end-of-life care using health administrative data: a scoping review’, https://doi.org/10.7910/DVN/R3S0RT30.\n\nData are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).",
"appendix": "Acknowledgements\n\nWe would like to thank Dr. José Pereira, Dr. Christopher Klinger, Dr. Irene Ying, Dr. Michelle Grinman, and Dr. Amy Hsu as subject matter experts for providing guidance on the indicator themes that were important to capture.\n\n\nReferences\n\nCare; OMoHaL-T: Patients First: A roadmap to strengthen home and community care. Ontario. 2015. Reference Source\n\nDe Lima L, Pastrana T: Opportunities for palliative care in public health. Annu Rev Public Health. 2016; 37: 357–74. Free Full Text Publisher Full Text |\n\nWorldwide Palliative Care Alliance, World Health Organization: Global Atlas of Palliative Care at the End of Life. 2014. Reference Source\n\nMorrison RS, Meier DE: America's Care of Serious Illness: 2015 State-By-State Report Card on Access to Palliative Care in our Nation's Hospitals. New York, New York,USA: Centre to Advance Palliative Care; National Palliative Care Research Centre; 2015. Reference Source\n\nLynch T, Connor S, Clark D: Mapping levels of palliative care development: a global update. J Pain Symptom Manage. 2013; 45(6): 1094–106. PubMed Abstract | Publisher Full Text\n\nDixon J, King D, Matosevic T, et al.: Equity in the Provision of Palliative Care in the UK:Review of Evidence. London, United Kingdom: Personal Social Services Research Unit London School of Economics and Political Science; 2015. Reference Source\n\nCanadian Institute for Health Information: CIHI Health Indicators 2017.Reference Source\n\nStatistics Canada: Use of administrative data 2015. Reference Source\n\nQuality Hospice Palliative Care Coalition of Ontario: Advancing High Quality, High Value Palliative Care in Ontario: A Declaration of Partnership and Commitment to Action. 2011. Reference Source\n\nCancer Quality Council of Ontario: Cancer System Quality Index 2016. 2016. Reference Source\n\nOffice of the Auditor General of Ontario: Palliative Care. Ontario, Canada; 2014. Reference Source\n\nOntario Health Technology Advisory Committee (OHTAC) End-of-Life Collaborative, Sikich N, Baidobonsoo S, et al.: Health Care for People Approaching the End of Life: An Evidentiary Framework. Ont Health Technol Assess Ser. 2014; 14(14): 1–45. PubMed Abstract | Free Full Text\n\nLorenz K, Lynn J, Dy S, et al.: Cancer Care Quality Measures: Symptoms and End-of-Life Care. Evid Rep Technol Assess (Full Rep). 2006; (137): 1–77. PubMed Abstract | Free Full Text\n\nNational Institute for Health and Care Excellence: Quality standards for end of life care for adults. 2013. Reference Source\n\nEarle CC, Park ER, Lai B, et al.: Identifying potential indicators of the quality of end-of-life cancer care from administrative data. J Clin Oncol. 2003; 21(6): 1133–8. PubMed Abstract | Publisher Full Text\n\nD'Angelo D, Mastroianni C, Vellone E, et al.: Palliative care quality indicators in Italy. What do we evaluate? Support Care Cancer. 2012; 20(9): 1983–9. PubMed Abstract | Publisher Full Text\n\nRaijmakers N, Galushko M, Domeisen F, et al.: Quality indicators for care of cancer patients in their last days of life: literature update and experts' evaluation. J Palliat Med. 2012; 15(3): 308–16. PubMed Abstract | Publisher Full Text\n\nMcMurchy D, Aggarwal M: Palliative Care in the Community: An Environmental Scan of Frameworks and Indicators. Canadian Hospice Palliative Care Association; 2013. Reference Source\n\nBarbera L, Paszat L, Chartier C: Indicators of poor quality end-of-life cancer care in Ontario. J Palliat Care. 2006; 22(1): 12–7. PubMed Abstract | Publisher Full Text\n\nEarle CC, Neville BA, Landrum MB, et al.: Evaluating claims-based indicators of the intensity of end-of-life cancer care. Int J Qual Health Care. 2005; 17(6): 505–9. PubMed Abstract | Publisher Full Text\n\nGrunfeld E, Lethbridge L, Dewar R, et al.: Towards using administrative databases to measure population-based indicators of quality of end-of-life care: testing the methodology. Palliat Med. 2006; 20(8): 769–77. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrunfeld E, Folkes A, Urquhart R: Do available questionnaires measure the communication factors that patients and families consider important at end of life? J Clin Oncol. 2008; 26(23): 3874–8. PubMed Abstract | Publisher Full Text\n\nBerenholtz SM, Dorman T, Ngo K, et al.: Qualitative review of intensive care unit quality indicators. J Crit Care. 2002; 17(1): 1–12. PubMed Abstract | Publisher Full Text\n\nBeaman R, McGee S, O'Connor D: Developing a Set of Indicators to Assess End-of-Life Care in Massachusetts Nursing Homes. Centre for Health Policy and Research, UMass Medical School; 2006. Reference Source\n\nCanadian Institute for Health Information: Health System Performance Frameworks: Aligning Frameworks for Sectors and Organizations to Health Systems. Ottawa, Canada; 2015. Reference Source\n\nArksey H, O'Malley L: Scoping studies: towards a methodological framework. Int J Soc Res Methodol. 2005; 8(1): 19–32. Publisher Full Text\n\nTricco AC, Lillie E, Zarin W, et al.: PRISMA extension for scoping reviews (PRISMA-ScR): checklist and explanation. Ann Intern Med. 2018; 169(7): 467–73. PubMed Abstract | Publisher Full Text\n\nMunn Z, Peters MDJ, Stern C, et al.: Systematic review or scoping review? Guidance for authors when choosing between a systematic or scoping review approach. BMC Med Res Methodol. 2018; 18(1): 143. PubMed Abstract | Publisher Full Text | Free Full Text\n\nU.S. National Library of Medicine: Medical Subject Headings.2017; [updated 14 November 2017]. Reference Source\n\nBudhwani S, Gill A, Scott M, et al.: Supplementary files for Measuring indicators of health system performance for palliative and end-of-life care using health administrative data: a scoping review, [Dataset]. Dataverse. http://www.doi.org/10.7910/DVN/R3S0RT\n\nStatistics Canada: Population estimates on July 1st, by age and sex. Canada Government of Canada 2019. Publisher Full Text\n\nTanuseputro P, Wodchis WP, Fowler R, et al.: The health care cost of dying: a population-based retrospective cohort study of the last year of life in Ontario, Canada. PLoS One. 2015; 10(3): e0121759. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOntario government: Home and community care Canada Government of Ontario. 2019. Reference Source\n\nOntario government: Get coverage for prescription drugs. Canada, 2020. Reference Source\n\nOntario government: Hospital Chronic Care Co-Payment: Questions and Answers: Government of Ontario. 2019. Reference Source\n\nAlsirafy SA, Hassan AA, Al-Shahri MZ: Hospitalization pattern in a hospital-based palliative care program: an example from Saudi Arabia. Am J Hosp Palliat Care. 2009; 26(1): 52–6. PubMed Abstract | Publisher Full Text\n\nBarbera L, Paszat L, Qiu F: End-of-life care in lung cancer patients in Ontario: aggressiveness of care in the population and a description of hospital admissions. J Pain Symptom Manage. 2008; 35(3): 267–74. PubMed Abstract | Publisher Full Text\n\nBurge F, Lawson B, Johnston G: Family physician continuity of care and emergency department use in end-of-life cancer care. Med Care. 2003; 41(8): 992–1001. PubMed Abstract | Publisher Full Text\n\nCanadian Cancer Society's Steering Committee: Canadian Cancer Statistics 2010. Toronto: Canadian Cancer Society; 2010. Reference Source\n\nCanadian Institute for Health Information: Health Care Use at the End of Life in Manitoba. Ottawa: Canadian Institute for Health Information; 2008. Reference Source\n\nCancer Care Ontario: Chronic Care Application of the Cancer Care Ontario Palliative Care Program's Models and Products. Cancer Care Ontario. Reference Source\n\nCancer Quality Council of Ontario: Patterns of End-of-Life Care. Cancer Care Ontario; 2007.\n\nComptroller and Auditor General: End of Life Care. In: Office. NA, editor. Norwich, UK, 2008.\n\nEthunandan M, Rennie A, Hoffman G, et al.: Quality of dying in head and neck cancer patients: a retrospective analysis of potential indicators of care. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2005; 100(2): 147–52. PubMed Abstract | Publisher Full Text\n\nTeno JM, Gozalo PL, Bynum JP, et al.: Change in end-of-life care for Medicare beneficiaries: site of death, place of care, and health care transitions in 2000, 2005, and 2009. JAMA. 2013; 309(5): 470–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCanadian Institute for Health Information: Health Care Use at the End of Life in Western Canada. Ottawa: Canadian Institute for Health Information; 2007. Reference Source\n\nDy SM, Lee A, Lashoher A: End-of-life health care utilization in hospitals with compared to those without palliative care programs. J Palliat Med. 2014; 17(8): 877–8. PubMed Abstract | Publisher Full Text\n\nEarle CC, Neville BA, Landrum MB, et al.: Trends in the aggressiveness of cancer care near the end of life. J Clin Oncol. 2004; 22(2): 315–21. PubMed Abstract | Publisher Full Text\n\nGonsalves WI, Tashi T, Krishnamurthy J, et al.: Effect of palliative care services on the aggressiveness of end-of-life care in the veteran's affairs cancer population. J Palliat Med. 2011; 14(11): 1231–5. PubMed Abstract | Publisher Full Text\n\nHu W, Yasui Y, White J, et al.: Aggressiveness of end-of-life care for patients with colorectal cancer in Alberta, Canada: 2006-2009. J Pain Symptom Manage. 2014; 47(2): 231–44. PubMed Abstract | Publisher Full Text\n\nLau F, Downing M, Tayler C, et al.: Toward a population-based approach to end-of-life care surveillance in Canada: initial efforts and lessons. J Palliat Care. 2013; 29(1): 13–21. PubMed Abstract | Publisher Full Text\n\nMittmann N, Liu N, Porter JM, et al.: End-of-life home care utilization and costs in patients with advanced colorectal cancer. J Community Support Oncol. 2014; 12(3): 92–8. PubMed Abstract | Publisher Full Text\n\nMizukawa S: [A survey on the view of end-of-life care in the elderly -a comparison among elderly patients, family members. physicians, nurses, and other caregivers]. Nihon Ronen Igakkai Zasshi. 2008; 45(1): 50–8. PubMed Abstract | Publisher Full Text\n\nTang ST, Wu SC, Hung YN, et al.: Trends in quality of end-of-life care for Taiwanese cancer patients who died in 2000-2006. Ann Oncol. 2009; 20(2): 343–8. PubMed Abstract | Publisher Full Text\n\nTemkin-Greener H, Zheng NT, Mukamel DB: Rural-urban differences in end-of-life nursing home care: facility and environmental factors. Gerontologist. 2012; 52(3): 335–44. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCanadian Institute for Health Information: End-of-Life Hospital Care for Cancer Patients. Ottawa: Canadian Institute for Health Information; 2013. Reference Source\n\nCanadian Partnership Against Cancer: Cancer System Performance 2017 Report. 2017. Reference Source\n\nErie St. Clair Hospice Palliative Care Network: Hospice Palliative Care System - Examples of Data elements and indicators by sector/service.\n\nGoodman DC, Esty AR, Fisher ES, et al.: Trends and Variation in End-of-Life Care for Medicare Bene ciaries with Severe Chronic Illness. The Dartmouth Institute for Health Policy and Clinical Practice; 2011. Reference Source\n\nGoodman DC, Fisher ES, Chang CH, et al.: Quality of End-of-Life Cancer Care for Medicare Beneficiaries: Regional and Hospital-Specific Analyses. The Dartmouth Institute for Health Policy & Clinical Practice; 2010. Reference Source\n\nGrunfeld E, Urquhart R, Mykhalovskiy E, et al.: Toward population-based indicators of quality end-of-life care: Testing stakeholder agreement. Cancer. 2008; 112(10): 2301–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHealth Improvement Scotland: Palliative and end of life care: Indicators. In: Scotland NHS, editor. 2013. Reference Source\n\nLiu TW, Chen JS, Wang HM, et al.: Quality of end-of-life care between medical oncologists and other physician specialists for Taiwanese cancer decedents, 2001-2006. Oncologist. 2009; 14(12): 1232–41. PubMed Abstract | Publisher Full Text\n\nNational Quality Forum: National Voluntary Consensus Standards for Quality of Cancer Care: A Consensus Report. Washington DC, USA: National Quality Forum; 2009. Reference Source\n\nNetwork for End-of-Life Studies (NELS) Interdisciplinary Care Enhancement (ICE): End of Life Care in Nova Scotia: Surveillance Report. Halifax, Nova Scotia, Canada: Dalhousie University; 2008. Reference Source\n\nOntario Ministry of Health and Long-Term Care - Residential Hospice Working Group: Environmental Scan for Strengthening Residential Hospice Care in Ontario. 2015. Reference Source\n\nPalliative Care Quality Indicators Working Group: Palliative Care Quality Indicator Measures Ontario. In: Tanuseputro P, editor. 2014.\n\nQuality Hospice Palliative Care Coalition of Ontario: Creating an Integrated Hospice Palliative Care System in Ontario: Setting the Stage for Change. 2010. Reference Source\n\nSetoguchi S, Glynn RJ, Stedman M, et al.: Hospice, opiates, and acute care service use among the elderly before death from heart failure or cancer. Am Heart J. 2010; 160(1): 139–44. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarbera L, Seow H, Sutradhar R, et al.: Quality of end-of-life cancer care in Canada: a retrospective four-province study using administrative health care data. Current Oncology. 2015; 22(5): 341. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarbera L, Seow H, Sutradhar R, et al.: Quality Indicators of End-of-Life Care in Patients With Cancer: What Rate Is Right? J Oncol Pract. 2015; 11(3): e279–e87. PubMed Abstract | Publisher Full Text\n\nCanadian Partnership Against Cancer: Palliative and end-of-life care. Toronto, ON: Canadian Partnership Against Cancer; 2017. Reference Source\n\nChiang JK, Lee YC, Kao YH: Trend analysis of end-of-life care between hospice and nonhospice groups of cancer patients in Taiwan for 2002-11. Medicine (Baltimore). 2017; 96(34): e7825. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCurtis JR, Sathitratanacheewin S, Starks H, et al.: Using Electronic Health Records for Quality Measurement and Accountability in Care of the Seriously Ill: Opportunities and Challenges. J Palliat Med. 2018; 21(S2): S52–S60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDe Roo ML, Miccinesi G, Onwuteaka-Philipsen BD, et al.: Actual and preferred place of death of home-dwelling patients in four European countries: making sense of quality indicators. PLoS One. 2014; 9(4): e93762. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDy SM, Walling AM, Mack JW, et al.: Evaluating the quality of supportive oncology using patient-reported data. J Oncol Pract. 2014; 10(4): e223–e30. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGourin CG, Frick KD, Blackford AL, et al.: Quality indicators of laryngeal cancer care in the elderly. Laryngoscope. 2014; 124(9): 2049–56. PubMed Abstract | Publisher Full Text\n\nSanders JJ, Curtis JR, Tulsky JA: Achieving Goal-Concordant Care: A Conceptual Model and Approach to Measuring Serious Illness Communication and Its Impact. J Palliat Med. 2018; 21(S2): S17–S27. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKao YH, Chiang JK: Effect of hospice care on quality indicators of end-of-life care among patients with liver cancer: a national longitudinal population-based study in Taiwan 2000-2011. BMC palliative care. 2015; 14(1): 39. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKrzyzanowska M, Barbera L, Elit L, et al.: Cancer-Chapter 4. 2009. Reference Source\n\nAslakson RA, Bridges JFP: Assessing the impact of palliative care in the intensive care unit through the lens of patient-centered outcomes research. Curr Opin Crit Care. 2013; 19(5): 504–10. PubMed Abstract | Publisher Full Text\n\nSessa C, Roggero E, Pampallona S, et al.: The last 3 months of life of cancer patients: medical aspects and role of home-care services in southern Switzerland. Support Care Cancer. 1996; 4(3): 180–5. PubMed Abstract | Publisher Full Text\n\nJakobsson E, Bergh I, Ohlen J, et al.: Utilization of health-care services at the end-of-life. Health Policy. 2007; 82(3): 276–87. PubMed Abstract | Publisher Full Text\n\nMiyashita M, Nakamura A, Morita T, et al.: Identification of quality indicators of end-of-life cancer care from medical chart review using a modified Delphi method in Japan. Am J Hosp Palliat Care. 2008; 25(1): 33–8. PubMed Abstract | Publisher Full Text\n\nCanadian Institute for Health Information: Health care use at the End of Life in Atlantic Canada. Canadian Institute for Health Information; 2011. Reference Source\n\nAlberta Health Services: Edmonton Zone Palliative Care Program Balanced Scorecard Report 2010-2011. 2012. Reference Source\n\nCanadian Institute for Health Information: Health care use at the end of life in British Columbia. Ottawa: CIHI; 2008. Reference Source\n\nDy SM, Wolff JL, Frick KD: Patient characteristics and end-of-life health care utilization among Medicare beneficiaries in 1989 and 1999. Med Care. 2007; 45(10): 926–30. PubMed Abstract | Publisher Full Text\n\nGoodridge DM, Marciniuk DD, Brooks D, et al.: End-of-life care for persons with advanced chronic obstructive pulmonary disease: report of a national interdisciplinary consensus meeting. Can Respir J. 2009; 16(5): e51–3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGreenberg A, Angus H, Sullivan T, et al.: Development of a set of strategy-based system-level cancer care performance indicators in Ontario, Canada. Int J Qual Health Care. 2005; 17(2): 107–14. PubMed Abstract | Publisher Full Text\n\nGrunfeld E, McIntyre P: Quality Indicators for end of life care: A comparison of elderly nursing home residents and non-residents dying of cancer in Nova Scotia, 2000-2003. NELS ICE Project 6 Update. 2008.\n\nQuality Hospice Palliative Care Coalition of Ontario: Hospice Palliative Care Data and Performance Measurement Subcommittee Techhnical Specificiations Working Group: Hospital Palliative Care Indicator Dictionary v7 &Technical Specs v.7, Appendix. In: Tanuseputro P, editor. Toronto. 2014.\n\nShugarman LR, Bird CE, Schuster CR, et al.: Age and gender differences in medicare expenditures and service utilization at the end of life for lung cancer decedents. Womens Health Issues. 2008; 18(3): 199–209. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWarren JL, Barbera L, Bremner KE, et al.: End-of-life care for lung cancer patients in the United States and Ontario. J Natl Cancer Inst. 2011; 103(11): 853–62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTanuseputro P, Budhwani S, Bai YQ, et al.: Understanding the Provision of End-of-Life and Palliative Care Services in Ontario. Health System Performance Research Network. 2013.\n\nTanuseputro P, Budhwani S, Bai YQ, et al.: Palliative Care in Ontario: A Population-Based Study across Health Sectors using Health Administrative Databases. Health System Performance Research Network. 2016.\n\nTanuseputro P, Budhwani S, Bai YQ, et al.: Palliative care delivery across health sectors: A population-level observational study. Palliat Med. 2017; 31(3): 247–257. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDe Schreye R, Houttekier D, Deliens L, et al.: Developing indicators of appropriate and inappropriate end-of-life care in people with Alzheimer’s disease, cancer or chronic obstructive pulmonary disease for population-level administrative databases: A RAND/UCLA appropriateness study. Palliat Med. 2017; 31(10): 932–45. PubMed Abstract | Publisher Full Text\n\nBrumley R, Enguidanos S, Jamison P, et al.: Increased satisfaction with care and lower costs: results of a randomized trial of in-home palliative care. J Am Geriatr Soc. 2007; 55(7): 993–1000. PubMed Abstract | Publisher Full Text\n\nCare Use at the End of Life in Saskatchewan. Ottawa: Canadian Institute for Health Information. 2008.\n\nCanadian Institute for Health Information: Health care use at the end of life in Saskatchewan. Ottawa: CIHI. 2008.\n\nNeuberg GW: The cost of end-of-life care: a new efficiency measure falls short of AHA/ACC standards. Circ Cardiovasc Qual Outcomes. 2009; 2(2): 127–33. PubMed Abstract | Publisher Full Text\n\nCanadian Partnership Against Cancer: The 2015 Cancer System Performance Report. Toronto, Ontario, Canada: Canadian Partnership Against Cancer. 2015.\n\nHodgson C: Cost-Effectiveness of Palliative Care: A Review of the Literature. Canadian Hospice Palliative Care Association. 2011. Reference Source\n\nMeloche J, Urowitz S, editors.: Hospice Palliative Care Data & Performance Subcommittee Performance Measurement Strategy. Canadian Association for Health Services & Policy Research; Toronto, Ontario, Canada. 2014.\n\nMack JW, Cronin A, Keating NL, et al.: Associations between end-of-life discussion characteristics and care received near death: a prospective cohort study. J Clin Oncol. 2012; 30(35): 4387. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLind S, Adolfsson J, Axelsson B, et al.: Quality indicators for palliative and end of life care: a review of Swedish policy documents. BMJ Support Palliat Care. 2015; 5(4): 413–9. PubMed Abstract | Publisher Full Text\n\nHuang J, Boyd C, Tyldesley S, et al.: Time spent in hospital in the last six months of life in patients who died of cancer in Ontario. J Clin Oncol. 2002; 20(6): 1584–92. PubMed Abstract | Publisher Full Text\n\nLewin SN, Buttin BM, Powell MA, et al.: Resource utilization for ovarian cancer patients at the end of life: how much is too much? Gynecol Oncol. 2005; 99(2): 261–6. PubMed Abstract | Publisher Full Text\n\nBarbera L, Elit L, Krzyzanowska M, et al.: End of life care for women with gynecologic cancers. Gynecol Oncol. 2010; 118(2): 196–201. PubMed Abstract | Publisher Full Text\n\nBergman J, Chi AC, Litwin MS: Quality of end-of-life care in low-income, uninsured men dying of prostate cancer. Cancer. 2010; 116(9): 2126–31. PubMed Abstract | Publisher Full Text\n\nHo TH, Barbera L, Saskin R, et al.: Trends in the aggressiveness of end-of-life cancer care in the universal health care system of Ontario, Canada. J Clin Oncol. 2011; 29(12): 1587–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKapadia NS, Mamet R, Zornosa C, et al.: Radiation therapy at the end of life in patients with incurable nonsmall cell lung cancer. Cancer. 2012; 118(17): 4339–45. PubMed Abstract | Publisher Full Text\n\nMorden NE, Chang CH, Jacobson JO, et al.: End-of-life care for Medicare beneficiaries with cancer is highly intensive overall and varies widely. Health Aff (Millwood). 2012; 31(4): 786–96. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorishima T, Lee J, Otsubo T, et al.: Impact of hospital case volume on quality of end-of-life care in terminal cancer patients. J Palliat Med. 2013; 16(2): 173–8. PubMed Abstract | Publisher Full Text\n\nPorter J, Earle C, Atzema C, et al.: Initiation of Chemotherapy in Cancer Patients with Poor Performance Status: A population-based analysis. J Palliat Care. 2014; 30(3): 166–72. PubMed Abstract | Publisher Full Text\n\nHuang J, Wai ES, Lau F, et al.: Palliative radiotherapy utilization for cancer patients at end of life in British Columbia: retrospective cohort study. BMC palliative care. 2014; 13(1): 49. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLangton JM, Blanch B, Drew AK, et al.: Retrospective studies of end-of-life resource utilization and costs in cancer care using health administrative data: a systematic review. Palliat Med. 2014; 28(10): 1167–96. PubMed Abstract | Publisher Full Text\n\nGoodman DC, Esty AR, Fisher ES, et al.: Trends and Variation in End-of-Life Care for Medicare Beneficiaries with Severe Chronic Illness. The Dartmouth Institute for Health Policy and Clinical Practice Center for Health Policy Research. 2011. Reference Source\n\nChoi JY, Shin DW, Kang J, et al.: Variations in process and outcome in inpatient palliative care services in Korea. Support Care Cancer. 2012; 20(3): 539–47. PubMed Abstract | Publisher Full Text\n\nHealth Quality Ontario: Palliative care at the end of life. 2016.\n\nBudhwani S: Using 'Death at Home' as a Health System Performance Outcome Measure - A Critical Analysis. [Final Paper for Course Age1001H]. In press. 2014.\n\nHughes SL, Cummings J, Weaver F, et al.: A randomized trial of the cost effectiveness of VA hospital-based home care for the terminally ill. Health Serv Res. 1992; 26(6): 801–7. PubMed Abstract | Free Full Text\n\nSutradhar R, Atzema C, Seow H, et al.: Is performance status associated with symptom scores? A population-based longitudinal study among cancer outpatients. J Palliat Care. 2014; 30(2): 99–107. PubMed Abstract | Publisher Full Text\n\nSaskatchewan Cancer Agency: Saskatchewan Cancer Agency Performance Dashboard. 2012.\n\nDy SM, Lorenz KA, Oneill SM, et al.: Cancer Quality-ASSIST supportive oncology quality indicator set: feasibility, reliability, and validity testing. Cancer. 2010; 116(13): 3267–75. PubMed Abstract | Publisher Full Text\n\nNeuss MN, Desch CE, McNiff KK, et al.: A process for measuring the quality of cancer care: the Quality Oncology Practice Initiative. J Clin Oncol. 2005; 23(25): 6233–9. PubMed Abstract | Publisher Full Text\n\nSeow H, Barbera L, Sutradhar R, et al.: Trajectory of performance status and symptom scores for patients with cancer during the last six months of life. J Clin Oncol. 2011; 29(9): 1151–8. PubMed Abstract | Publisher Full Text\n\nDeCato TW, Engelberg RA, Downey L, et al.: Hospital variation and temporal trends in palliative and end-of-life care in the ICU. Crit Care Med. 2013; 41(6): 1405–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAllan DE, Stajduhar KI, Reid RC: The uses of provincial administrative health databases for research on palliative care: Insights from British Columbia, Canada. BMC palliative care. 2005; 4(1): 2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarbera L, Sussman J, Viola R, et al.: Factors associated with end-of-life health service use in patients dying of cancer. Healthc Policy. 2010; 5(3): e125–e43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRizzi F, Pizzuto M, Lodetti L, et al.: Quality for home palliative care: an Italian metropolitan multicentre JCI-certified model. BMJ Qual Saf. 2011; 20(7): 592–8. PubMed Abstract | Publisher Full Text\n\nWalker H, Anderson M, Farahati F, et al.: Resource use and costs of end-of-Life/palliative care: Ontario adult cancer patients dying during 2002 and 2003. J Palliat Care. 2011; 27(2): 79–88. PubMed Abstract | Publisher Full Text\n\nSeow H, Barbera L, Howell D, et al.: Using more end-of-life homecare services is associated with using fewer acute care services: a population-based cohort study. Med Care. 2010; 48(2): 118–24. PubMed Abstract | Publisher Full Text\n\nHealth Quality Ontario: Palliative care at the end of life. Toronto, ON: Health Quality Ontario; 2016.\n\nCancer Care Ontario: Ontario Cancer Registry - Cancer Care Ontario. Reference Source\n\nGlavan BJ, Engelberg RA, Downey L, et al.: Using the medical record to evaluate the quality of end-of-life care in the intensive care unit. Crit Care Med. 2008; 36(4): 1138–46. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmith T: Using the interRAI Palliative Care to assist with client - centred care: examples from Ontario, Canada. 1st Flanders Study to Improve End of Life Care & Evaluation Tools Conference on Palliative Care. Leuven, Belgium. 2013.\n\nReich O, Signorell A, Busato A: Place of death and health care utilization for people in the last 6 months of life in Switzerland: a retrospective analysis using administrative data. BMC Health Serv Res. 2013; 13: 116. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHealth Quality Ontario: Quality Indicators for Measuring the Ontario Health Technology Advisory Committee (OHTAC) Recommendations for End-of-Life Care. 2015.\n\nGozalo P, Teno JM, Mitchell SL, et al.: End-of-life transitions among nursing home residents with cognitive issues. N Engl J Med. 2011; 365(13): 1212–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDi Giulio P, Toscani F, Villani D, et al.: Dying with advanced dementia in long-term care geriatric institutions: a retrospective study. J Palliat Med. 2008; 11(7): 1023–8. PubMed Abstract | Publisher Full Text\n\nKeay TJ, Alexander C, McNally K, et al.: Nursing home physician educational intervention improves end-of-life outcomes. J Palliat Med. 2003; 6(2): 205–13. PubMed Abstract | Publisher Full Text\n\nHwang SJ, Chang HT, Hwang IH, et al.: Hospice offers more palliative care but costs less than usual care for terminal geriatric hepatocellular carcinoma patients: a nationwide study. J Palliat Med. 2013; 16(7): 780–5. PubMed Abstract | Publisher Full Text\n\nCenters for Medicare & Medicaid Services: Hospice Quality Reporting Program Current Measures USA2017. [updated Nov 14, 2017]. Reference Source\n\nCanadian Institute for Health Information: A Performance Measurement Framework for the Canadian Health System. Ottawa, Canada; 2013. Reference Source\n\nMiccinesi G, Crocetti E, Morino P, et al.: Palliative home care reduces time spent in hospital wards: a population-based study in the Tuscany Region, Italy. Cancer Causes Control. 2003; 14(10): 971–7. PubMed Abstract | Publisher Full Text\n\nMiller A: Issues Related to Utilization of Health Care at the End of Life. Medicare Patient Management. 2008; 27–30.\n\nNational Institute for Health and Care Excellence: End of Life Care Quality Assessment Tool. United Kingdom2013 [updated October 2013].\n\nDe Roo ML, Francke AL, Van den Block L, et al.: Hospitalizations of cancer patients in the last month of life: quality indicator scores reveal large variation between four European countries in a mortality follow-back study. BMC Palliat Care. 2014; 13(1): 54. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKelley AS, Bollens-Lund E: Identifying the Population with Serious Illness: The \"Denominator\" Challenge. J Palliat Med. 2018; 21(S2): S7–S16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAggarwal G: The Effect of the Palliative Care Team on Quality Indicators for End-of-life Care in the Intensive Care Unit: a randomised, controlled study. [Presentation]. In press.\n\nEvers MM, Purohit D, Perl D, et al.: Palliative and aggressive end-of-life care for patients with dementia. Psychiatr Serv. 2002; 53(5): 609–13. PubMed Abstract | Publisher Full Text\n\nDe Lima L, Krakauer EL, Lorenz K, et al.: Ensuring palliative medicine availability: the development of the IAHPC list of essential medicines for palliative care. J Pain Symptom Manage. 2007; 33(5): 521–6. PubMed Abstract | Publisher Full Text\n\nWalling AM, Asch SM, Lorenz KA, et al.: The quality of supportive care among inpatients dying with advanced cancer. Support Care Cancer. 2012; 20(9): 2189–94. PubMed Abstract | Publisher Full Text\n\nGiovanis P, De Leonardis G, Garna A, et al.: Clinical governance benchmarking issues in oncology: aggressiveness of cancer care and consumption of strong opioids. A single-center experience on measurement of quality of care. Tumori. 2010; 96(3): 443–7. PubMed Abstract | Publisher Full Text\n\nValuck TMR, Montgomery R: Data Needs for a Serious Illness Care Accountability System: A Framework and Recommendations. J Palliat Med. 2018; 21(S2): S1–S6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPrice AR, Elliott MN: Measuring Patient-Centeredness of Care for Seriously Ill Individuals: Challenges and Opportunities for Accountability Initiatives. J Palliat Med. 2018; 21(S2): S28–S35. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "80857",
"date": "06 Apr 2021",
"name": "Catherine J Evans",
"expertise": [
"Reviewer Expertise Research on palliative care delivery",
"new models and evaluation and use of outcome measures in routine clinical care."
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThank you, this is an interesting and comprehensive scoping review on health system quality/performance indicators for palliative care and end of life care. The scope of the review is impressive in aiming to encompass all settings and conditions, and inclusion and searching of grey literature. As the authors conclude, the findings are important to inform quality indicator selection for commissioners and policy makers and acknowledge important limitations that the review identifies the types and variety of indicators available, but not what indicators should form criteria for quality and policy makers.\nCreation of a universally accepted performance framework for palliative and end of life care is identified as a priority research area, developed in consultation with key stakeholders. Patients and families would be key stakeholders, with country context specific consideration for low, middle, and high income countries.\nThe review identifies 339 indicators. The exploration of these by respective setting to identify common categories is important to distinguish key quality indicators by respective settings, convergence and gaps. The themes identified by respective settings are helpful to consider key indicators for respective settings.\nAs the authors discuss, the main area of convergence across settings is measurement of health service use towards the end of life as a quality indicator and identification of gaps of patient level indicators such as patient experience, and outcome data such as symptom distress. The authors call for use of use of outcome measures in routine care to address this. Helpful for the authors to refer to examples of this approach at a national level, such as the:\nPalliative Care Outcomes Collaborative (PCOC) in Australia https://www.uow.edu.au/ahsri/pcoc/ and publications on embedding objective measures of quality in routine care1.\n\nProgression of this work in the UK in the Outcome Assessment and Complexity Collaborative suite of measures for use in palliative and end of life at King’s College London https://www.arc-sl.nihr.ac.uk/research-and-implementation/our-research-areas/palliative-and-end-life-care/outcomes-health-and, https://www.kcl.ac.uk/cicelysaunders/attachments/Studies-OACC-Brief-Introduction-Booklet.pdf and Hull York Medical School https://www.hyms.ac.uk/research/research-centres-and-groups/wolfson/resolve\nI have a few main points about the robustness of the scoping review:\nThe scoping review is well defined in terms of the rationale and methods, and concepts of quality indicators and heath systems. However, important to have defined the concepts of palliative care and end of life care. This is eluded to as a limitation. The paper would have been strengthened to have defined these concepts in the methods, to enable consistent application of this eligibility criteria\n\nAlthough the search is comprehensive, the last search was in July 2018. This is a fast moving area of health care. As such, there is an expectation of a search in the last 12 months as there is no justification to not updated for nearly three years.\n\nAs a scoping review, it is important to identify other systematic reviews on this topic area to cite in the background to justify the review and/or refer to in the discussion. For example, Henderson et al. (2019)2\n\nMethods:\nSearch strategy – the abstract states searched Medline and EMBASE, but the supplementary file details Pubmed and EMBASE. Please correct to give correct reporting of the databases searchers.\n\nSearch terms for palliative care – this would have been strengthened by using/drawing on Sladek’s 2006 JMLA search filter for studies on palliative care3.\n\nSentence on \"Studies outside of the study time period of 2010-2018...\" is confusing. It implies that publications were identified after 2018. When what reporting is the use of reference chaining to identify publications prior to 2010. This needs to be updated for clarity.\n\nFigure 1 needs to align with PRISMA reporting to include number of studies identified by respective sources e.g. data bases searched, grey literature, screening at title and abstract and number excluded, and screening at full paper and reasons for exclusion against the eligibility criteria. The PRISMA for scoping reviews checklist is included as a supplementary file, but this can’t be opened. Please review to ensure full reporting on the identification and screening for eligible publications.\n\nDetail on the analysis to form the themes for the indicators is very brief. Important to understand the process, how were disagreements managed, how ensured robust process across the 339 indicators.\n\nResults – tables interesting and clear. Supplementary table interesting re-reporting on nominator and denominator, and how rarely presented.\n\nNo quality appraisal of the included studies, or indicators. This limits consideration on the quality of the included studies, and the indicators identified and for policy makers to discern between the indicators identified. This needs to be detailed in the limitations, for conclusions to be considered from this crucial limitation of the study.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "95959",
"date": "28 Oct 2021",
"name": "Joachim Cohen",
"expertise": [
"Reviewer Expertise Palliative care",
"Public health"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study provides a useful and welcomed overview of EOLC quality indicators that have been developed or used in published studies. Overall the presentation is clear, but we have a number of comments and suggestions the authors may wish to address.\n\nRationale for the Canadian perspective versus the conclusions about comparability across jurisdictions: the emphasis on Canada is strongly integrated in the article, both in the choices for the classifications of health sectors used to classify the indicators and in the analysis criterion for “conceivably measurable using (Ontario/Canadian) health administrative data”. That seems to be somewhat in conflict with the conclusions and plea the authors seem to make about comparability across jurisdictions.\n\nWe endorse the call for cross-national comparability. However, we would welcome a bit more reflection about the plea for more comparability across jurisdictions. There are likely several reasons for variations, including data availability but also different discourses around quality care. It is not so unsurprising then that experts within one jurisdiction face-validate indicators in a different manner than those in another. What would be needed to develop these? What rigorous efforts, using what methods, would the authors suggest for such a process? Our own experiences in cross-national comparisons learn us that issues around measurement equivalence, but also conceptual equivalence are a huge challenge.\n\nAims: the aim includes measurement concepts AND indicators, whereas the results talk about indicators. This creates confusion about the concept of ‘measurement concept’ and how it differs from the indicators. Either it needs some operationalization, or, alternatively, you could remove it from the aim.\n\nMethods: literature search: the search strategy and inclusion/exclusion criteria were well done, are provided in an online format and were described precisely. It is easy to access the files through the citation to the data repository. However we have three comments: First, why is the search limited to July 2018? As this is a quickly expanding field this seems like an important limitation. What is the risk of missing important recent development efforts? Second, no validation strategy for the search string was followed and this could be mentioned as a limitation. Third, inclusion criteria for study selection are OK but there is no mention for the criteria for indicator selection. Authors remove 722 indicators for irrelevancy, but based on what criteria? Reasons for removal would also be useful in the flowchart.\n\nResults: Quality assessment: the authors did not do a formal quality assessment for the studies being reviewed. Knowing the heterogeneity of included studies/literature, we can understand the practical difficulties in using existing quality assessment tools. However, as a reader we would benefit from at least some reflection or discussion about the methodological quality and rigor in indicator selection (eg convenience selection vs. formal validation efforts? What methods are used?). It would also be very helpful to have an overview of the level of scientific evidence underlying an indicator. It is likely that for most indicators the evidence-level is expert opinion and that only a limited number are based on evidence of causal impact on quality of life or related concepts. As such an overview would probably present demand a huge undertaking it is perhaps something that the authors may wish to stipulate as an attention point for future research?\n\nDiscussion: There is quite a strong emphasis on indicators being sector-specific. Not sure if this is accurate as the idea is not to use these indicators within a certain sector. We believe that for many of the indicators there is a system-wide responsibility to assess and improve quality. For instance, referral to home care is the responsibility of not only the actors within home care but also those outside that sector. Careful not to suggest an even stronger echeloning of health care.\n\nDiscussion: The authors focus on the fact that the measurement of patient-centered or patient-reported outcomes was infrequent at the population-level using health administrative database. This is an important point but we think it may be important to provide readers with more discussion about why this is so (not routinely collected?) and what would be needed to move forward to allow for inclusion of such indicators.\n\nDiscussion: strengths and limitations: would be good to have a separate paragraph discussing these. Also, some of the limitation mentioned in the points above may need to be included.\n\nA minor point about the classifications: it seems that you approached the classification of indicators both deductively (i.e. using Ontario’s health sector classification) and inductively (i.e. based on the types of indicators you found. The inductive approach is a result from your review (and should also be presented as such) and we would welcome more insight about how you came to the classification.\n\nSpecific sentences:\n\n“One of the most commonly occurring measurement themes across sectors focused on health system utilization (…)\" --> health system utilization could be used as a term to describe most of the other indicators as well. Consider a more specific term.\n\n!” As health system resources become finite (…)” > Have they ever been infinite?\n\nDiscretionary: a visual summary of the themes would make the results more comprehensible. Similarly, the ‘description’ column of table 3 is a bit unclear, it might be clearer to give some examples of specific indicators.\n\nSmall language issues: adequately use end-of-life (adjective) versus end of life (noun); same for consistent use of population level vs population-level.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1438
|
https://f1000research.com/articles/9-182/v1
|
11 Mar 20
|
{
"type": "Research Article",
"title": "Effects of polyisoprenoids from Avicennia lanata and Avicennia alba leaves on the gene expression of PI3K, Akt1, mTOR, P53, and EGFR in human colorectal adenocarcinoma WiDr cells using reverse transcription-PCR",
"authors": [
"Taufiq Qurrohman",
"Poppy Anjelisa Zaitun Hasibuan",
"Arif Nuryawan",
"Sumaiyah Sumaiyah",
"Etti Sartina Siregar",
"Mohammad Basyuni",
"Taufiq Qurrohman",
"Poppy Anjelisa Zaitun Hasibuan",
"Arif Nuryawan",
"Sumaiyah Sumaiyah",
"Etti Sartina Siregar"
],
"abstract": "Background: Mangrove plants produce a polyisoprenoid compound. Polyisoprenoids have been proven to have anticancer properties. This study investigated the inhibitory activity of polyisoprenoids derived from the leaves of mangrove plants Avicennia alba and Avicennia lanata regarding the expression of PI3K, Akt1, mTOR, P53, and EGFR genes against human colorectal adenocarcinoma WiDr cells. Methods: Anticancer activity was tested through the MTT assay method performed on WiDr cells. The inhibited cell cycle and apoptosis were analysed by flow cytometry and double staining. Gene expression of PI3K, Akt1, mTOR, P53, and EGFR was observed using the RT-PCR method. Results: Cytotoxic activity against WiDr cells showed that the IC50 for A. alba and A. lanata was 258.14 ug/mL and 243.32 ug/mL, respectively. This indicated that their classification as anticancer agents was moderate. The cell cycle showed that inhibition of A. alba and A. lanata occurred in the late phase of apoptosis S (10.60 and 10.51%) and G2-M1 (22.05 and 23.84%), which was higher than negative and positive control cells. Furthermore, the polyisoprenoids derived from A. alba and A. lanata leaves exhibited anticancer activity in WiDr cells through the downregulated gene expression of PI3K, Akt1, mTOR, and EGFR as well as the upregulated gene expression of P53. Conclusion: This study demonstrated that polyisoprenoids obtained from A. alba and A. lanata leaves are promising chemopreventive agents for colon cancer.",
"keywords": [
"Avicennia",
"cytotoxic",
"polyisoprenoids",
"colon cancer",
"mangrove"
],
"content": "Introduction\n\nCancer is a disease characterised by uncontrolled cell growth. Cancer cells can evade apoptosis and avoid signals that suppress its growth, impede the ability to form new blood vessels (angiogenesis), and halt its invasion and metastasis1. According to the Global Cancer Observatory, in 2018, Asia had the highest incidence of colon cancer with 51.8% of the global cases. Colon cancer is one of the top three causes of death in the world1. The use of chemotherapeutic agents constitutes a treatment for colon cancer, in addition to surgery and radiation therapy. Chemotherapeutic agents generally suppress the growth or proliferation of cancer cells, simultaneously causing toxicity in the body2.\n\nNatural ingredients developed as potential chemotherapeutic agents include mangrove leaves. Mangroves are vegetation formations found in littoral areas in tropics and subtropics3. Polyisoprenoids are secondary metabolites found in several mangroves, distributed as dolichol and polyprenol on the leaves and roots of mangrove plants4. So far, few studies have reported the pharmacological activity of polyisoprenoids obtained from mangrove species. Thus, it is essential to study the potential and mechanisms of polyisoprenoids in mangroves as a natural ingredient for anticancer pharmaceuticals and medication4. For instance, methanol extracts of Avicennia alba (bark and leaves) present anti-proliferative activity in MCF-7 and T47D5. Additionally, this extract has cytotoxic effects on a variety of cancer cells, including colon cancer cells – HT-296. Previous research suggests that polyisoprenoids induce cancer cell cycle inhibition in adenocarcinoma of the colon (COLO 320 HSR, WiDr, and LS174 cells) in the G2/M phase and reduce the percentage of Bcl-2 and Bcl-xL7. Polyisoprenoids have been previously reported as chemopreventive agents for colon cancer8, given that polyisoprenoids in A. lanata leaves have displayed anticancer activity for the same9. On the other hand, A. alba contains polyisoprenoids that induce cell cycle, apoptosis, and gene expression of COX-2 in colon cancer cells WiDr10. This extract has a mechanism for inhibiting the cell cycle at the G0-G1 phase, and apoptotic analysis occurs in the early phase of apoptosis in WiDr cells9,10.\n\nThe present study analysed the effect of immune-related genes’ expression on WiDr cells in vitro using reverse transcription-polymerase chain reaction (RT-PCR). RT-PCR was developed as an in vitro test to measure the biological activity of plasmid DNA-based products (pDNA). The said test measures RNA-specific transgenic messengers (mRNA) derived from transfected cultured cells. Forward and reverse primers have been designed to trigger selective RT-PCR reactions for plasmid mRNAs and differentiate between the level of individual plasmid expression in multivalent pDNA11. Therefore, the present study aims to investigate the inhibitory activity of polyisoprenoids obtained from the leaves of mangrove plants A. alba and A. lanata concerning the expression of PI3K, Akt1, mTOR, P53, and EGFR genes in human colorectal adenocarcinoma WiDr cells.\n\n\nMethods\n\nThe leaves of two mangrove species – A. alba and A. lanata – were collected from the village of Lubuk Kertang, Brandan Barat, Langkat, North Sumatra, Indonesia. The sample site is situated at 04° 07' 39.71'' North latitude and at 98° 30'97.87'' East longitude.\n\nEvery 500 g of powdered simplicia mangrove leaves of A. alba and A. lanata was macerated with a mixture of chloroform/methanol (2:1, v/v) (CM21) for 48 h. The cell wall debris insouble in CM21 was removed by filtration paper (Advantec, Japan) and the extract was partially purified as lipid extract. The simplicia procedure is described briefly; the wet mangrove leaves were sorted to remove dirt; the leaf bones were removed and cleaned from attached leaf spines; the leaves were cut into 3 cm long pieces, after which they were washed under running water, drained, and weighed. Subsequently, the mangrove leaves were dried in a drying cabinet, dry sorted to separate unwanted plant parts and other impurities that were still present, remain in dry simplicia, and weighed and stored in tightly closed plastic containers8. The lipid extracts of leaves were refluxed at a temperature of 65°C for 24 h in 86% ethanol containing KOH 2 M. The portions were further saponified and diluted with 2 mg/mL n-hexane. The extract in n-hexane was concentrated using rotary evaporator at 40°C. Subsequently, a thick extract was obtained and the concentration was adjusted to 1 mg/mL n-hexane.\n\nLeaf (50–100 mg) extracts were applied to thin layer chromatography (TLC) plates to identify the polyisoprenoid composition4,12,13. The polyisoprenoid standards were generously provided by Dr Ewa Swiezewska (Institute of Biochemistry and Biophysics, Polish Academy of Sciences, Warsaw, Poland). The polyisoprenoid compounds in the A. alba (PAA) and A. lanata (PAL) leaves were confirmed to belong to the dolichol family 100% (C60–C100) and (C70–C100), respectively4.\n\nWiDr cells (isolated human colon cancer cells) and normal cells (Vero) were kindly provided by the Laboratory of Parasitology Collection, Faculty of Medicine, Gadjah Mada University (Yogyakarta, Indonesia). The WiDr cell lines were cultured in Roswell Park Memorial Institute (RPMI) 1640 medium (Sigma Aldrich, Singapore) and the Vero cell lines were nurtured in M199 (Gibco, USA). Both cells were supplemented with 10% (v/v) foetal bovine serum (FBS) (Gibco), 1% penicillin and streptomycin (Gibco), and 0.5% fungizone (Amphoterin B) in a 37°C incubator with 5% CO211.\n\nCytotoxicity tests were conducted on the WiDr and Vero cells using the MTT method. Cells were grown in 96 well microplates to obtain a density of 1 × 104 cells/well and incubated in a 5% CO2 incubator at a temperature of 37°C for 48 h to ensure good growth. Once the new medium had been replaced, the cells were treated with varying concentrations of PAA and PAL (1000, 500, 250, 125, and 62.5 µg/mL) as previously reported10. 5-Fu (Sigma Aldrich) was used as a positive control with the same concentration of PAA and PAL and incubated in 5% CO2 at 37°C for 48 h. At the end of the incubation, the culture media was removed, and the cells were washed with PBS. In each of the wells, 100 mL of culture medium (RPMI) and 10 mL MTT (Sigma Aldrich) were added. The cells were incubated again for 3-6 h in 5% CO2 at 37°C. The reaction was stopped with 10% SDS reagent (Sigma Aldrich) in 0.01 N HCl (Merck). The plate was wrapped to protect it from the light so that the wells were opaque, and it was left overnight at room temperature14. Absorption was measured by the ELISA reader (Benchmark 10431, BioRad) at a wavelength of 595 nm.\n\nSelectivity index (SI) was determined from the IC50 of the polyisoprenoid extract from PAA and PAL leaves in Vero cells versus WiDr cells to exhibit the cytotoxic selectivity of the polyisoprenoid extract, as previously reported9. IC50 was calculated from concentrations that caused death among 50% of the cell population analysed using probit analysis in SPSS version 23 with a significance of 0.058.\n\nThe WiDr cells (5 × 103 cells/well) were added to a 6-well plate which was incubated for 24 h for optimal growth. Subsequently, the cells were exposed to selected concentrations of PAL and PAA (1/5 IC50) and incubated again10. Floating as well as attached cells were collected by adding 0.025% trypsin. The cells from each well were transferred to a separate eppendorf tube. 1 mL PBS was added and the PBS was removed with a micropipette and centrifuged at 2500 rpm for 5 min. The supernatant was removed and 1 µL RNase/PI staining solution (Thermo Fisher Scientific) was added and kept for 10 min a dark place (avoiding light) at 37°C. The cell cycle distribution was analysed using the FAC Scan Flow Cytometer (BD Biosciences) and the percentage of cells obtained in each cell cycle phase (G1/S and G2/M) was calculated using the software ModFit LT. 3.0 s for Windows (Verity Software House).\n\nThe double staining method was used to determine the level of apoptosis. WiDr cells were grown in a 6-well microplate at density of 5 × 105 cells/well and incubated for 24 h. The following day, the cells were treated with concentrations of PAA and PAL (1/5 IC50) and incubated for 24 h. The procedure for apoptosis with double staining was conducted by taking cells from the CO2 incubator and observing the conditions. Then, the calculated cells were used to prepare 24-well plates and slipcovers. 200 µl of cell suspension was evenly and slowly transferred just above the coverslip. The cell was kept in the incubator for 3–30 min to attach to the coverslip with 800 μl of culture media which was added for 48 h of incubation. The culture media was slowly disposed, and the cells were washed with PBS (500 µl). The sample and media were added into the well for control cells and then incubated. All media from the well was slowly removed using a Pasteur pipette. Cells in the wells were washed with PBS. The coverslip was removed using tweezers, placed on a glass slide, and labelled. 10 μl of the reagent mixture of ethidium bromide acridine orange (Sigma Aldrich) was added over the slipcover. The mixture was flattened and gently rocked. Apoptosis was observed under a fluorescence microscope (Olympus CKX41)9. The fluorescent green cells were alive and the fluorescent red cells were dead15.\n\nTotal RNA was extracted from the WiDr cells treated with PAA and PAL (7.5 × 108 cells/well) using the Total RNA Mini Kit (Geneaid), according to the manufacturer’s protocol. The total RNA (3 mg each) was reverse-transcribed with 1 µg random primer to produce cDNA in a total volume of 20 µl using ReverTra Ace kit (Toyobo) with 10 mM dNTP to incubate for 10 min at 30°C, for 60 min at 42°C, and for 5 min at 99°C according to manufacturer’s procedure. The resulting cDNA mixture was diluted using 100 μL TE buffer (10 mM Tris/HCl, 1 mM EDTA, ph 8.0) and directly used for the subsequent PCR.\n\nSemi-quantitative RT-PCR for genes p53, EGFR, PI3K, Akt1, and mTOR16–20 were assessed using 1 μL cDNA added to 25 μL PCR Master Mix which contained 12.5 μL GoTaxGreen, 1 μL primer forward, and 1 μL primer reverse (as listed in Table 1), and 9.5 μL DNase/RNase free water. 35–40 cycles of semi-quantitative RT-PCR (ProFlex PCR system, Thermo Fisher Scientific) were conducted under the following cycling conditions: 15–30 sec at 94°C, 45 sec at 94°C, and 10 sec at 55–60°C, with the final extension phase at 72°C for 5 min and then storage at -20°C21. Semi-quantitative RT-PCR products were observed using 2% agarose gel and stained with ethidium bromide. The bands were documented using the image scanner Doc XR Gel (Bio-Rad)22.\n\nAll the data were analysed using SPSS version 23. The data are presented as mean ± standard error of the mean (SEM). One-way variance analysis (ANOVA) was used to compare the results for different conditions. P < 0.05 was considered significantly different.\n\n\nResults and discussion\n\nThe cytotoxicity test is a preliminary parameter to determine the potential toxicity of a test substance, particularly cancer cells. The toxicity is expressed by IC50 parameters. However, the cytotoxic test can also be performed to assess the toxicity of a test substance on normal cells. So, it can be used to demonstrate selective cytotoxic effects against a cancer line. In this study, the test material is cytotoxic for the polyisoprenoids of leaves derived from two mangrove species (PAA and PAL) against colon cancer cells (WiDr) with a concentration series of 1000, 500, 250, 125, and 62.5 μg/mL. The purpose of testing the extract was obtaining the smallest IC50 value for subsequent use as an advanced test for anticancer activity. The results showed that the smallest IC50 value obtained from the polyisoprenoids of leaves from PAL was 243.32 ug/mL and from PAA was 258.14 ug/mL. Therefore, these concentrations were used in the remaining experiments of the study. The cytotoxic effects were indicated by absorbance values and analysed using a probit analysis to obtain IC50, as shown in Table 2.\n\nSI: selectivity index.\n\nThe greatest cytotoxic activity against WiDr cells, shown by the smallest IC50 value, was obtained from PAL. Therefore, PAL has the most active anticancer activity because the IC50 value showed that PAL could block 50% of the WiDr cell growth. Some extracts are only considered active if they have IC50 values ≤100 μg/mL16. However, it has been demonstrated that an extract value of IC50 of 100–500 μg/mL can be classified as moderate and, therefore, can potentially be developed as an anticancer agent17. Even though a study has reported that an extract is considered active if IC50 > 500 μg/mL18. As shown in Table 2, the SI of PAL and PAA were lower than that of the positive control. The cytotoxicity of PAL and PAA in the present study included an interesting SI against WiDr cells in a dose-dependent manner17.\n\nA previous study showed that methanol and water extracts of A. alba leaves have distinctive properties in regulators and mediators of cancer19. This study tested the cell cycle using flow cytometry to determine the distribution of cells in each phase of the cell cycle at sub G1, S, and G2/M after treatment and obtained predictable pathway inhibition using PAA and PAL to inhibit the cycle cell20.\n\nThe inhibition of the cell cycle in this study is shown in Figure 1 and Table 3. Table 3 shows the control group’s WiDr cell accumulation in the G0/G1, S, and G2/M phase as 76.63%, 7.22%, and 17.93%, respectively. The accumulation of cells in the S phase and G2-M cells increased by 10.60%, 10.51% and 23.84%, and 22.05%, respectively after being administered with a concentration of PAL 1/5 IC50 and PAA with the concentration of 1/5 IC50. The phase change is considered related to the concentration. However, the overall mechanism of inhibition of the cell cycle for PAA and PAL occurred at S and G2-M phases.\n\n(a) Control cell; (b) PAL 1/5 IC50; (c) PAA1/5IC50; and (d) 5-Fu1/5 IC50.\n\nAs Table 3 illustrates, the administration of 5-Fu with 1/5 concentration IC50 decreases the accumulation of WiDr cells in the G2-M phase at 6.42%. The increase in cell accumulation occurred in the G0-G1 phase – S was 88.12 and 9.52%. However, it can be confirmed that the overall mechanisms of cell cycle inhibition of 5-Fu (in the G0-G1 and S phases) had a different mechanism compared with PAL and PAA. Treatment of cancer cells with 5-Fu can accumulate cells at the G1 phase and at the beginning of the synthesis phase (G1/S arrest)21. However, the cell cycle’s inhibitory activity using 5-Fu depends on the type of cancer cell. In colon cancer cells HCT-15 and HT-29, 5-Fu inhibited at the G2/M phase. 5-Fu increases the expression of cyclin A, cyclin B, and CDC2, which is a regulatory protein in the G2/M phase22. The mechanism that mediates the activity in this phase needs to be explored further. In Lovo and WiDr cells, 5-Fu inhibits the cell cycle in the S phase18. This suggests that the activity of 5-Fu is not always associated with thymidylate synthase inhibitory activity, and the activity of 5-Fu in the cell cycle if used in a different cell needs to be researched further. The antineoplastic activity of PAL and PAA occur in the S phase of the cell cycle, which involves the possibility of bonding with the DNA through intercalation between the base pairs as well as the inhibition of the DNA and RNA synthesis18.\n\nIn the present study, increased apoptosis (seen using the reagent acridine orange-ethidium bromide through the fluorescence microscope) was obtained by a percentage increase in each of the phases. Apoptosis is generally characterised by different morphological characteristics and biochemical mechanisms that depend on energy23. Apoptosis usually occurs during development and aging and as a homeostatic mechanism to maintain the population of cells in the network. It also is a defence mechanism when cells are damaged by disease or harmful agents or during immune reaction24. Cytotoxic data test samples indicate the presence of a cytotoxic effect as shown in Figure 2. The green cells are the live ones while the red ones are dead. The range of red fluorescent cells represents the necrotic cells.\n\n(a) Control cell; (b) PAL1/5 IC50; (c) PAA1/5 IC50; and (d) 5-Fu1/5 IC50. death cells (apoptotic) living cells\n\nThe present analysis used raster imaging to count the dead cells and control cells after observing them under a fluorescence microscope. In total, 95.15% green colour (living cells) and 4.84% red colour (dead cells) were observed in control cells. With the PAL 1/5 IC50 treatment, the cells produce 41.09% green colour (living cells) and 58.91% red (dead cells). With PAA 1/5 IC50 treatment, cells produce 90.14% green colour (living cells) and nearly 9.86% red colour (dead cell). With the extract of PAL, the 1/5 IC50 treatment cells produce 78.11% green and as much as 21.88% red colour. With 5-Fu 1/5 IC50 treatment, the cells produce 77.48% green (live cells) and 22.51% red colour (dead cells) (Figure 2).\n\nThe results of the control cells were observed to be green/alive cells. The green colour comes from the orange acridine penetrating the entire living cell with intact membranes and nuclei. In cells treated with PAL and PAA, there was a predominantly red colour which illustrates that the WiDr cells were dead. The orange colour is produced by ethidium bromide interacting with damaged cell membranes and nuclei25. The test results showed that the extract can inhibit the growth of cancer cells, especially in WiDr cancer cells. The inhibition capability through the mechanism of apoptosis can also be evidenced through testing and analysis of double staining flow cytometry. The results of both analyses can illustrate the mechanism of cell death caused by apoptosis both quantitatively and qualitatively. Thus, the induction of apoptosis shows that these treatments are a promising treatment for cancer. The cancer cells undergo apoptosis and lose their ability to proliferate rapidly. This way of treatment may induce usual apoptotic signalling, thereby, potentially eliminating the cancer cells26.\n\nThe potential working mechanism of PAL and PAA are in the late phase of apoptosis. The potency of PAL and PAA in triggering apoptosis may be caused by compound isoprenoids (based on the results of phytochemical screening of PAL and PAA). Steroids/triterpenoids are compounds that have high anticancer activity, by blocking nuclear factor-kappa B, inducing apoptosis, and activating transcription and angiogenesis, which can be useful in the treatment of various types of cancer27.\n\nThe measurement of the expression of PI3K, Akt1, mTOR, P53, and EGFR genes using RT-PCR produces the band illustrated in Figure 3. Gene expression density was quantified using a computerised system (Table 4). Table 4 shows significant differences between the treatment groups. Gene expression results on PI3K, Akt1, mTOR, P53, and EGFR differed significantly (p < 0.05).\n\na. Sig (P) < 0.05: There is a significant difference with the normal group (control cell).\n\nb. Sig (P) < 0.05: There is a significant difference with the group PAL.\n\nc. Sig (P) < 0.05: There is a significant difference with the group PAA.\n\nd. Sig (P) < 0.05: There was no significant difference in the positive control group (5-Fu).\n\nIn this study, RT-PCR showed that the anti-apoptotic gene expression of P53 increases compared to control, whereas the expression of pro-apoptotic genes (PI3K, Akt1, mTOR, and EGFR) tend to decrease. P53 is a tumour-suppressing protein that can affect the permeability of the mitochondrial membrane and directly induce apoptosis without inducing the transcription of the target gene associated with apoptosis in advance28. This condition causes apoptosis when the gene Bax mRNA expression does not increase and, thus, the pathway of apoptosis by P53 has two paths to the mitochondria – directly and indirectly through the activation of the transcription of genes under it. P53 molecule is a tumour-suppressing protein found at a low level under normal conditions and has a short life span. P53 is activated when the cells are exposed to stimuli such as agents that cause DNA damage, hypoxia, lack of nucleotide, or tumour cell activation. As a tumour suppressor, P53 protects the genome and regulates the growth and proliferation of the critical points in response to stress. The P53 molecule is an upstream regulator of the cell cycle as well as the intrinsic apoptotic pathway mediated by the Bcl-2 protein29.\n\nBesides from being a tumour suppressor protein, P53 also acts as a transcription factor for the activation of the expression of multiple target genes involved in various biological functions, such as apoptosis and cell cycle arrest30. The results of the present study show that PAL and PAA significantly increase the gene expression of P53 at high concentrations, and the increase in P53 gene expression is concentration dependant; the higher the concentration, the higher the level of gene expression. These results are consistent with previous studies stating that the expression of P53 protein increases apoptosis31.\n\nThe density of PI3K, Akt1, EGFR and mTOR gene expression was been significantly downregulated in treatment cells compared to the control cell in the present study. PAL administration demonstrated a more significant reduction in PI3K, Akt1 and EGFR gene expression than PAA and 5-Fu, while mTOR was downregulated more with 5-Fu than PAL and PAA. The density of the P53 gene expression was significantly upregulated in the treatment cells compared to the control cell, and this was more significant with PAL than PAA and 5-Fu.\n\n\nConclusions\n\nOverall, the present study confirmed that PAL and PAA can affect anti-apoptotic P53 gene expression by upregulating this gene than the controls, while the expression of pro-apoptotic genes PI3K, Akt1, mTOR, and EGFR were downregulated compared to the controls. In addition, PAL and PAA inhibited the WiDr cell cycle in later apoptosis (S and G2-M1). Therefore, this study confirms that the polyisoprenoids derived from A. alba and A. lanata leaves are promising chemopreventive agents for colon cancer.\n\n\nData availabilty\n\nFigshare: Dataset for manuscript: Effects of polyisoprenoids from Avicennia lanata and Avicennia alba leaves on the gene expression of P13K, Akt1, mTOR, P53, and EGFR in human colorectal adenocarcinoma WiDr cells using reverse transcription-PCR, https://doi.org/10.6084/m9.figshare.11839350.v132.\n\nThis project contains the following data:\n\n- Underlying data for Table 2–Table 4.\n\nFigshare: Dataset for manuscript: Effects of polyisoprenoids from Avicennia lanata and Avicennia alba leaves on the gene expression of P13K, Akt1, mTOR, P53, and EGFR in human colorectal adenocarcinoma WiDr cells using reverse transcription-PCR, https://doi.org/10.6084/m9.figshare.11856039.v233.\n\n- Uncropped, unedited images for Figure 2 and Figure 3.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "References\n\nGupta SC, Kim JH, Prasad S, et al.: Regulation of survival, proliferation, invasion, angiogenesis, and metastasis of tumor cells through modulation of inflammatory pathways by nutraceuticals. Cancer Metastasis Rev. 2010; 29(3): 405–434. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMitchison TJ: The proliferation rate paradox in antimitotic chemotherapy. Mol Biol Cell. 2012; 23(1): 1–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGiri C, Ochieng E, Tieszen LL, et al.: Status and distribution of mangrove forests of the world using earth observation satellite data. Glob Ecol Biogeogr. 2011; 20(1): 154–159. Publisher Full Text\n\nBasyuni M, Sagami H, Baba S, et al.: Distribution and occurrence of new polyprenyl acetone and other polyisoprenoids in Indonesian mangroves. Dendrobiology. 2017; 78: 18–31. Publisher Full Text\n\nNurhanan MY, Asiah O, Ilham MAM, et al.: Antiproliferative activities of 32 Malaysian plant species in breast cancer cell lines. J Tropical Forest Sci. 2008; 20(2): 77–81. Reference Source\n\nAkter R, Uddin SJ, Grice ID, et al.: Cytotoxic activity screening of Bangladeshi medicinal plant extracts. J Nat Med. 2013; 2013(1): 1–7. PubMed Abstract | Publisher Full Text\n\nNiranjana R, Gayathri R, Mol SN, et al.: Carotenoids modulate the hallmarks of cancer cells. J Funct Foods. 2015; 18(Part B): 968–985. Publisher Full Text\n\nSari DP, Basyuni M, Hasibuan PA, et al.: Cytotoxic and Antiproliferative Activity of Polyisoprenoids in Seventeen Mangroves Species Against WiDr Colon Cancer Cells. Asian Pac J Cancer Prev. 2018; 19(12): 3393–3400. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIllian ND, Basyuni M, Wati R, et al.: Polyisoprenoids from Avicennia marina and Avicennia lanata inhibit WiDr cells proliferation. Pharmacogn Mag. 2018; 14(58): 513–518. Publisher Full Text\n\nIllian DN, Hasibuan PAZ, Sumardi S, et al.: Anticancer activity of polyisoprenoids from Avicennia alba Blume. in WiDr Cells. Iran J Pharm Res. 2019; 18(3): 1477–1487. Publisher Full Text\n\nTropberger P, Pott S, Keller C, et al.: Regulation of transcription through acetylation of H3K122 on the lateral surface of the histone octamer. Cell. 2013; 152(4): 859–872. PubMed Abstract | Publisher Full Text\n\nKurisaki A, Sagami H, Ogura K: Distribution of polyprenols and dolichols in soybean plant. Phytochemistry. 1997; 44(1): 45–50. Publisher Full Text\n\nSagami H, Kurisaki A, Ogura K, et al.: Separation of dolichol from dehydrodolichol by a simple two-plate thin-layer chromatography. J Lipid Res. 1992; 33(12): 1857–1862. PubMed Abstract\n\nSari DP, Basyuni M, Hasibuan PAZ, et al.: Cytotoxic effect of polyisoprenoids from Rhizophora mucronata and Ceriops tagal leaves against WiDr colon cancer cell lines. Sains Malaysiana. 2018; 47(9): 1953–1959. Publisher Full Text\n\nKucsko G, Maurer, PC, Yao NY, et al.: Nanometre-scale thermometry in a living cell. Nature. 2013; 500(7460): 54–58. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu HC, Chen GG, Vlantis AC, et al.: 5-fluorouracil mediates apoptosis and G1/S arrest in laryngeal squamous cell carcinoma via a p53-independent pathway. Cancer J. 2006; 12(6): 482–493. PubMed Abstract | Publisher Full Text\n\nLim YJ, Rhee JC, Bae YM, et al.: Celecoxib attenuates 5-fluorouracil-induced apoptosis in HCT-15 and HT-29 human colon cancer cells. World J Gastroenterol. 2007; 13(13): 1947–1952. PubMed Abstract | Free Full Text\n\nBackus HH, Pinedo HM, Wouters D, et al.: Differences in the induction of DNA damage, cell cycle arrest, and cell death by 5-fluorouracil and antifolates. Oncol Res. 2000; 12(5): 231–239. PubMed Abstract | Publisher Full Text\n\nTyagi AK, Agarwal C, Chan DC, et al.: Synergistic anti-cancer effects of Silibinin with conventional cytotoxic agents doxorubicin, cisplatin and carboplatin against human breast carcinoma MCF-7 and MDA-MB468 cells. Oncol Rep. 2004; 11(2): 493–499. PubMed Abstract | Publisher Full Text\n\nWeerapreeyakul N, Nonpunya A, Barustux S, et al.: Evaluation of the anticancer potential of six herbs against a hepatoma cell line. Chin Med. 2012; 7(1): 15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMachana S, Weerapreeyakul N, Barusrux S, et al.: Cytotoxic and apoptotic effects of six herbal plants against the human hepatocarcinoma (HepG2) cell line. Chin Med. 2011; 6(1): 39. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSatapathy S, Jena BK: Antitumor and growth effector screen of leaf extracts of selected mangroves of Bhitarkanika, Odisha. Int J Technol Enhanc Emerg Eng Res. 2013; 1(4): 25–30. Reference Source\n\nRoss JJ, Kennedy GA, Macnab F, et al.: The effectiveness of spiritual/religious interventions in psychotherapy and counselling: a review of the recent literature. Psychotherapy and Counselling Federation of Australia. 2015; 1–23. Reference Source\n\nNorbury CJ, Hickson ID: Cellular responses to DNA damage. Annu Rev Pharmacol Toxicol. 2001; 41: 367–401. PubMed Abstract | Publisher Full Text\n\nFink SL, Cookson BT: Apoptosis, pyroptosis, and necrosis: Mechanistic description of dead and dying eukaryotic cells. Infect Immun. 2005; 73(4): 1907–1916. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWong RS: Apoptosis in cancer: from pathogenesis to treatment. J Exp Clin Cancer Res. 2011; 30: 87. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRios JL, Andujar I, Escandell JM, et al.: Cucurbitacins as inducers of cell death and a rich source of potential anticancer compounds. Curr Pharm Des. 2012; 18(12): 1663–1676. PubMed Abstract | Publisher Full Text\n\nChipuk JE, Maurer U, Green DR, et al.: Pharmacologic activation of p53 elicits Bax-dependent apoptosis in the absence of transcription. Cancer Cell. 2003; 4(5): 371–381. PubMed Abstract | Publisher Full Text\n\nEl-Deiry WS: Regulation of p53 downstream genes. Semin Cancer Biol. 1998; 8: 345–357. PubMed Abstract | Publisher Full Text\n\nZhang X, Mukerji R, Samadi AK, et al.: Down-regulation of estrogen receptor-alpha and rearranged during transfection tyrosine kinase is associated with withaferin a-induced apoptosis in MCF-7 breast cancer cells. BMC Complement Altern Med. 2011; 11(1): 84. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYeap S, Akhtar MN, Lim KL, et al.: Synthesis of an anthraquinone derivative (DHAQC) and its effect on induction of G2/M arrest and apoptosis in breast cancer MCF-7 cell line. Drug Des Devel Ther. 2015; 9: 983–992. PubMed Abstract | Publisher Full Text | Free Full Text\n\nQurrohman T, Hasibuan PAZ, Nuryawan A, et al.: Dataset for manuscript: Effects of polyisoprenoids from Avicennia lanata and Avicennia alba leaves on the gene expression of P13K, Akt1, mTOR, P53, and EGFR in human colorectal adenocarcinoma WiDr cells using reverse transcription-PCR. figshare. 2020. Dataset. https://doi.org/10.6084/m9.figshare.11839350.v1\n\nQurrohman T, Hasibuan PAZ, Nuryawan A, et al.: Dataset for manuscript: Effects of polyisoprenoids from Avicennia lanata and Avicennia alba leaves on the gene expression of P13K, Akt1, mTOR, P53, and EGFR in human colorectal adenocarcinoma WiDr cells using reverse transcription-PCRDataset for manuscript: Effects of polyisoprenoids from Avicennia lanata and Avicennia alba leaves on the gene expression of P13K, Akt1, mTOR, P53, and EGFR in human colorectal adenocarcinoma WiDr cells using reverse transcription-PCR. figshare. Dataset. 2020. https://doi.org/10.6084/m9.figshare.11856039.v2"
}
|
[
{
"id": "61218",
"date": "25 Mar 2020",
"name": "Ewa Swiezewska",
"expertise": [],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nWe, Liliana Surmacz and Ewa Swiezewska, found the manuscript entitled ‘Effects of polyisoprenoids from Avicennia lanata and Avicennia alba leaves on the gene expression of PI3K, Akt1, mTOR, P53, and EGFR in human colorectal adenocarcinoma WiDr cells using reverse transcription-PCR’ prepared by Dr T. Qurrohman and coworkers not acceptable for publication in the journal F1000Research in its current form.\n\nThis manuscript describes elucidation of the biological activity of two polyisoprenoid-containing extracts isolated from leaves of two mangrove plant species Avicennia lanata and Avicennia alba. To meet the generally acceptable standard of the manuscript quality this preliminary study requires extremely extensive revision. Firstly, all the biological data were obtained using crude and poorly characterized extracts and consequently all these experiments should be performed once again with highly purified polyisoprenoid mixture (polyisoprenoid content at least 90%). Such purification procedure is doable within a reasonable time period and is not very difficult. Secondly, several methodological issues require clarification, e.g. selection of the drug concentration points, appropriate control experiments (composition of the cell growth media), number of replicates, methods of quantification of PCR results and statistical analysis of the data. Thirdly, interpretation of the obtained results seems not really clear, in particular results of the analysis of the toxicity should be reconsidered (see below). Finally, the Authors should carefully discuss the available literature data to document the novelty of their study. A detailed list of comments follows.\n\nThis manuscript describes study on the biological effects of the polyisoprenoid-containing extracts isolated from leaves Avicennia lanata and Avicennia alba. All tests were performed for two cell lines, WiDr and Vero. Shown are data on the effect of PAA and PAL on the cytotoxic activity, cell-cycle progression and expression of the selected genes: PI3K, Akt1, mTOR, and EGFR.\n\nGeneral comments: All the biological tests described in this study were performed using the crude mixture of lipids obtained after alkaline hydrolysis of the leaf extracts. The Authors did not provide the approximate polyisoprenoid content in the mixtures used. Such complex mixture of natural compounds could be used just for preliminary tests of their potential biological activity. What is more, data presented in this study in fact confirm the weakness of the strategy used – e.g. different effects of pro-apoptocic activity of PAA and PAL on the level of expression of analyzed genes might suggest that this is NOT the effect of polyisoprenoids (very similar profile for both extracts) but rather some other components of these mixtures. In order to present any reasonable conclusion on the biological effect of polyisoprenoids highly purified polyisoprenoid samples (purity higher than 90% based on the HPLC/UV assay) should be used. For this reason the experiments have to be reproduced with purified polyisoprenoid samples. In this context the title of the manuscript does not describe the real content of the work.\n\nData presented in Table 2 suggest that analyzed extracts exert higher toxicity against Vero than WiDr cells. Consequently the Selectivity Index is below 1. It raises the question whether analyzed extracts could be considered as candidates for anticancer drugs since they posses higher cytotoxic activity against normal than cancer cells. Authors should carefully comment on this observation. Keeping in mind that analyzed extracts were not pure these data do not preclude the potential usefulness of polyisoprenoids as drugs although this conclusion cannot be made on the basis of results shown in Table 2. Furthermore, interpretation of the quantitative results of cytoxicity against WiDr cells is questionable – the difference between both IC50 values obtained for PAA and PAL is approx. 6% (Table 2); thus the conclusion presented by Authors seems highly exaggerated. Finally, why are no SD values are presented here? Have these experiments been performed only once?\n\nExperiments on the effect of plant extracts on the cell-cycle progression are summarized as follows: ‘The antineoplastic activity of PAL and PAA occur in the S phase of the cell cycle, which involves the possibility of bonding with the DNA through intercalation between the base pairs as well as the inhibition of the DNA and RNA synthesis.’ Such conclusion definitely requires additional data supporting the suggested mechanism(s). Moreover, presented data were not analyzed using statistical method. Fig.2 presents representative data but how many biological replica were performed? How many microscopic snapshots were used for quantitative analysis?\n\nExpression of selected genes - Authors should add a general comment on the biological relevance of the transcriptional data in the context of transcript – translated protein – its function in the cell, e.g. is the level of transcript directly related to the level and/or activity of the appropriate protein?\n\nWhat concentartion of PAL or PAA was used to obtain data presented in Fig.3 and Table 4? Since Fig.3 and Table 4 present data for just one concentration point thus the statement is not justified: ‘The results of the present study show that PAL and PAA significantly increase the gene expression of P53 at high concentrations, and the increase in P53 gene expression is concentration dependent; the higher the concentration, the higher the level of gene expression.’ Furthermore, how quantification of the PCR product, i.e. bands intensity was performed? What software was used for this analysis? Why no reference gene was used as a control for reaction efficiency? How many RT-PCR repetitions were used for quantitative analysis? The term ‘Gene expression density’ should be replaced with ‘Gene expression level’ or ‘Transcript expression level’ Finally, a general comment concerning the RT-PCR technique - currently a qPCR method is considered an acceptable standard for trnascriptomic analysis. It is recommended by these reviewers to repeat analyses using this approach.\n\nIn the Introduction the rationale behind the choice of the particular genes to be analyzed (PI3K, Akt1, mTOR, P53, and EGFR) has to be presented together with a brief (one sentence) description of their function in the cancer cell. Additionally, rewrite the sentence and explain the meaning of the expression ‘and differentiate between the level of individual plasmid expression in multivalent pDNA’. How this expression is relevant to the method used in this study?\n\nMethodology requires clarification:\nDelete the description of the cleaning of the leaves from ‘The simplicia procedure …’ to ‘…stored in tightly closed plastic containers.’ Instead, describe the date of the collection of the leaves, for how long the plant material was stored prior to extraction, the temperature of the extraction procedure.\n\nWhat is meant in the sentence: ‘The portions were further saponified and ……..’ All the ester have already been saponified upon KOH treatment.\n\nWas a crude unsaponifiable lipid fraction used for further tests? It has to be clearly mentioned in the text.\n\nIf so what was the approximate polyisoprenoid content in this fraction?\n\nData obtained in the biological test are not fully justified keeping in mind the complexity of the lipid fraction used.\n\nA type of the TLC plates should be mentioned in the text, moreover what is the meaning of the expression: ‘… were confirmed to belong to the dolichol family 100%...’.\n\nWhat type of cell line are Vero cells? Which organ is it derived from? Is it an immortalized cell line?\n\nConcentration of 5-Fu used in this study – please provide literature data on the concentration of 5-FU used. Were the values used in this study within the range of those used in similar type of assays?\n\nThe method of administration of the leaf extracts to the cells is not described at all. Polyisoprenoids are not soluble in water so what type of solvent was used to prepare PAA and PAL solutions for all biological tests? And how the control cells were treated – there is no information, neither in the Methods nor in the Figure legends, on the supplementation of the growth media of the control cell cultures with the appropriate amount of the same solvent.\n\nFor how long the cells were exposed to PAL in the cell cycle analysis? What concentrations of PAL and PAA were used?\n\nThe description of the apoptosis analysis is unclear and requires rewriting. What type of software was used to count green and red fluorescence signals?\n\nThe amount of total RNA used in this study in the reverse transcription reaction (according to the manuscript 3 mg in the 20 µl reaction mixture) seems unrealistic. Manufacturer's protocols usually recommend to use ≤1 µg RNA. Please correct.\n\nThe novelty of the data presented in this study has to be indicated by Authors in the section Discussion or Conclusions – novel data presented in the current manuscript in comparison to the previous already published ones should be clearly depicted.\n\nThe manuscript requires an English language edit. Some sentences are misleading, some are just awkward, e.g. ‘Cytotoxic activity against WiDr cells showed that the IC50 for A. alba and A. lanata was 258.14 ug/mL and 243.32 ug/mL, respectively. This indicated that their classification as anticancer agents was moderate.’ ‘Natural ingredients developed as potential chemotherapeutic agents include mangrove leaves.’ should rather read as follows: ‘Natural substances developed as potential chemotherapeutic agents include components of mangrove leaves.’ ‘This extract has a mechanism for inhibiting the cell cycle at the G0-G1 phase…’ ‘In this study, the test material is cytotoxic for the polyisoprenoids of leaves derived from…’ The word ‘simplicia’ is used in the text several times– please replace with any typical expression.\n\nMinor remarks:\nTitle – ‘using reverse transcription-PCR’ - in my opinion these last four words are not necessary and should be deleted; it is nothing special in this RT-PCR technique.\n\np.1 – ‘Mangrove plants produce a polyisoprenoid compound. Polyisoprenoids have been proven to have anticancer properties.’ – these two sentences have to be rewritten. As it is now they do not provide enough information about the subject of study. Moreover, why the term ‘polyisoprenoid compound’ is used here in a singular form? It is not consistent either with the natural composition of the leaf extract or with the text below.\n\np.1 – delete ‘inhibited’ in the sentence ‘The inhibited cell cycle and apoptosis...’.\n\np.3 – rewrite the second sentence of Introduction – what is meant by the expression ‘halt its invasion and metastasis’ in the context of this sentence?\n\np.3 – replace ‘…dolichol and polyprenol on the leaves…’ with ‘…dolichol and polyprenol in the leaves…’.\n\np.3 – ‘MCF-7 and T47D’ – a brief description of the cell line type is missing.\n\np.3 – ‘gene expression of COX-2 in colon cancer cells’ spell out the name of the enzyme when first used here and throughout the entire manuscript.\n\np.3 – delete the word ‘Every’ in the sentence ‘Every 500 g of powdered simplicia mangrove leaves…’. Explain the word ‘simplicia’ in the text or delete.\n\np.3 – ‘The cell wall debris insouble in CM21…’ – how about other cellular components, e.g. proteins? – replace ‘cell wall debris’ with ‘Precipitate’.\n\np.3 – delete the description of the cleaning of the leaves from ‘The simplicia procedure …’ to ‘…stored in tightly closed plastic containers.’ Instead, describe the date of the collection of the leaves, for how long the plant material was stored prior to extraction, the temperature of the extraction procedure.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "5433",
"date": "27 Apr 2020",
"name": "Mohammad Basyuni",
"role": "Author Response",
"response": "We, Liliana Surmacz and Ewa Swiezewska, found the manuscript entitled ‘Effects of polyisoprenoids from Avicennia lanata and Avicennia alba leaves on the gene expression of PI3K, Akt1, mTOR, P53, and EGFR in human colorectal adenocarcinoma WiDr cells using reverse transcription-PCR’ prepared by Dr T. Qurrohman and coworkers not acceptable for publication in the journal F1000Research in its current form.This manuscript describes elucidation of the biological activity of two polyisoprenoid-containing extracts isolated from leaves of two mangrove plant species Avicennia lanata and Avicennia alba. To meet the generally acceptable standard of the manuscript quality this preliminary study requires extremely extensive revision.Response: We would like to thank the Reviewers’ comments and suggestions which significantly improve the manuscript and enrich the content.Firstly, all the biological data were obtained using crude and poorly characterized extracts and consequently all these experiments should be performed once again with highly purified polyisoprenoid mixture (polyisoprenoid content at least 90%). Such purification procedure is doable within a reasonable time period and is not very difficult.Response: We agreed with Reviewers’ suggestions to revise the procedures. The polyisoprenoid extraction was performed using the established procedures. The lipid extracts of Avicennia alba and A. lanata was concentrated to dryness and saponified at 65°C for 24 h in 86% ethanol containing KOH 2 M. The unsaponifiable lipid partitioned into hexane by vigorous mixing was analyzed by silica gel 60 TLC and RP-18 high performance thin layer chromatography (HPTLC) plates. The unsanifiable lipid basically denotes simple lipid fractions except for fatty acids (saponifiable lipid). The leaf extracts (50-100 mg) were applied to TLC plate. The quantity of polysioprenoids in A. alba leaves was 5.5±0.8 mg/g dry weight and A. lanata leaves was 14.9±1.2 mg/g dry weight. Data are represented as the means ± SEM (n=3). We enclose two-dimensional TLC (2D TLC) chromatograms. Figure 1 shows 2D TLC chromatograms of polyisoprenoids from leaves of A. alba and A. lanata. Dolichols with the chain length of C60-C100 and C70-C100 were detected as major polyisoprenoids alcohols in A. alba leaves (A) and A. lanata leaves (B), respectively. No polyprenol was found in both mangrove leaves. The 2D TLC has been performed triplicates and showed an identical pattern. We confirmed that leaves extracts contained 100% dolichols as highly purified polyisoprenoid mixture to meet Reviewers’ requirement. Therefore it is not needed to purify the polyisoprenoid samples and used for further investigation.New and revised sentences have added to the revised manuscript to incorporate Reviewers’ suggestions. Please refer to the revised version of Preparation of isolation polyisoprenoid alcohols. Figure 1. Two-dimensional TLC chromatograms of polyisoprenoids from Avicennia alba leaves (PAA) (A) and A. lanata leaves (PAL) (B)Secondly, several methodological issues require clarification, e.g. selection of the drug concentration points, appropriate control experiments (composition of the cell growth media), number of replicates, methods of quantification of PCR results and statistical analysis of the data.Response: To meet Reviewers’ suggestions, methodological issues, e.g. selection of the drug concentration points, appropriate control experiments (composition of the cell growth media), number of replicates, methods of quantification of PCR results and statistical analysis of the data have been revised. The raw data of experiments with three independent repetitions has been deposited in https://doi.org/10.6084/m9.figshare.11839350.v4 and https://doi.org/10.6084/m9.figshare.11856039.v4Please refer to revised Methods. Thirdly, interpretation of the obtained results seems not really clear, in particular results of the analysis of the toxicity should be reconsidered (see below).Response: To incorporate Reviewers’the results of the analysis of the toxicity has been revised by adding the standard errors of the means and statistical analysis. Please refer to revised Table 2. Finally, the Authors should carefully discuss the available literature data to document the novelty of their study.Response: We agreed with Reviewers’ suggestions, new sentences have been added to meet Reviewers’ comments, please refer to revised Discussion. A detailed list of comments follows.This manuscript describes study on the biological effects of the polyisoprenoid-containing extracts isolated from leaves Avicennia lanata and Avicennia alba. All tests were performed for two cell lines, WiDr and Vero. Shown are data on the effect of PAA and PAL on the cytotoxic activity, cell-cycle progression and expression of the selected genes: PI3K, Akt1, mTOR, and EGFR.General comments:All the biological tests described in this study were performed using the crude mixture of lipids obtained after alkaline hydrolysis of the leaf extracts. The Authors did not provide the approximate polyisoprenoid content in the mixtures used. Such complex mixture of natural compounds could be used just for preliminary tests of their potential biological activity.Response: Unsaponifiable lipid was analysed using TLC and 2D-TLC plates to identify polyisoprenoids that contained 100% dolichols with the chain length of C60-C100 and C70-C100 detected as main polyisoprenoids alcohols in A. alba leaves and A. lanata leaves. New sentences have been added to the revised manuscript. Please refer to revised Preparation of isolation polyisoprenoid alcohols.What is more, data presented in this study in fact confirm the weakness of the strategy used – e.g. different effects of pro-apoptocic activity of PAA and PAL on the level of expression of analyzed genes might suggest that this is NOT the effect of polyisoprenoids (very similar profile for both extracts) but rather some other components of these mixtures.Response: We clarified that leaves extracts of PAA and PAL contained 100% dolichols based on the HPTLC chromatograms (Figure 1), we believe that different effects of pro-apoptotic activity of PAL and PAL on the level of expression of analysed from the effect of polyisoprenoids (dolichols). We examined several pro-apoptotic genes as questioned by Reviewers as the weakness of the strategy used, we agreed to this point, however, we also examined an anti-apoptotic gene, p53 to prevent the cancer formation involving a mechanism on the tumour suppressor protein p53.In order to present any reasonable conclusion on the biological effect of polyisoprenoids highly purified polyisoprenoid samples (purity higher than 90% based on the HPLC/UV assay) should be used. For this reason the experiments have to be reproduced with purified polyisoprenoid samples.Response: We clarified that leaves extracts of both mangrove contained 100% dolichols based on 2D-TLC chromatograms, no other compunds found. These samples met the criteria as Reviewers’ suggestion for purity samples was higher than 90%. Therefore it is not required to purify the polysioprenoid samples.In this context the title of the manuscript does not describe the real content of the work.Response: To incoporporate Reviewers’s suggestion, the title has been revised to be “Effects of polyisoprenoids drom Avicennia lanata and Avicennia alba leaves on the gene expression of PI3K, Akt1, MTOR, P53 and EGFR in human colorectal adenocarcinoma WiDr cells”Data presented in Table 2 suggest that analyzed extracts exert higher toxicity against Vero than WiDr cells. Consequently the Selectivity Index is below 1. It raises the question whether analyzed extracts could be considered as candidates for anticancer drugs since they posses higher cytotoxic activity against normal than cancer cells. Authors should carefully comment on this observation. Keeping in mind that analyzed extracts were not pure these data do not preclude the potential usefulness of polyisoprenoids as drugs although this conclusion cannot be made on the basis of results shown in Table 2.Response: We agreed with Reviewers’ comments on the results of Selectivity Index of PAA and PAL is below 1, showing higher cytotoxic selectivity We also clarified that the analysed extracts of both samples were pure, contained 100% dolichols. However, as Reviewers also mentioned that PAA and PAL have potential usefulness as drugs.Furthermore, interpretation of the quantitative results of cytoxicity against WiDr cells is questionable – the difference between both IC50 values obtained for PAA and PAL is approx. 6% (Table 2); thus the conclusion presented by Authors seems highly exaggerated.Response: We agreed with Reviewers’ suggestion on small difference between both IC50 values obtained for PAA and PAL, we revised Table 2 and the conclusion to incorporate Reviewers’ comments.Finally, why are no SD values are presented here? Have these experiments been performed only once?Response: SEM values have been added to revised Table 2 to incorporate Reviewers’ suggestion. These experiments have been performed triplicate analyses. The raw data on Table 2 has been deposited on this link: https://doi.org/10.6084/m9.figshare.11839350.v4Experiments on the effect of plant extracts on the cell-cycle progression are summarized as follows: ‘The antineoplastic activity of PAL and PAA occur in the S phase of the cell cycle, which involves the possibility of bonding with the DNA through intercalation between the base pairs as well as the inhibition of the DNA and RNA synthesis.’ Such conclusion definitely requires additional data supporting the suggested mechanism(s).Response: We do not have additional data supporting the suggested mechanism, to incorporate Reviewers’ suggestion the sentence ‘The antineoplastic activity of PAL and PAA occur in the S phase of the cell cycle, which involves the possibility of bonding with the DNA through intercalation between the base pairs as well as the inhibition of the DNA and RNA synthesis,’ has been deleted from the revised manuscript.Moreover, presented data were not analyzed using statistical method. Fig.2 presents representative data but how many biological replica were performed? How many microscopic snapshots were used for quantitative analysis?Response: Observation of apoptotic cells in a fluorescence microscope with a magnification of 40x and 3x in cell control, PAA, PAL, and 5-Fu treatments was analysed using SPSS version 23, followed by Duncan’s multiple range test for treatment comparisons. ImageRaster 4.0.5 was used to count green and red fluorescence signals of three microscopic snapshots of individual experiments. Please refer to revised raw data of Fig. 2 and revised Apoptosis analysis.Expression of selected genes - Authors should add a general comment on the biological relevance of the transcriptional data in the context of transcript – translated protein – its function in the cell, e.g. is the level of transcript directly related to the level and/or activity of the appropriate protein? Response: New sentence has been added to revised Discussion to incorporate Reviewers’ suggestions.What concentartion of PAL or PAA was used to obtain data presented in Fig.3 and Table 4? Since Fig.3 and Table 4 present data for just one concentration point thus the statement is not justified: ‘The results of the present study show that PAL and PAA significantly increase the gene expression of P53 at high concentrations, and the increase in P53 gene expression is concentration dependent; the higher the concentration, the higher the level of gene expression.’Response: We used 5μL PCR product to obtain data presented in Fig. 3 and Table 4. We agreed with Reviewers’ comments to correct the sentence to read ‘The results of the present study show that PAL and PAA significantly increase the gene expression of P53 comparing to control cell and 5-Fu’.Furthermore, how quantification of the PCR product, i.e. bands intensity was performed? What software was used for this analysis? Why no reference gene was used as a control for reaction efficiency? How many RT-PCR repetitions were used for quantitative analysis?Response: New sentences and revised Figure 3 have been included to incorporate Reviewers’ comments. Each data represents the average of three independents RT-PCR measurements with standard errors of individual experiments. To quantify the PCR product, Quantity One® 1-D analysis software (Bio-Rad) used to assess bands intensity of genes analysed. b-actin was reference gene to normalize the PCR efficiency. The term ‘Gene expression density’ should be replaced with ‘Gene expression level’ or ‘Transcript expression level’Response: ‘Gene expression density’ has been replaced with ‘Gene expression level’. Plese refer to revised part Results of Discussion of Expression of PI3K, Akt1, mTOR, P53, and EGFR genesFinally, a general comment concerning the RT-PCR technique - currently a qPCR method is considered an acceptable standard for trnascriptomic analysis. It is recommended by these Reviewers to repeat analyses using this approach.Response: We agreed with Reviewers’ suggestion, however, in the present situation, the authors are unable to repeat analysis using transcriptomic approach.In the Introduction the rationale behind the choice of the particular genes to be analyzed (PI3K, Akt1, mTOR, P53, and EGFR) has to be presented together with a brief (one sentence) description of their function in the cancer cell.Response: To incorporate Reviewers’ suggestion, a new sentences relating to description function of PI3K, Akt1, mTOR, p53, and EGFR in the cancer cell has been added to Introduction. Please refer to last sentence of Introduction.Additionally, rewrite the sentence and explain the meaning of the expression ‘and differentiate between the level of individual plasmid expression in multivalent pDNA’. How this expression is relevant to the method used in this study?Response: The sentence ‘and differentiate between the level of individual plasmid expression in multivalent pDNA’ has been deleted from revised manuscript.Methodology requires clarification: Delete the description of the cleaning of the leaves from ‘The simplicia procedure …’ to ‘…stored in tightly closed plastic containers.’ Instead, describe the date of the collection of the leaves, for how long the plant material was stored prior to extraction, the temperature of the extraction procedure. Response: The description of the cleaning of the leaves from ‘The simplicia procedure …’ to ‘…stored in tightly closed plastic containers.’ has been deleted from revised manuscript. What is meant in the sentence: ‘The portions were further saponified and ……..’ All the ester have already been saponified upon KOH treatment. Response: The sentence: ‘The portions were further saponified and ……..’ has been deleted and changed to new sentence to read ‘The unsaponifiable lipid partitioned into 2 mg/mL n-hexane’. Was a crude unsaponifiable lipid fraction used for further tests? It has to be clearly mentioned in the text. Response: A crude unsaponifiable lipid used for further test and has been clearly mentioned in the text of revised manuscript. If so what was the approximate polyisoprenoid content in this fraction? The quantity of polysioprenoids in A. alba leaves was 5.5±0.8 mg/g dry weight and A. lanata leaves was 14.9±1.2 mg/g dry weight. Data are represented as the means ± SEM (n=3). Data obtained in the biological test are not fully justified keeping in mind the complexity of the lipid fraction used. Response: We agreed with Reviewers’ comments on the complexity of the lipid fraction, however, the biological tests have been clarified to be 100% dolichols. A type of the TLC plates should be mentioned in the text, moreover what is the meaning of the expression: ‘… were confirmed to belong to the dolichol family 100%...’. Response: Types of TLC plates were Silica gel 60 thin layer chromatography (TLC) and RP-18 HPTLC plates have been added to revised Methods, the expression: ‘….were confirmed to belong to the dolichol family 100%...’ has been revised to read, ‘…were detected to be 100% dolichol family...’ What type of cell line are Vero cells? Which organ is it derived from? Is it an immortalized cell line? Response: Vero ATCC® CCL-81™, an immortalized cell line, derived from the kidney of an African green monkey Concentration of 5-Fu used in this study – please provide literature data on the concentration of 5-FU used. Were the values used in this study within the range of those used in similar type of assays? Response: New sentence has been added to meet Reviewers’ comment. 50 μM of 5-fluorouracil (5-Fu) were dissolved in a 100 μL dimethyl sulfoxide (DMSO) co-solvent. The concentration used in this study within the rage of those used in colon cancer cells. The method of administration of the leaf extracts to the cells is not described at all. Polyisoprenoids are not soluble in water so what type of solvent was used to prepare PAA and PAL solutions for all biological tests? And how the control cells were treated – there is no information, neither in the Methods nor in the Figure legends, on the supplementation of the growth media of the control cell cultures with the appropriate amount of the same solvent. Response: To incorporate Reviewers’ comments, new sub title in the Methods have been added. Please refer to Administration of the leaf extracts to the cells of Methods. For how long the cells were exposed to PAL in the cell cycle analysis? What concentrations of PAL and PAA were used? Response: The cell was exposed to serially diluted concentrations of PAA and PAL (1000, 500, 250, 125, and 62.5 µg/mL) in the cell cycle analysis for 48 h. Please refer to revised sentence in the Cytotoxicity analysis of Methods. The description of the apoptosis analysis is unclear and requires rewriting. What type of software was used to count green and red fluorescence signals? Response: New sentence has been added to meet Reviewers’ comments. ImageRaster 4.05. (Micronos, Yogyakarta, Indonesia) was used to count green and red fluorescence signals. Please refer to revised Apoptosis analysis method. The amount of total RNA used in this study in the reverse transcription reaction (according to the manuscript 3 mg in the 20 µl reaction mixture) seems unrealistic. Manufacturer's protocols usually recommend to use ≤1 µg RNA. Please correct. Response: The total RNA (0.3 µg each) has been corrected.The novelty of the data presented in this study has to be indicated by Authors in the section Discussion or Conclusions – novel data presented in the current manuscript in comparison to the previous already published ones should be clearly depicted. Response: We agreed with Reviewers’ suggestion, new and revised sentences have added to revised Conclusion.The manuscript requires an English language edit. Some sentences are misleading, some are just awkward, e.g.‘Cytotoxic activity against WiDr cells showed that the IC50 for A. alba and A. lanata was 258.14 ug/mL and 243.32 ug/mL, respectively. This indicated that their classification as anticancer agents was moderate.’Response: To incorporate Reviewers’ comments, the sentences have been revised to read ‘Cytotoxic activity against WiDr cells showed that the IC50 for A. alba and A. lanata was 258.14 ug/mL and 243.32 ug/mL, respectively. This observation indicated the possibility to develop moderate anticancer agents.’ ‘Natural ingredients developed as potential chemotherapeutic agents include mangrove leaves.’ should rather read as follows: ‘Natural substances developed as potential chemotherapeutic agents include components of mangrove leaves.’Response: Natural ingredients developed as potential chemotherapeutic agents include mangrove leaves.’ has been corrected to ‘Natural substances developed as potential chemotherapeutic agents include components of mangrove leaves.’ ‘This extract has a mechanism for inhibiting the cell cycle at the G0-G1 phase…’Response: ‘This extract has a mechanism for inhibiting the cell cycle at the G0-G1 phase…’ has been revised to ‘Polyisoprenoids have a mechanism to inhibit the cell cycle at the G0-G1 phase…’‘In this study, the test material is cytotoxic for the polyisoprenoids of leaves derived from…’Response: ‘In this study, the test material is cytotoxic for the polyisoprenoids of leaves derived from…’has been corrected to ‘In this study, the cytotoxic test material was derived from polyisoprenoids of mangrove leaves (PAA and PAL)..’The word ‘simplicia’ is used in the text several times– please replace with any typical expression.The word ‘simplicia’ has been deleted to incorporate Reviewers’ suggestion throughout the revised manuscript. Minor remarks: Title – ‘using reverse transcription-PCR’ - in my opinion these last four words are not necessary and should be deleted; it is nothing special in this RT-PCR technique. Response: ‘using reverse transcription-PCR’ has been deleted from the title. p.1 – ‘Mangrove plants produce a polyisoprenoid compound. Polyisoprenoids have been proven to have anticancer properties.’ – these two sentences have to be rewritten. As it is now they do not provide enough information about the subject of study. Moreover, why the term ‘polyisoprenoid compound’ is used here in a singular form? It is not consistent either with the natural composition of the leaf extract or with the text below. ‘Mangrove plants produce polyisoprenoid compounds. Polyisoprenoids have been proven to have anticancer properties.’ p.1 – delete ‘inhibited’ in the sentence ‘The inhibited cell cycle and apoptosis...’. Response: ‘inhibited’ has been deleted. p.3 – rewrite the second sentence of Introduction – what is meant by the expression ‘halt its invasion and metastasis’ in the context of this sentence? Response: The second sentence of Introduction ‘halt its invasion and metastasis’ has been corrected to ‘spread its invasion and metastasis’ in the context of this sentence. p.3 – replace ‘…dolichol and polyprenol on the leaves…’ with ‘…dolichol and polyprenol in the leaves…’. Response: ‘…dolichol and polyprenol on the leaves…’ has been replaced with ‘…dolichol and polyprenol in the leaves…’. p.3 – ‘MCF-7 and T47D’ – a brief description of the cell line type is missing. Response: human breast cancer cell lines has been added before MCF-7 and T47D p.3 – ‘gene expression of COX-2 in colon cancer cells’ spell out the name of the enzyme when first used here and throughout the entire manuscript. Response: ‘gene expression of COX-2 in colon cancer cells’ has been corrected to ‘gene expression of cyclooxygenase-2 (COX-2) in colon cancer cells’ p.3 – delete the word ‘Every’ in the sentence ‘Every 500 g of powdered simplicia mangrove leaves…’. Explain the word ‘simplicia’ in the text or delete. Response: The words ‘Every’ and ‘simplicia’ in the sentence ‘Every 500 g of powdered simplicia mangrove leaves…’ has been deleted to read ‘Five hundreds g of powdered mangrove leaves…’ p.3 – ‘The cell wall debris insouble in CM21…’ – how about other cellular components, e.g. proteins? – replace ‘cell wall debris’ with ‘Precipitate’. Response: ‘The cell wall debris insouble in CM21…’ has been corrected to ‘Precipitate insoluble in CM21…’ p.3 – delete the description of the cleaning of the leaves from ‘The simplicia procedure …’ to ‘…stored in tightly closed plastic containers.’ Instead, describe the date of the collection of the leaves, for how long the plant material was stored prior to extraction, the temperature of the extraction procedure. Response: The simplicia procedure …’ to ‘…stored in tightly closed plastic containers.’ Has been deleted from the revised manuscript.Is the work clearly and accurately presented and does it cite the current literature? Partly Is the study design appropriate and is the work technically sound? Partly Are sufficient details of methods and analysis provided to allow replication by others? No If applicable, is the statistical analysis and its interpretation appropriate? No Are all the source data underlying the results available to ensure full reproducibility? Partly Are the conclusions drawn adequately supported by the results? Partly Competing InterestsNo competing interests were disclosed. We confirm that we have read this submission and believe that we have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however we have significant reservations, as outlined above."
}
]
}
] | 1
|
https://f1000research.com/articles/9-182
|
https://f1000research.com/articles/9-662/v1
|
30 Jun 20
|
{
"type": "Review",
"title": "SMARCB1/INI1-deficient tumors of adulthood",
"authors": [
"Nathaniel A. Parker",
"Ammar Al-Obaidi",
"Jeremy M. Deutsch",
"Ammar Al-Obaidi",
"Jeremy M. Deutsch"
],
"abstract": "The SMARCB1/INI1 gene was first discovered in the mid-1990’s, and since then it has been revealed that loss of function mutations in this gene result in aggressive rhabdoid tumors. Recently, the term “rhabdoid tumor” has become synonymous with decreased SMARCB1/INI1 expression. When genetic aberrations in the SMARCB1/INI1 gene occur, the result can cause reduced, complete loss, and mosaic expression. Although SMARCB1/INI1-deficient tumors are predominantly sarcomas, this is a diverse group of tumors with mixed phenotypes, which can often make the diagnosis challenging. Prognosis for these aggressive tumors is often poor. Moreover, refractory and relapsing progressive disease is common. As a result, accurate and timely diagnosis is imperative. Despite the SMARCB1/INI1 gene itself and its implications in tumorigenesis being discovered over two decades ago, there is a paucity of rhabdoid tumor cases reported in the literature that detail SMARCB1/INI1 expression. Much work remains if we hope to provide additional therapeutic strategies for patients with aggressive SMARCB1/INI1-deficient tumors.",
"keywords": [
"SMARCB1",
"INI1",
"loss of function mutation",
"rhabdoid",
"sarcoma"
],
"content": "History of the SMARCB1/INI1 Gene\n\nSWI/SNF-related matrix-associated actin-dependent regulator of chromatin subfamily B member 1 (SMARCB1), also known as integrase interactor 1 (INI1), is a crucial component of a chromatin-remodeling protein complex. SMARCB1/INI1 was first identified in yeast in the late 1980’s1. By 1994, its human homologue was isolated in fibroblast cells2,3. Subsequent molecular investigations showed this nuclear protein complex enhances DNA transcription by interactions with HIV-1 integrase2. Nuclear SMARCB1/INI1 exists ubiquitously in all normal cells, and acts as a tumor suppressor gene4. It was revealed in the early 2000’s by studies in mice that biallelic knockout of the SMARCB1/INI1 gene resulted in early lethality5. Mice with heterozygous loss before birth, or who had later conditional single-allele knockout after birth, of SMARCB1/INI1 developed aggressive rhabdoid tumors6–8. Since its discovery, much work has revealed this chromatin-remodeling protein has crucial roles in multiple signaling pathways that function to suppress tumorigenesis and tumor growth9. Although these pathways are highly complex, the development and use of targeted anti-cancer therapies has practically become ubiquitous for nearly all solid tumors. Thus, continued investigations are needed if we hope to provide additional therapeutic strategies for patients with aggressive SMARCB1/INI1-deficient tumors9.\n\nInterestingly, the genetic signatures of SMARCB1/INI1-deficient tumors are far from monotonous. Three distinct patterns of abnormal SMARCB1/INI1 gene expression have been identified – reduced, complete loss, and mosaic9.\n\n\nEpidemiology, clinical, prognosis\n\nComplete loss of SMARCB1/INI1 expression has been linked to a number of pediatric and adult sarcomas (Table 1). Malignant rhabdoid tumor (MRT) and epithelioid sarcoma (ES) both result from biallelic deletions or mutations causing a complete loss of SMARCB1/INI1 expression41. Commonly arising before the age of three years old, MRTs are considered one of the most aggressive childhood neoplasms associated with high mortality41. MRTs have been reported in adults42–49. Based on MRT of adulthood being primarily reported anecdotally, estimated rates of incidence remain unclear. Data concerning the 5-year survival rate for MRT in adults is difficult to determine as well, as various percentages have been reported in literature13,14. However, estimated average survival following MRT diagnosis has been reported to be six months10.\n\nSTS, soft tissue sarcomas; MRT, malignant rhabdoid tumor; MPNST, malignant peripheral nerve sheath tumor; NF-1, neurofibromatosis type 1; NF-2; neurofibromatosis type 2, GI, gastrointestinal; NA, data not available.\n\nES is now categorized into two subgroups: distal and proximal. Conventional or distal-type ES tends to be histologically similar squamous cells. Also, distal-type ES immunohistochemical (IHC) profiles can be diverse. Proximal-type ES is thought to be the more aggressive variant, and has an affinity for the proximal limbs of young adults. Microscopically, sheets of large rhabdoid tumor cells are predominantly observed50. Based on more recent clinicopathologic and IHC data, many tumors that were previously diagnosed as a MRT are now classified as proximal ES51.\n\nIn addition to ES, atypical teratoid/rhabdoid tumor, renal medullary carcinoma, and pediatric chordoma are rare sarcomas that result from the complete loss of SMARCB1/INI1 expression (Table 1). They predominantly occur in pediatric or young adult patients. Collectively, these neoplasms typically develop in the head/neck, CNS, thorax, kidneys, other visceral organs, retroperitoneum, trunk, and extremities12,17,19,26,52. Exceedingly rare SMARCB1/INI1-deficient tumors that occur more commonly in adults include synovial sarcomas, epithelioid malignant peripheral nerve sheath tumor, myoepithelial carcinoma, extraskeletal myxoid chondrosarcoma, chordoma, schwannomatosis, gastrointestinal stromal tumors (GIST), and ossifying fibromyxoid tumor (Table 1). On light microscopy, these sarcomatous neoplasms exist on a morphological spectrum. Tissue specimens are often composed of epithelioid or rhabdoid cells53. However, other morphologic patterns have been described50. Thus, the diagnosis of SMARCB1/INI1-deficient tumors can be difficult based on their polyphenotypic variation4. SMARCB1/INI1 immunostaining can be used to confirm the diagnosis of an epithelioid or rhabdoid sarcoma because loss of SMARCB1/INI1 expression is rarely observed in other tumor types54,55. Thus, in the absence of this genetic alteration, other malignant soft tissue tumors with epithelioid-like morphologies can be more confidently ruled out, such as melanoma, rhabdomyosarcoma, and undifferentiated carcinoma.\n\nAside from SMARCB1/INI1-deficient tumors sharing an aberration in the same gene, the relationship between these malignancies remains unclear. Following diagnosis in any age or organ, nearly all SMARCB1/INI1-deficient malignancies characteristically follow an aggressive clinical pattern and prognosis if often poor (Table 1). Survival rates are often reportedly low, but they may not be accurate given low rates of incidence, and considerations for newer treatments. Also, survival can be highly dependent on surgical intervention and completeness of tumor resection, especially for chordomas. GIST are the most common sarcomas of the gastrointestinal (GI) tract. They commonly develop in the sixth decade of life and have no gender predominance35. Following the diagnosis of a GIST, survival rates are highly variable and depend on specific biologic characteristics of the tumor, the type of treatment, and the risk of post-treatment recurrence36.\n\n\nChallenges in retrospective data collection for adult cases of SMARCB1/INI-deficient tumors\n\nRecently, the term “rhabdoid tumor” has become synonymous with tumors that harbor loss of function mutations in the SMARCB1/INI1 gene56. We reviewed the literature and found a paucity of cases reporting SMARCB1/INI1 genetic aberrations in adult patients with sarcomas. A total of 450 cases of rare sarcomas were found to be described in single case reports, case series, or systematic reviews published between the years 2000 – 2020 (Table 2)57–92. This number is likely far lower than the actual accounts of reported sarcoma cases in the literature. However, reports were excluded if it was apparent the case did not meet our inclusion criteria based on the publicly-available title or abstract information. Despite the SMARCB1/INI1 gene being discovered in the mid-1990’s, the majority of previous reports were excluded for not mentioning the tumor’s SMARCB1/INI1-deficiency status. Also, tumor occurrence in the pediatric patient population accounted for multiple exclusions.\n\nExclusion criteria was as follows: 1.) individual patient age could not be confirmed; 2.) pediatric study population (less than 18 years of age); 3.) absence of documentation noting the loss of SMARCB1/INI1 expression by immunohistochemistry or genetic studies; 4.) intact SMARCB1/INI1 expression by immunochemistry or genetic studies; and 5.) non-sarcomatous histologic tumor type. PMID, PubMed Central © unique article identifier; GU, genitourinary; PNS, peripheral nervous system; GI, gastrointestinal.\n\nWe located 25 cases of adult SMARCB1/INI1-deficient sarcomas that were described in 18 reports (Table 3)42,50,93–108. Median age at the time of diagnosis was 36 years old. A male predominance was mildly observed (14 cases, 56%), which is consistent with other larger reviews. Presentation in the head and neck (e.g. brain, eye, nose, and scalp) occurred more frequently (6 cases, 24%). No descriptive data analysis was performed to determine if our observations were significant. The majority of reports were originally described as proximal epithelioid sarcoma, but overall these remained a morphologically diverse group of cases that also included rhabdoid and mixed phenotypes.\n\nInclusion criteria was as follows: ability to confirm an individual case patient was greater than 18 years of age; documentation of a loss of SMARCB1/INI1 expression by immunohistochemistry or genetic studies; and confirmed sarcomatous histologic tumor type. “ - “ denotes complete, reduced, or mosaic loss of SMARCB1/INI1 expression (exp.). M, male; F, female.\n\n\nTreatment\n\nPrior to, and still after, the discovery that SMARCB1/INI1-deficient tumors contribute to the large majority of soft tissue sarcomas, systemic cytotoxic agents have been used to treat this diverse group of neoplasms. Doxorubicin and ifosfamide have remained the mainstay of first-line treatment for advanced disease for the last few decades. Currently, the most widely used regimen for soft tissue sarcomas is termed AIM, which includes doxorubicin plus ifosfamide and mesna109–111. Therapies such as these, and other cytotoxic agents, exhibit intermediate to improved anti-cancer activity, and prolong survival in metastatic soft tissue sarcoma (Table 4). However, refractory or progressive disease can occur. With the hopes of improving outcomes in patients who develop aggressive sarcomas, multiple new therapies are being introduced. Olaratumab, a monoclonal antibody that targets platelet-derived growth factor alpha and beta (PDGFRA/B), has been approved for first-line therapy in combination with doxorubicin due to improved progression and overall survival in sarcoma patients112. The use of tyrosine kinase-inhibitors (TKIs) has transformed the treatment of advanced GIST. Imatinib, a TKI, as monotherapy is now approved for upfront treatment of metastatic GIST due to improved side effect profiles and outcomes in these patients113–115. Given its mechanism of action, imatinib is also approved for first-line treatment of the fibrosarcomatous variant of dermatofibrosarcoma protuberans116,117.\n\nORR, overall response rate; PFS, progression free survival; OS, overall survival; STS, soft tissue sarcoma; GIST, gastrointestinal stromal tumor; D, doxorubicine; I, ifosfamide; P, palifosfamide; E, Evofosfamide; T, trabectedin; O, Olaratumab; G, gemcitabine; Doc, docetaxel; NA, data not available.\n\nAdditional TKIs have recently been introduced, with clinical trial data showing promise for their use in sarcomas. Sunitinib and regorafenib significantly improve overall survival in imatinib-resistant GIST patients118. Pazopanib, a TKI that targets angiogenesis by inhibiting vascular endothelial growth factor receptor, PDGFRA/B, and KIT, has been shown to improve progression free survival in certain histologic types of sarcoma. This led to its approval for advanced, refractory non-lipomatous sarcoma119,120. Alveolar sarcomas appear to respond well to anti-angiogenetic sorafenib and cediranib121,122. In phase II studies tivozanib, which mechanism of action mimics pazopanib, exhibits promising anti-cancer activity in metastatic or nonresectable soft tissue sarcomas123.\n\nRecently, much work studying the complex mechanisms involved in sarcoma tumorigenesis has revealed the potential for numerous new drug targets. Targeting mTOR by serine/threonine kinase inhibition has been widely studied. However, thus far either only equivocal or minor benefits have been shown with the administration of these agents124. In contrast, phase II trial data is reassuring for the future use of palbociclib, a cyclin-dependent kinase 4 and 6 inhibitor approved in breast cancer, for liposarcoma125,126.\n\nPreliminary data from pre-clinical and phase I/II trials is encouraging for small molecule inhibitors, such as with MDM2-antagonists, histone deacetylase inhibitors, and histone methylation inhibitors124. A possible breakthrough in small molecular inhibition is represented by the recent discovery of histone-lysine N-methyltransferase EZH2 upregulation in SMARCB1/INI1-deficient tumors127. Given the defining characteristic of SMARCB1/INI1 deficiency in the nearly all soft tissue sarcomas, tazemetostat has emerged as a highly intriguing compound for its direct inhibition of histone-lysine N-methyltransferase EZH2138,139. Another new agent that hopes to improve outcomes for patients with these rare and aggressive SMARCB1/INI1-deficient rhabdoid sarcomas comes from the proteasome inhibitor drug class. Ixazomib selectively targets proteasomes involved in protein anabolism and cellular apoptosis, whose activity is directly enhanced by the transcription factor MYC in SMARCB1/INI1-deficient states. Currently, ixazomib plus gemcitabine and doxorubicin is being studied in the phase II trial setting for renal medullary carcinoma140,141.\n\n\nData availability\n\nNo data are associated with this article.",
"appendix": "References\n\nAbrams E, Neigeborn L, Carlson M: Molecular analysis of SNF2 and SNF5, genes required for expression of glucose-repressible genes in Saccharomyces cerevisiae. Mol Cell Biol. 1986; 6(11): 3643–3651. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKalpana GV, Marmon S, Wang W, et al.: Binding and stimulation of HIV-1 integrase by a human homolog of yeast transcription factor SNF5. Science. 1994; 266(5193): 2002–2006. PubMed Abstract | Publisher Full Text\n\nMuchardt C, Sardet C, Bourachot B, et al.: A human protein with homology to Saccharomyces cerevisiae SNF5 interacts with the potential helicase hbrm. Nucleic Acids Res. 1995; 23(7): 1127–1132. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHollmann TJ, Hornick JL: INI1-deficient tumors: diagnostic features and molecular genetics. Am J Surg Pathol. 2011; 35(10): e47–63. PubMed Abstract | Publisher Full Text\n\nGuidi CJ, Sands AT, Zambrowicz BP, et al.: Disruption of Ini1 leads to peri-implantation lethality and tumorigenesis in mice. Mol Cell Biol. 2001; 21(10): 3598–3603. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoberts CW, Orkin SH: The SWI/SNF complex--chromatin and cancer. Nat Rev Cancer. 2004; 4(2): 133–142. PubMed Abstract | Publisher Full Text\n\nRoberts CW, Galusha SA, McMenamin ME, et al.: Haploinsufficiency of Snf5 (integrase interactor 1) predisposes to malignant rhabdoid tumors in mice. Proc Natl Acad Sci U S A. 2000; 97(25): 13796–13800. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoberts CW, Leroux MM, Fleming MD, et al.: Highly penetrant, rapid tumorigenesis through conditional inversion of the tumor suppressor gene Snf5. Cancer Cell. 2002; 2(5): 415–425. PubMed Abstract | Publisher Full Text\n\nKohashi K, Oda Y: Oncogenic roles of SMARCB1/INI1 and its deficient tumors. Cancer Sci. 2017; 108(4): 547–552. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFerrari A, Sultan I, Huang TT, et al.: Soft tissue sarcoma across the age spectrum: a population-based study from the Surveillance Epidemiology and End Results database. Pediatr Blood Cancer.. 2011; 57(6): 943–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRare Soft Tissue Tumors: Synovial Sarcoma. National Cancer Institute Web site. 2019. Reference Source\n\nShonka NA, Armstrong TS, Prabhu SS, et al.: Atypical teratoid/rhabdoid tumors in adults: a case report and treatment-focused review. J Clin Med Res. 2011; 3(2): 85–92. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDoyle LA: Malignant rhabdoid tumor: Pathology of Human Disease. Academic press; 2014.\n\nHorazdovsky R, Manivel JC, Cheng EY: Successful salvage and long-term survival after recurrent malignant rhabdoid tumor. Sarcoma. 2007; 2007: 53549. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGeller JI, Waltz AL: Malignant Rhabdoid Tumor. Medscape. 2018. Reference Source\n\nChan V, Marro A, Findlay JM, et al.: A Systematic Review of Atypical Teratoid Rhabdoid Tumor in Adults. Front Oncol. 2018; 8: 567. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNeeds T, Fillman EP: Cancer, Epithelioid Sarcoma. StatPearls. Treasure Island (FL): StatPearls. 2020. PubMed Abstract\n\nJawad MU, Extein J, Min ES, et al.: Prognostic factors for survival in patients with epithelioid sarcoma: 441 cases from the SEER database. Clin Orthop Relat Res. 2009; 467(11): 2939–48. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShetty A, Matrana MR: Renal medullary carcinoma: a case report and brief review of the literature. Ochsner J. 2014; 14(2): 270–5. PubMed Abstract | Free Full Text\n\nMsaouel P: Renal Medullary Carcinoma. National organization for Rare Disorders. 2019. Reference Source\n\nDu P, Zhu J, Zhang ZD, et al.: Recurrent epithelioid malignant peripheral nerve sheath tumor with neurofibromatosis type 1: A case report and literature review. Oncol Lett. 2019; 18(3): 3072–3080. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMinagawa T, Shioya R, Sato C, et al.: Advanced epithelioid malignant peripheral nerve sheath tumor showing complete response to combined surgery and chemotherapy: a case report. Case Rep Oncol Med. 2011; 2011: 705345. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSalivary Gland Cancer Treatment: Myoepithelial carcinoma. National Cancer Institute Web site. 2019. Reference Source\n\nSu YX, Roberts DB, Hanna EY, et al.: Risk Factors and Prognosis for Myoepithelial Carcinoma of the Major Salivary Glands. Ann Surg Oncol. 2015; 22(11): 3701–7. PubMed Abstract | Publisher Full Text\n\nPaoluzzi L, Ghesani M: Extraskeletal myxoid chondrosarcoma with massive pulmonary metastases. Clin Sarcoma Res. 2018; 8: 20. Published 2018 Dec 5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBeccaria K, Sainte-Rose C, Zerah M, et al.: Paediatric Chordomas. Orphanet J Rare Dis. 2015; 10: 116. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRare Bone Tumors: Chordoma. National Cancer Institute Web site. 2019; Reference Source\n\nSmoll NR, Gautschi OP, Radovanovic I, et al.: Incidence and relative survival of chordomas: the standardized mortality ratio and the impact of chordomas on a population. Cancer. 2013; 119(11): 2029–37. PubMed Abstract | Publisher Full Text\n\nAgaimy A, Haller F, Frohnauer J, et al.: Pancreatic undifferentiated rhabdoid carcinoma: KRAS alterations and SMARCB1 expression status define two subtypes. Mod Pathol. 2015; 28(2): 248–60. PubMed Abstract | Publisher Full Text\n\nAgaimy A, Koch M, Lell M, et al.: SMARCB1(INI1)-deficient sinonasal basaloid carcinoma: a novel member of the expanding family of SMARCB1-deficient neoplasms. Am J Surg Pathol. 2014; 38(9): 1274–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee JS, Ko IJ, Jun SY, et al.: Basaloid squamous cell carcinoma in nasal cavity. Clin Exp Otorhinolaryngol. 2009; 2(4): 207–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nÖfkeli Ö, Bulut A, Yüksek YN, et al.: Malignant rhabdoid tumour in the stomach. Ulus Cerrahi Derg. 2013; 31(1): 55–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRomera Barba E, Sánchez Pérez A, Duque Pérez C, et al.: Malignant rhabdoid tumor of the colon: a case report. Cir Esp. 2014; 92(9): 638–40. PubMed Abstract | Publisher Full Text\n\nYohay K, Bergner A: Schwannomatosis. UpToDate. 2020. Reference Source\n\nMorgan J, Raut CP, Duensing A, et al.: Epidemiology, classification, clinical presentation, prognostic features, and diagnostic work-up of gastrointestinal stromal tumors (GIST). UpToDate. 2019.Reference Source\n\nDeMatteo RP, Lewis JJ, Leung D, et al.: Two hundred gastrointestinal stromal tumors: recurrence patterns and prognostic factors for survival. Ann Surg. 2000; 231(1): 51–58. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJoensuu H, Fletcher C, Dimitrijevic S, et al.: Management of malignant gastrointestinal stromal tumours. Lancet Oncol. 2002; 3(11): 655–664. PubMed Abstract | Publisher Full Text\n\nHowlader N, Noone AM, Krapcho M, et al.: (eds). SEER Cancer Statistics Review, 1975-2016. National Cancer Institute, Bethesda, MD, based on November 2018 SEER data submission, posted to the SEER website, April. 2019. Reference Source\n\nVelasco IA, Zhang R, Li T, et al.: Ossifying Fibromyxoid tumor of soft parts in head and neck: case report and literature review. Diagn Pathol. 2018; 13(1): 21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBakiratharajan D, Rekhi B: Ossifying Fibromyxoid Tumor: An Update. Arch Pathol Lab Med. 2016; 140(4): 371–375. Publisher Full Text\n\nPawel BR: SMARCB1-deficient Tumors of Childhood: A Practical Guide. Pediatr Dev Pathol. 2018; 21(1): 6–28. PubMed Abstract | Publisher Full Text\n\nParker NA, Al-Obaidi A, Lalich D, et al.: SMARCB1/INI1-Deficient Extrarenal Rhabdoid Tumor: A Case Report of a Rare and Aggressive Soft Tissue Sarcoma. Cureus. 2020; 12(5): e8273. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPodduturi V, Campa-Thompson MM, Zhou XJ, et al.: Malignant rhabdoid tumor of the kidney arising in an adult patient. Proc (Bayl Univ Med Cent). 2014; 27(3): 239–41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPeng HQ, Stanek AE, Teichberg S, et al.: Malignant rhabdoid tumor of the kidney in an adult: a case report and review of the literature. Arch Pathol Lab Med. 2003; 127(9): e371–e373. PubMed Abstract | Publisher Full Text\n\nLowe W, Weiss RM, Todd MB, et al.: Malignant rhabdoid tumor of the kidney in an adult. J Urol. 1990; 143(1): 110–112. PubMed Abstract | Publisher Full Text\n\nClausen HV, Horn T, Anagnostaki L, et al.: Malignant rhabdoid tumour of the kidney in an adult: a case report with immunohistochemical and ultrastructural investigation. Scand J Urol Nephrol Suppl. 1994; 157: 123–128. PubMed Abstract\n\nEbbinghaus SW, Herrera G, Marshall ME: Rhabdoid tumor of the kidney in an adult: review of the literature and report of a case responding to interleukin-2. Cancer Biother. 1995; 10(3): 237–241. PubMed Abstract | Publisher Full Text\n\nCaballero JM, Collera P, Marti L, et al.: [Malignant Rhabdoid Tumor of the Kidney in the Adult]. Actas Urol Esp. 1996; 20(4): 377–379. PubMed Abstract\n\nZhao G, Na R, Yang Y, et al.: Pure malignant rhabdoid tumor of the left kidney in an adult: A case report and review of the literature. Oncol Lett. 2013; 5(5): 1481–1484. PubMed Abstract | Publisher Full Text | Free Full Text\n\nModena P, Lualdi E, Facchinetti F, et al.: SMARCB1/INI1 tumor suppressor gene is frequently inactivated in epithelioid sarcomas. Cancer Res.. 2005; 65(10): 4012–9. PubMed Abstract | Publisher Full Text\n\nMargol AS, Judkins AR: Pathology and diagnosis of SMARCB1-deficient tumors. Cancer Genet. 2014; 207(9): 358–64. PubMed Abstract | Publisher Full Text\n\nNg WK, Toe BP, Lau HY: Malignant Rhabdoid Tumor of the Mediastinum: A Case Report and Literature Review. J Clin Imaging Sci. 2019; 9: 7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGupta RK, Batra VV, Das MC, et al.: Malignant extra-renal rhabdoid tumor with unusual presentation: A report of two cases. J Cancer Res Ther. 2015; 11(4): 963–6. PubMed Abstract | Publisher Full Text\n\nHoot AC, Russo P, Judkins AR, et al.: Immunohistochemical analysis of hSNF5/INI1 distinguishes renal and extra-renal malignant rhabdoid tumors from other pediatric soft tissue tumors. Am J Surg Pathol.. 2004; 28(11): 1485–91. PubMed Abstract | Publisher Full Text\n\nChbani L, Guillou L, Terrier P, et al.: Epithelioid sarcoma: a clinicopathologic and immunohistochemical analysis of 106 cases from the French sarcoma group. Am J Clin Pathol. 2009; 131(2): 222–227. PubMed Abstract | Publisher Full Text\n\nKohashi K, Tanaka Y, Kishimoto H, et al.: Reclassification of rhabdoid tumor and pediatric undifferentiated/unclassified sarcoma with complete loss of SMARCB1/INI1 protein expression: three subtypes of rhabdoid tumor according to their histological features. Mod Pathol. 2016; 29(10): 1232–1242. Publisher Full Text\n\nZhang XW, Deng YJ, Zhou L, et al.: Epithelioid sarcoma of the scalp: a case report and literature review. Int J Clin Exp Pathol. 2019; 12(10): 3908–3914. PubMed Abstract | Free Full Text\n\nKubota Y, Arakawa Y, Sekiguchi M, et al.: A case of malignant rhabdoid tumor mimicking yolk sac tumor. Pediatr Blood Cancer. 2019; 66(8): e27784. PubMed Abstract | Publisher Full Text\n\nKolin DL, Dong F, Baltay M, et al.: SMARCA4-deficient undifferentiated uterine sarcoma (malignant rhabdoid tumor of the uterus): a clinicopathologic entity distinct from undifferentiated carcinoma. Mod Pathol. 2018; 31(9): 1442–1456. PubMed Abstract | Publisher Full Text\n\nKim YS, Kwak HS, Chung GH, et al.: Epithelioid sarcoma arising from the temporal space: A case report. Medicine (Baltimore). 2018; 97(38): e12529. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStrehl JD, Wachter DL, Fiedler J, et al.: Pattern of SMARCB1 (INI1) and SMARCA4 (BRG1) in poorly differentiated endometrioid adenocarcinoma of the uterus: analysis of a series with emphasis on a novel SMARCA4-deficient dedifferentiated rhabdoid variant. Ann Diagn Pathol. 2015; 19(4): 198–202. PubMed Abstract | Publisher Full Text\n\ndos Santos LM, Nogueira L, Matsuo CY, et al.: Proximal-type epithelioid sarcoma -- case report. An Bras Dermatol. 2013; 88(3): 444–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPatrizi L, Corrado G, Saltari M, et al.: Vulvar \"proximal-type\" epithelioid sarcoma: report of a case and review of the literature. Diagn Pathol. 2013; 8: 122. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTocco I, Bassetto F, Vindigni V: Primary epithelioid sarcoma of the scalp complicated by humoral hypercalcemia of malignancy. Eplasty. 2012; 12: ic18. PubMed Abstract | Free Full Text\n\nRizzo D, Fréneaux P, Brisse H, et al.: SMARCB1 deficiency in tumors from the peripheral nervous system: a link between schwannomas and rhabdoid tumors? Am J Surg Pathol. 2012; 36(7): 964–72. PubMed Abstract | Publisher Full Text\n\nKuge A, Sato S, Sakurada K, et al.: Atypical teratoid rhabdoid tumor located in the pineal region following prophylactic irradiation for acute lymphoblastic leukemia. Brain Tumor Pathol. 2012; 29(3): 177–81. PubMed Abstract | Publisher Full Text\n\nHagström J, Mesimäki K, Apajalahti S, et al.: A rare case of oral epithelioid sarcoma of the gingiva. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2011; 111(4): e25–8. PubMed Abstract | Publisher Full Text\n\nNarendra H, Ray S, Rao L, et al.: Malignant extrarenal rhabdoid tumor of the vulva in an adult. J Cancer Res Ther. 2010; 6(1): 82–5. PubMed Abstract | Publisher Full Text\n\nTholpady A, Lonergan CL, Wick MR: Proximal-type epithelioid sarcoma of the vulva: relationship to malignant extrarenal rhabdoid tumor. Int J Gynecol Pathol. 2010; 29(6): 600–4. PubMed Abstract | Publisher Full Text\n\nHornick JL, Dal Cin P, Fletcher CDM: Loss of INI1 expression is characteristic of both conventional and proximal-type epithelioid sarcoma. Am J Surg Pathol. 2009; 33(4): 542–50. PubMed Abstract | Publisher Full Text\n\nKim JH, Choi YS, Lee TS: A case of epithelioid sarcoma arising in the vulva. J Gynecol Oncol. 2008; 19(3): 202–204. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRekhi B, Gorad BD, Chinoy RF: Clinicopathological features with outcomes of a series of conventional and proximal-type epithelioid sarcomas, diagnosed over a period of 10 years at a tertiary cancer hospital in India. Virchows Arch. 2008; 453(2): 141–153. PubMed Abstract | Publisher Full Text\n\nArgenta PA, Thomas S, Chura JC: Proximal-type epithelioid sarcoma vs. malignant rhabdoid tumor of the vulva: a case report, review of the literature, and an argument for consolidation. Gynecol Oncol. 2007; 107(1): 130–135. PubMed Abstract | Publisher Full Text\n\nBourdeaut F, Fréneaux P, Thuille B, et al.: hSNF5/INI1-deficient tumours and rhabdoid tumours are convergent but not fully overlapping entities. J Pathol. 2007; 211(3): 323–330. PubMed Abstract | Publisher Full Text\n\nWeisskopf M, Münker R, Hermanns-Sachweh B, et al.: Epithelioid sarcoma in the thoracic spine. Eur Spine J. 2006; 15 Suppl 5(Suppl 5): 604–609. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOnol FF, Tanidir Y, Kotiloğlu E, et al.: Proximal type epithelioid sarcoma of the scrotum: a source of diagnostic confusion that needs immediate attention. Eur Urol. 2006; 49(2): 406–407. PubMed Abstract | Publisher Full Text\n\nZevallos-Giampietri EA, Barrionuevo C: Proximal-type epithelioid sarcoma: report of two cases in the perineum: differential diagnosis and review of soft tissue tumors with epithelioid and/or rhabdoid features. Appl Immunohistochem Mol Morphol. 2005; 13(3): 221–230. PubMed Abstract | Publisher Full Text\n\nMasunaga A, Ikeda K, Suzuki T, et al.: Proximal-type epithelioid sarcoma in a 36-year-old man: closer immunoelectron-microscopic resemblance of the tumor cells to epithelial cells than to mesenchymal cells. Pathol Int. 2004; 54(8): 616–622. PubMed Abstract | Publisher Full Text\n\nChang JH, Dikranian AH, Johnston WH, et al.: Malignant extrarenal rhabdoid tumor of the bladder: 9-year survival after chemotherapy and partial cystectomy. J Urol. 2004; 171(2 Pt 1): 820–821. PubMed Abstract | Publisher Full Text\n\nAltundag K, Dikbas O, Oyan B, et al.: Epithelioid sarcoma of vulva: a case report and review of the literature. Med Oncol. 2004; 21(4): 367–372. PubMed Abstract | Publisher Full Text\n\nLee MW, Jee KJ, Ro JY, et al.: Proximal-type epithelioid sarcoma: case report and result of comparative genomic hybridization. J Cutan Pathol. 2004; 31(1): 67–71. PubMed Abstract | Publisher Full Text\n\nHanna SL, Kaste S, Jenkins JJ, et al.: Epithelioid sarcoma: clinical, MR imaging and pathologic findings. Skeletal Radiol. 2002; 31(7): 400–412. PubMed Abstract | Publisher Full Text\n\nEtienne-Mastroianni B, Falchero L, Chalabreysse L, et al.: Primary sarcomas of the lung: a clinicopathologic study of 12 cases. Lung Cancer. 2002; 38(3): 283–289. PubMed Abstract | Publisher Full Text\n\nMoore RG, Steinhoff MM, Granai CO, et al.: Vulvar epithelioid sarcoma in pregnancy. Gynecol Oncol. 2002; 85(1): 218–222. PubMed Abstract | Publisher Full Text\n\nHaidopoulos D, Elsheikh A, Vlahos G, et al.: Malignant rhabdoid tumor of the clitoris in an elderly patient: report of a case. Eur J Gynaecol Oncol. 2002; 23(5): 447–449. PubMed Abstract\n\nTzilinis A, Clarke LE, Affuso C, et al.: Successful treatment of malignant rhabdoid tumor of the vulva in an older patient: a case report and review of the literature. Curr Surg. 2002; 59(6): 570–571. PubMed Abstract | Publisher Full Text\n\nAmrikachi M, Ro JY, Ordonez NG, et al.: Adenocarcinomas of the gastrointestinal tract with prominent rhabdoid features. Ann Diagn Pathol. 2002; 6(6): 357–363. PubMed Abstract | Publisher Full Text\n\nHasegawa T, Matsuno Y, Shimoda T, et al.: Proximal-type epithelioid sarcoma: a clinicopathologic study of 20 cases. Mod Pathol. 2001; 14(7): 655–663. PubMed Abstract | Publisher Full Text\n\nKasamatsu T, Hasegawa T, Tsuda H, et al.: Primary epithelioid sarcoma of the vulva. Int J Gynecol Cancer. 2001; 11(4): 316–320. PubMed Abstract | Publisher Full Text\n\nKnapik J, Yachnis AT, Ripley D, et al.: Aggressive uterine sarcoma with rhabdoid features: diagnosis by peritoneal fluid cytology and absence of INI1 gene mutation. Hum Pathol. 2001; 32(8): 884–886. PubMed Abstract | Publisher Full Text\n\nBiegel JA, Fogelgren B, Wainwright LM, et al.: Germline INI1 mutation in a patient with a central nervous system atypical teratoid tumor and renal rhabdoid tumor. Genes Chromosomes Cancer. 2000; 28(1): 31–37. PubMed Abstract | Publisher Full Text\n\nSpillane AJ, Thomas JM, Fisher C: Epithelioid sarcoma: the clinicopathological complexities of this rare soft tissue sarcoma. Ann Surg Oncol. 2000; 7(3): 218–225. PubMed Abstract | Publisher Full Text\n\nAhmad Z, Stanazai Q, Wright S, et al.: Primary pleural epithelioid sarcoma of the proximal type: a diagnostic and therapeutic challenge. Transl Lung Cancer Res. 2019; 8(5): 700–705. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBodi I, Giamouriadis A, Sibtain N, et al.: Primary intracerebral INI1-deficient rhabdoid tumor with CD34 immunopositivity in a young adult. Surg Neurol Int. 2018; 9: 45. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGurwale S, Gore C, Kaur S, et al.: Primary epithelioid sarcoma of scalp. Asian J Oncol. 2017; 3: 71–73. Publisher Full Text\n\nSaha D, Basu A, Maiti A, et al.: Primary proximal epithelioid sarcoma of the lung successfully treated with pneumonectomy and adjuvant chemotherapy. BMJ Case Rep. 2016; 2016: bcr2015213966. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRego JL, Cintra GF, Netto AKJ, et al.: Extremely rare case of vulvar myxoid epithelioid sarcoma. Case Rep Obstet Gynecol. 2015; 2015: 971217. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWetzel SL, Kerpel S, Reich RF, et al.: Malignant rhabdoid tumor of the floor of mouth: first reported case in the oral cavity of an adult. Head Neck Pathol. 2015; 9(2): 269–272. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAgaimy A, Rau TT, Hartmann A, et al.: SMARCB1 (INI1)-negative rhabdoid carcinomas of the gastrointestinal tract: clinicopathologic and molecular study of a highly aggressive variant with literature review. Am J Surg Pathol. 2014; 38(7): 910–920. PubMed Abstract | Publisher Full Text\n\nMadsen GA, Rasmussen TR, Bærentzen S: A case of pleural epithelioid sarcoma of proximal type presenting as malignant pleural mesothelioma. J Thorac Oncol. 2013; 8(10): e89–e90. PubMed Abstract | Publisher Full Text\n\nFrank R, Sadri N, Bhatti T, et al.: Proximal-type Epithelioid Sarcoma of the Head and Neck (HN): A Study with Immunohistochemical and Molecular Analysis of SMARCB1. J Clin Exp Oncol. 2013; 2(2): 1000106. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim HJ, Kim MH, Kwon J, et al.: Proximal-type epithelioid sarcoma of the vulva with INI1 diagnostic utility. Ann Diagn Pathol. 2012; 16(5): 411–415. PubMed Abstract | Publisher Full Text\n\nMannan AA, Rifaat AA, Kahvic M, et al.: Proximal-type epithelioid sarcoma in the groin presenting as a diagnostic dilemma. Pathol Oncol Res. 2010; 16(2): 181–188. PubMed Abstract | Publisher Full Text\n\nTakei H, Adesina AM, Mehta V, et al.: Atypical teratoid/rhabdoid tumor of the pineal region in an adult. J Neurosurg. 2010; 113(2): 374–379. PubMed Abstract | Publisher Full Text\n\nRaoux D, Péoc'h M, Pedeutour F, et al.: Primary epithelioid sarcoma of bone: report of a unique case, with immunohistochemical and fluorescent in situ hybridization confirmation of INI1 deletion. Am J Surg Pathol. 2009; 33(6): 954–958. PubMed Abstract | Publisher Full Text\n\nRobbens C, Vanwyck R, Wilms G, et al.: An extrarenal rhabdoid tumor of the cervical spine with bony involvement. Skeletal Radiol. 2007; 36(4): 341–345. PubMed Abstract | Publisher Full Text\n\nSigauke E, Rakheja D, Maddox DL, et al.: Absence of expression of SMARCB1/INI1 in malignant rhabdoid tumors of the central nervous system, kidneys and soft tissue: an immunohistochemical study with implications for diagnosis. Mod Pathol. 2006; 19(5): 717–725. PubMed Abstract | Publisher Full Text\n\nPerry A, Fuller CE, Judkins AR, et al.: INI1 expression is retained in composite rhabdoid tumors, including rhabdoid meningiomas. Mod Pathol. 2005; 18(7): 951–958. PubMed Abstract | Publisher Full Text\n\nEdmonson JH, Ryan LM, Blum RH, et al.: Randomized comparison of doxorubicin alone versus ifosfamide plus doxorubicin or mitomycin, doxorubicin, and cisplatin against advanced soft tissue sarcomas. J Clin Oncol. 1993; 11(7): 1269–75. PubMed Abstract | Publisher Full Text\n\nCesne A, Judson I, Crowther D, et al.: Randomized phase III study comparing conventional-dose doxorubicin plus ifosfamide versus high-dose doxorubicin plus ifosfamide plus recombinant human granulocyte-macrophage colony-stimulating factor in advanced soft tissue sarcomas: A trial of the European Organization for Research and Treatment of Cancer/Soft Tissue and Bone Sarcoma Group. J Clin Oncol. 2000; 18(14): 2676–84. PubMed Abstract | Publisher Full Text\n\nWorden FP, Taylor JM, Biermann JS, et al.: Randomized phase II evaluation of 6 g/m2 of ifosfamide plus doxorubicin and granulocyte colony-stimulating factor (G-CSF) compared with 12 g/m2 of ifosfamide plus doxorubicin and G-CSF in the treatment of poor-prognosis soft tissue sarcoma. J Clin Oncol. 2005; 23(1): 105–12. PubMed Abstract | Publisher Full Text\n\nTap WD, Jones RL, Van Tine BA, et al.: Olaratumab and doxorubicin versus doxorubicin alone for treatment of soft-tissue sarcoma: an open-label phase 1b and randomised phase 2 trial. [published correction appears in Lancet. 2016 Jul 30; 388(10043):464]. Lancet. 2016; 388(10043): 488–497. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvan Oosterom AT, Judson I, Verweij J, et al.: Safety and efficacy of imatinib (STI571) in metastatic gastrointestinal stromal tumours: a phase I study. Lancet. 2001; 358(9291): 1421–1423. PubMed Abstract | Publisher Full Text\n\nDemetri GD, von Mehren M, Blanke CD, et al.: Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl J Med. 2002; 347(7): 472–480. PubMed Abstract | Publisher Full Text\n\nBlanke CD, Rankin C, Demetri GD, et al.: Phase III randomized, intergroup trial assessing imatinib mesylate at two dose levels in patients with unresectable or metastatic gastrointestinal stromal tumors expressing the kit receptor tyrosine kinase: S0033. J Clin Oncol. 2008; 26(4): 626–632. PubMed Abstract | Publisher Full Text\n\nRutkowski P, Van Glabbeke M, Rankin CJ, et al.: Imatinib mesylate in advanced dermatofibrosarcoma protuberans: pooled analysis of two phase II clinical trials. J Clin Oncol. 2010; 28(10): 1772–1779. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStacchiotti S, Pantaleo MA, Negri T, et al.: Efficacy and Biological Activity of Imatinib in Metastatic Dermatofibrosarcoma Protuberans (DFSP). Clin Cancer Res. 2016; 22(4): 837–846. PubMed Abstract | Publisher Full Text\n\nCall JW, Wang Y, Montoya D, et al.: Survival in advanced GIST has improved over time and correlates with increased access to post-imatinib tyrosine kinase inhibitors: results from Life Raft Group Registry. [published correction appears in Clin Sarcoma Res. 2019 May 7;9:7]. Clin Sarcoma Res. 2019; 9: 4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvan der Graaf WTA, Blay JY, Chawla SP, et al.: Pazopanib for metastatic soft-tissue sarcoma (PALETTE): a randomised, double-blind, placebo-controlled phase 3 trial. Lancet. 2012; 379(9829): 1879–86. PubMed Abstract | Publisher Full Text\n\nSleijfer S, Ray-Coquard I, Papai Z, et al.: Pazopanib, a multikinase angiogenesis inhibitor, in patients with relapsed or refractory advanced soft tissue sarcoma: a phase II study from the European organisation for research and treatment of cancer-soft tissue and bone sarcoma group (EORTC study 62043). J Clin Oncol. 2009; 27(19): 3126–32. PubMed Abstract | Publisher Full Text\n\nRay-Coquard I, Italiano A, Bompas E, et al.: Sorafenib for Patients With Advanced Angiosarcoma: A Phase II Trial From the French Sarcoma Group (GSF/GETO). Oncologist. 2012; 17(2): 260–266. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKummar S, Allen D, Monks A, et al.: Cediranib for Metastatic Alveolar Soft Part Sarcoma. J Clin Oncol. 2013; 31(18): 2296–2302. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAgulnik M, Costa RLB, Milhem M, et al.: A Phase II Study of Tivozanib in Patients With Metastatic and Nonresectable Soft-Tissue Sarcomas. Ann Oncol. 2017; 28(1): 121–127. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFrezza AM, Stacchiotti S, Gronchi A: Systemic Treatment in Advanced Soft Tissue Sarcoma: What Is Standard, What Is New. BMC Med. 2017; 15(1): 109. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBinh MB, Sastre-Garau X, Guillou L, et al.: MDM2 and CDK4 Immunostainings Are Useful Adjuncts in Diagnosing Well-Differentiated and Dedifferentiated Liposarcoma Subtypes: A Comparative Analysis of 559 Soft Tissue Neoplasms With Genetic Data. Am J Surg Pathol. 2005; 29(10): 1340–1347. PubMed Abstract | Publisher Full Text\n\nDickson MA, Tap WD, Keohan ML, et al.: Phase II trial of the CDK4 inhibitor PD0332991 in patients with advanced CDK4-amplified well-differentiated or dedifferentiated liposarcoma. J Clin Oncol. 2013; 31(16): 2024–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWilson BG, Wang X, Shen X, et al.: Epigenetic antagonism between polycomb and SWI/SNF complexes during oncogenic transformation. [published correction appears in Cancer Cell. 2011 Jan 18; 19(1): 153] Cancer Cell. 2010; 18(4): 316–28. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJudson I, Verweij J, Gelderblom H, et al.: Doxorubicin alone versus intensified doxorubicin plus ifosfamide for first-line treatment of advanced or metastatic soft-tissue sarcoma: a randomised controlled phase 3 trial. Lancet Oncol. 2014; 15(4): 415–423. PubMed Abstract | Publisher Full Text\n\nRyan CW, Merimsky O, Agulnik M, et al.: PICASSO III: A Phase III, Placebo-Controlled Study of Doxorubicin With or Without Palifosfamide in Patients With Metastatic Soft Tissue Sarcoma. J Clin Oncol. 2016; 34(32): 3898–3905. PubMed Abstract | Publisher Full Text\n\nTap WD, Papai Z, Van Tine BA, et al.: Doxorubicin plus evofosfamide versus doxorubicin alone in locally advanced, unresectable or metastatic soft-tissue sarcoma (TH CR-406/SARC021): an international, multicentre, open-label, randomised phase 3 trial. Lancet Oncol. 2017; 18(8): 1089–1103, [published correction appears in Lancet Oncol. 2018 Feb; 19(2):e78]. PubMed Abstract | Publisher Full Text\n\nMartin-Broto J, Pousa AL, de Las Peñas R, et al.: Randomized Phase II Study of Trabectedin and Doxorubicin Compared With Doxorubicin Alone as First-Line Treatment in Patients With Advanced Soft Tissue Sarcomas: A Spanish Group for Research on Sarcoma Study. J Clin Oncol. 2016; 34(19): 2294–2302. PubMed Abstract | Publisher Full Text\n\nBui-Nguyen B, Butrynski JE, Penel N, et al.: A phase IIb multicentre study comparing the efficacy of trabectedin to doxorubicin in patients with advanced or metastatic untreated soft tissue sarcoma: the TRUSTS trial. Eur J Cancer. 2015; 51(10): 1312–1320. PubMed Abstract | Publisher Full Text\n\nChawla SP, Papai Z, Mukhametshina G, et al.: First-Line Aldoxorubicin vs Doxorubicin in Metastatic or Locally Advanced Unresectable Soft-Tissue Sarcoma: A Phase 2b Randomized Clinical Trial. JAMA Oncol. 2015; 1(9): 1272–1280. PubMed Abstract | Publisher Full Text\n\nGupta S, Gouw L, Wright J, et al.: Phase II study of amrubicin (SM-5887), a synthetic 9-aminoanthracycline, as first line treatment in patients with metastatic or unresectable soft tissue sarcoma: durable response in myxoid liposarcoma with TLS-CHOP translocation. Invest New Drugs. 2016; 34(2): 243–252. PubMed Abstract | Publisher Full Text\n\nSeddon B, Strauss SJ, Whelan J, et al.: Gemcitabine and docetaxel versus doxorubicin as first-line treatment in previously untreated advanced unresectable or metastatic soft-tissue sarcomas (GeDDiS): a randomised controlled phase 3 trial. Lancet Oncol. 2017; 18(10): 1397–1410. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGelderblom H, Blay JY, Seddon BM, et al.: Brostallicin versus doxorubicin as first-line chemotherapy in patients with advanced or metastatic soft tissue sarcoma: an European Organisation for Research and Treatment of Cancer Soft Tissue and Bone Sarcoma Group randomised phase II and pharmacogenetic study. Eur J Cancer. 2014; 50(2): 388–96. PubMed Abstract | Publisher Full Text\n\nPenel N, Bui BN, Bay JO, et al.: Phase II trial of weekly paclitaxel for unresectable angiosarcoma: the ANGIOTAX Study. J Clin Oncol. 2008; 26(32): 5269–74. PubMed Abstract | Publisher Full Text\n\nAgulnik M, Tannir NM, Pressey JG, et al.: A phase II, multicenter study of the EZH2 inhibitor tazemetostat in adult subjects with INI1-negative tumors or relapsed/refractory synovial sarcoma. J Clin Oncol. 2016; 34(15_suppl): TPS11071–TPS11071. (suppl; abstr TPS11071). Publisher Full Text\n\nWilson BG, Wang X, Shen X, et al.: Epigenetic antagonism between polycomb and SWI/SNF complexes during oncogenic transformation. [published correction appears in Cancer Cell. 2011 Jan 18; 19(1): 153] Cancer Cell. 2010; 18(4): 316–28. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMsaouel P, Carugo A, Genovese G: Targeting proteostasis and autophagy in SMARCB1-deficient malignancies: where next? Oncotarget. 2019; 10(40): 3979–3981. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMsaouel P, Slack-Tidwell R, Genovese G, et al.: Phase II trial of ixazomib combined with gemcitabine and doxorubicin in patients with SMARCB1-deficient kidney malignancies. J Clin Oncol. 2019; 37(7_suppl): (suppl; abstr TPS678–TPS678). Publisher Full Text"
}
|
[
{
"id": "74306",
"date": "05 Nov 2020",
"name": "Conor Patrick Malone",
"expertise": [
"Reviewer Expertise Orbital rhabdomyosarcoma."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article clearly and concisely reviews the role of SMARCB1/INI1 in rhabdoid tumours, as well as summarising the literature and discussing management options. The structure is good, the language is accessible, and the references are appropriate and comprehensive.\nI would suggest a short conclusion to recap the main points and to ensure that there are clear learning outcomes for readers of varying experience levels.\nBelow are minor grammar/punctuation corrections and suggestions:\n\n\"the result can cause reduced, complete loss, and mosaic expression.\" - this is not clear - I would suggest changing to \"the result can cause reduced expression, complete loss of expression, and mosaic expression\" ?\nThere should be no apostrophe in 1980’s or 2000's, i.e. 1980s and 2000s are correct.\nIn Table 1 \"typically presents in intraabdominally in adult males\" the first \"in\" is an error.\n\"CNS\" is used without expansion/explanation of the acronym.\n\"prognosis if often poor\" - should read \"is often poor\".\nIn Tables 2 and 3 \"criteria was as follows\" should be \"were as follows\".\n\"A male predominance was mildly observed (14 cases, 56%), which is consistent with other larger reviews.\" - this wording is unclear - suggest \"Consistent with other larger reviews, there was a slight male predominance (14 cases, 56%).\"\n\"is termed AIM, which includes doxorubicin plus ifosfamide and mesna\" - explain this more clearly so that the initialism (AIM) makes sense.\n\"KIT\" is not an acronym/initialism but suggest \"KIT proto-oncogene\" so that it is clear what it is.\nExplain what \"mTOR\" and \"MDM2\" and \"EZH2\" stand for.\n\"as a highly intriguing\" - remove the word \"highly\".\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-662
|
https://f1000research.com/articles/9-1415/v1
|
07 Dec 20
|
{
"type": "Research Article",
"title": "Catheter placement selection for convection-enhanced delivery of therapeutic agents to brain tumors",
"authors": [
"Lisa H. Antoine",
"Roy P. Koomullil",
"Timothy M. Wick",
"Louis B. Nabors",
"Ahmed K. Abdel Aal",
"Mark S. Bolding",
"Roy P. Koomullil",
"Timothy M. Wick",
"Louis B. Nabors",
"Ahmed K. Abdel Aal",
"Mark S. Bolding"
],
"abstract": "Background: Convection-enhanced delivery (CED) of therapeutic agents to brain tumors allows clinicians to bypass the blood-brain barrier (BBB) to infuse virus therapy, biological, or chemotherapy directly into a brain tumor through convection. However, the effectiveness of infusions via CED may depend on catheter placement. Methods: This study used diffusion maps from magnetic resonance imaging (MRI) of human brain tumors and computational fluid dynamics (CFD) simulations to assess therapy volume distribution percentages based on catheter placement locations. Results: The primary outcome showed differences in volume distribution based on the catheter placement location. Total tumor volume filled ranged from 144.40 mm3 to 317.98 mm3. Percent filled of tumor volume ranged from 2.87% to 6.32%. Conclusions: The selection of the location for catheter placement using the region with the highest volume filled may provide optimal therapeutic effect. The researchers conclude that CFD may provide guidance for catheter placement in CED of therapeutic agents.",
"keywords": [
"tumor",
"catheter",
"convection-enhanced delivery",
"diffusion",
"computational fluid dynamics"
],
"content": "Introduction\n\nWhen using intravenous therapy to treat brain tumors, the blood-brain barrier (BBB) appears to hinder the effectiveness of the therapy. The BBB consists of glial cells and pericytes, which enclose the endothelial layer of the blood vessels, and restricts fluid flow into the brain1. Glucose, electrolytes, vitamins, peptides, regulatory proteins, and ~2% of small-molecule drugs (<~400 Dalton (Da) with <~8 hydrogen bonds) can cross the BBB; however, large-molecule drugs and ~98% of small-molecule drugs cannot adequately cross the BBB2,3. Therefore, clinicians may consider direct injection of therapeutic agents into the tumor instead of delivering the agent intravenously.\n\nMolecules diffuse from the site of injection at different rates. Wolak et al.4 compiled diffusion coefficients for selected molecules at 37°C and found that the diffusion rate for sucrose (342 Da) is 7×10-4 mm2/s; whereas, the diffusion rates for inulin (5000 Da), transferrin (80,000 Da), and PHPMA (515,000 Da) were 2.44×10-4 mm2/s, 7.50×10-5 mm2/s, and 1.29×10-5 mm2/s, respectively. Therapeutic agents are typically larger molecules that diffuse at slower rates.\n\nWithin the brain, stand-alone diffusion is slow and therapeutic agents may migrate only 1 mm from the source in three days5. Stand-alone diffusion in simulations for our study caused therapy to migrate a maximum distance of 6.68 mm from the source. Convection (i.e. injection of therapeutic agents under pressure) can improve the distribution of the therapeutic agents, as opposed to diffusion, which is passive and occurs during molecule movement from an area of higher concentration to an area of lower concentration.\n\nBobo et al.6 developed the CED method to increase therapeutic distribution in the brain. To bypass the BBB and deliver therapeutic agents directly to tumors, CED uses a catheter inserted into a cannula, which connects to a syringe pump. The pump introduces and maintains a pressure gradient, which activates mass fluid movement or convection. Using convection, Bobo et al.6 observed at least a 100-fold increase in therapy distribution. However, this 100-fold increase did not necessarily translate to success in clinical trials.\n\nStine et al.1 reviewed more than 15 trials and found that the outcomes using CED did not produce better results than standard treatment. For these trials, 10 clinical groups reported median survival, which ranged from 18.5 to 60 weeks7–13. These results suggest that either the therapy or the procedure needs improvement.\n\nWhen modeling therapy distribution using CED, Støverud et al.14 considered a region of the gray and white matter of the brain, not including tumor, along with porosity and patient-specific diffusivity and permeability parameters. They found that the therapy distribution followed the white matter region more than the gray matter region. Using tumor shapes that ranged from oblate to prolate, Sefidgar et al.15, found that larger tumors decreased therapeutic agent distribution in the interstitial fluid and that the prolate shape resulted in a wider range of distribution than other shapes. Zhan et al.16 modeled distribution using a patient-specific tumor geometry and constant parameters for pressure, permeability, and diffusivity and found that the therapy type and infusion locations influenced the distribution. When infusing therapy from the tumor center, uniform distribution mostly occurred. However, infusion from peripheral locations seemed to limit spatial distribution closer to that location.\n\nRecently Bhandari et al.17 compared distribution of three different therapies within the brain tumor using constant diffusivity, but variable permeability and porosity based on tumor to contrast agent correlation. These researchers found that at the outset distribution appears greater in higher permeability regions, but then shifts and appears greater in higher porosity regions with more available space. Tumor cell density appeared to be an effectiveness indicator and noticeably varied by therapy and time. Conversely, Zhan et al.18 observed that the permeability minimally affected the therapy distribution volume but did alter the distribution shape in space. CED appeared to cause the interstitial fluid pressure to increase near the infusion location.\n\nThe aim of this study is to assess therapy distribution in the tumor based on catheter placement location. Measurement of the volume distribution was a predictor of therapy effectiveness19–23.\n\nUsing CFD this study calculated therapy distribution percentages and total tumor volume filled with therapy. The study simulated transient therapy distribution in a patient-specific brain tumor using a pressure-based solver. The model analyzes the effect of convection and tumor properties such as geometry, diffusivity, permeability, porosity, and interstitial fluid pressure on therapy distribution within the tumor.\n\nT1-weighted imaging (T1W) and diffusion-weighted imaging (DWI) provided patient-specific geometry, diffusivity, and permeability information. Using patient-specific brain tumor characteristics may improve the accuracy of the volume distribution prediction. The researchers conducted sixteen simulations by dividing the tumor into four regions and within each region introduced therapy into four random locations.\n\n\nMethods\n\nPreviously, the investigators, using a two-dimensional domain, detailed the image processing and simulation pre-processing stages, the order of convergence for the mesh, and the numerical scheme accuracy for this analysis24. In the model development, simulation results for a two-dimensional domain closely agreed with the analytical results and mesh spacing was 0.5% of the domain length24. Average mesh spacing for this phase of the study was approximately 0.64% of the domain length. An examination of results from the numerical schemes, which included the first order, second order, power law, quick scheme, and third order scheme suggested that higher order schemes compared favorably to the analytical results24.\n\nThe tumor geometry, which represents initial diagnosis, for this simulation resulted from an MRI brain exam obtained on a 3 Tesla Phillips Ingenia MRI machine (Phillips Ingenia scanner, Netherlands) at the University of Alabama at Birmingham Hospital. The use of patient-specific information requires a review by the Institutional Review Board (IRB) to determine human subjects research status. After reviewing the submitted application the IRB determined that this research is not human subjects research. The acquired sagittal T1W gradient-echo images (Figure 1a) included slice thickness, repetition time, and echo time of 1.2 mm, 6.8 ms, and 3.3 ms, respectively. Axial diffusion tensor images resulted from the DWI of the same patient (Figure 1b). The patient received the standard of care, which is fractionated radiation therapy and chemotherapy with the drug temozolomide at a dose of 0.1 mL/hour for 6 hours. Repetition time and echo time of the DWI were 7927.7 ms, and 70.0 ms, respectively. The b-value, in-plane resolution, and slice thickness were 800 s/mm2, 1.75 mm, and 2 mm, respectively, with 33 diffusion directions performed. Images used in this study are available as Underlying data25.\n\n(a) Sagittal view of the T1-weighted scan. The tumor (red arrow) locates in the frontal lobe. (b) Axial view of the DWI shows restricted diffusion in tumor (red arrow)\n\nThe stereolithography (stl) tumor geometry (see Figure 2) resulted from the Lesion GNB version 2 software26. Investigators utilized the Ansys version 19.1 ICEM meshing tool to generate the computational mesh from the stl27.\n\nAnsys calculated the therapy concentration at the center of each element. Lengths along the x, y, and z axis were 12.86 mm, 30.4 mm, and 37.43 mm, respectively. Tumor volume was 5031.35 mm3\n\nThe researchers used DSI Studio to transform the DWI onto the T1W28,29. The transformation, which produced the diffusion tensor and fiber tracking images (Figure 3), was necessary to ensure that the diffusion tensors aligned to the geometry spatial locations. MATLAB allowed further validation of the transformation with the spatial plotting of the geometry with the diffusion tensors overlaid (Figure 4)30. Validation can also be performed using GNU Octave.\n\n(a) Diffusion tensor orientation at the spatial locations. Red represents the x component of the tensor. Green represents the y component. Blue represents z. (b) Fiber tracking included the anisotropy threshold of 0.16, angular threshold from 15 ° to 90 °, step size from 0.5 voxels to 1.5 voxels, and track length from 30 mm to 300 mm\n\nTumor geometry displayed in the xyz volume with diffusion tensor locations overlaid (points).\n\nVariable tensors included Dxx, Dyy, Dzz (principal and main diagonal diffusion rates and directions) and Dxy, Dxz, Dyz (off-diagonal diffusion rates and directions) aligned with the x, y, and z locations within the tumor. Minimum domain extents of the mesh were -21.24 mm, -0.75 mm, and 20.11 mm for x, y, and z. Maximum domain extents of the mesh were -8.38 mm, 29.65 mm, and 57.54 mm for x, y, and z. Because the diffusion tensors from the medical imaging analysis are Cartesian in nature, the researchers used these tensors and trilinear interpolation to derive tensors based on the cell center of the mesh element. The tumor geometry (Figure 5) encompassed four regions with four catheter placement locations per region. See Table 1 for input coordinates for each location. For each location, the radius was constant at 1.5 mm, which accounts for the radius of the catheter.\n\nEach region includes locations A, B, C, and D with the corresponding x, y, and z coordinates as shown in Table 1.\n\nUsing Ansys, the researchers conducted 16 simulations of therapy distribution within the tumor27. Diffusivity and main diagonal permeability location-dependent variable tensors resulted from the following governing equations6,7,31.\n\n\n\n∂C∂t represents concentration change with respect to time, u · ∇c represents convection, and ∇ · (D∇c) represents diffusion due to the concentration gradient.\n\n\n\nK0 represents permeability tensor. ξ represents diffusion tensor eigenvectors. Λ′ represents permeability and eigenvalues.\n\nA generation rate, defined as follows, was necessary to consider the effects of the infusion rate, source volume, and the density in the transport equation.\n\n\n\nSϕ represents generation rate. IR represents infusion rate. Sv represents source volume and ρ is density. The source represents the catheter placement location infused with one milliliter of an oncolytic herpes simplex virus for six hours32. Therefore, IR was 0.1 mL/hr. The minimum dose was 1×10-10 mL/m3 (1×10-19 mL/mm3).\n\nFor the simulations, the solver type was pressure-based with absolute velocity formulation and transient solver time. The model was viscous and laminar. The solution method was a third-order spatial discretization. The stability condition for a numerical scheme determined an acceptable time step size that did not cause non-physical solutions. The following equation is useful for calculating the maximum time step size (MTSS) and resolving the unsteadiness of the instance27.\n\n\n\nwhere Lscale (conservative) represents MIN(Lvol, Lext), U is maximum velocity at the domain boundary, Lvol represents V3, V is the domain volume, Lext represents MAX(Lx, Ly, Lz), Lx, Ly, Lz are domain extents (x,y,z direction), and ν is the kinematic viscosity. The MTSS and time steps were 10 seconds and 2,160 steps, respectively. Tumor porosity and average interstitial fluid pressure (IFP) of 0.6 and 266.65 Pa resulted from previous studies33,34.\n\n\nResults\n\nInfusion time for all simulations was 21,600 seconds (6 hours) with 0.6 mL of therapy infused per catheter. Simulations for four unique infusion locations per region resulted in therapy concentration ranges as depicted in Figure 6–Figure 9. The average velocity magnitude and pressure were 3.43×10-7 mm/s and 443.66 Pa, respectively. Average distance from the source was 7.4 mm. The investigators ranked each location based on at least 1×10-10 mL/m3 (1×10-19 mL/mm3) distribution throughout the tumor (see Table 2).\n\nConcentration range from 0.0 to 3.3×10-2 mL/mm3. Location A shows the highest percentage filled at 6.26%.\n\nConcentration range from 0.0 to 3.3×10-2 mL/mm3. Location D shows the highest percentage filled at 6.32%\n\nConcentration range from 0.0 to 3.3×10-2 mL/mm3. Location B shows the highest percentage filled at 5.3%.\n\nConcentration range from 0.0 to 3.3×10-2 mL/mm3. Location B shows the highest percentage filled at 4.47%.\n\nIn region one the location with maximum therapy distribution was A with 314.96 mm3 or 6.26% and a maximum Euclidean distance from this location of 9.6 mm. Location B (-12.09 mm, 28.16 mm, 39.74 mm) represented the least therapy distribution at 213.83 mm3 or 4.25%. In region two the location with maximum distribution was D with 317.98 mm3 or 6.32% and a maximum distance from this location of 8.56 mm. Location B (-10.16 mm, 19.82 mm, 21.63 mm) represented the least therapy distribution at 191.69 mm3 or 3.81%. In region three the location with the maximum distribution was B with 266.66 mm3 or 5.30% and a maximum distance from this location of 8.27 mm. Location C (-9.20 mm, 3.07 mm, 33.60 mm) represented the least distribution for region three and was 209.81 mm3 or 4.17%. In region four the location with the maximum distribution was B with 224.90 mm3 or 4.47% and a maximum distance from this location of 8.32 mm. Location D (-12.33 mm, 8.80 mm, 53.16 mm) in region four represented the lowest distribution percentage and was 2.87% or 144.40 mm3.\n\nThe tumor volume was 5031.35 mm3. The number of locations with volume filled greater than 250 mm3 was six, while the locations less than 250 mm3 was ten (see Figure 10).\n\n\nDiscussion\n\nThe primary outcome of this research showed differences in the therapy volume distribution based on the catheter placement location and suggested that location may influence the distribution and therapeutic value. Total volume filled by the therapy ranged from 144.40 mm3 to 317.98 mm3. Percent filled of tumor volume ranged from 2.87% to 6.32%. The average velocity range of 2.45×10-7 mm/s to 4.49×10-7 mm/s caused the therapy to displace from the source by an average of 7.4 mm, which was reasonable based on an infusion time of six hours.\n\nPreviously, research groups defined therapeutic value using the ratio between specific therapy volume of distribution (Vd) and volume of infusion (Vi), which is the therapy plus carrier fluid22,35–37. In vivo rodent and nonhuman primates’ experiments allowed these researchers to determine Vd by using image processing to measure the distribution of the specific therapy in the brain. Distribution for oncolytic viruses appeared to be unavailable because this distribution varies based on infusion location, tumor parameters, and therapy clearance38. The researchers in this study did not identify any in vivo Vd results for the oncolytic virus. However, clinicians at the authors’ institution are currently conducting a clinical trial to investigate immunotherapy in canines39. The trial is a regional collaboration designed to assess brain tumor therapies in humans and animals. Nevertheless, oncolytic viruses can reproduce in tumor cells destroying these cells without harming normal cells32.\n\nClinicians in the authors’ institution select four catheter placement locations to maximize therapy volume distribution. If the clinicians select the four previously mentioned patient-specific locations from regions 1, 2, 3, and 4, the effectiveness of therapy may improve. The selection of locations within each region with the highest total volume filled or highest tumor percentage filled may provide the most optimal therapeutic value.\n\nResearchers in this study analyzed 16 random locations, which provided a baseline mathematical prediction of the optimal catheter placement location. Clinicians may not currently use a mathematical model to select catheter placement locations, but instead select locations based on the avoidance of cell areas with visible signs of necrosis. This mathematical model may be an improvement over the current clinical method. In the next phase of this study the investigators will use a design optimization technique, which will allow the analysis of additional locations.\n\n\nConclusions\n\nIn this study, porosity and IFP resulted solely from previous studies; while diffusivity and permeability were mostly patient-specific. Patient-specific porosity and IFP may improve volume distribution accuracy. Although the authors did not consider retrograde infusion flow depending on tumor density and placement, it may improve the effectiveness of the treatment if future researchers consider the impact of retrograde flow.\n\nThe results presented suggest that computational fluid dynamic approaches using diffusivity and permeability parameters of actual patient data could greatly improve the treatment of adult and pediatric brain tumor patients by optimizing the placement of catheters in convection enhanced therapy. Using the specific anatomy of the patient, this novel method would optimize catheter placement to provide maximal tumor coverage of the therapeutic agent. This is the first report using diffusivity and permeability of real patient data and computational fluid dynamic modeling to guide catheter placement for convection enhanced delivery of a therapeutic agent. This predictive quantitative method to determine the ideal catheter placement location will assist clinicians in effectively treating brain tumors using CED.\n\n\nData availability\n\nHarvard Dataverse: Catheter placement selection for convection-enhanced delivery of therapeutic agents to brain tumors. https://doi.org/10.7910/DVN/H7C6A225.\n\nThis project contains T1-weighted and diffusion-weighted images used in the present study.\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).",
"appendix": "Author contributions\n\n\n\nLisa H Antoine substantially contributed to the study design, data (T1-weighted images and diffusion-weighted images) acquisition, analysis and interpretation, drafting and critically revising the article, and the final approval of the published version. Roy P Koomullil substantially contributed to the study conception and design, data analysis and interpretation, critically revising the article, and the final approval of the published version. Timothy M Wick, Louis B Nabors, and Mark S Bolding substantially contributed to the study design, data analysis and interpretation, critically revising the article, and the final approval of the published version. Ahmed K Abdel Aal substantially contributed to data acquisition, analysis and interpretation, critically revising the article, and the final approval of the published version.\n\n\nReferences\n\nStine CA, Munson JM: Convection-Enhanced Delivery: Connection to and impact of interstitial fluid flow. Front Oncol. 2019; 9: 966. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBanks WA: From blood-brain barrier to blood-brain interface: new opportunities for CNS drug delivery. Nat Rev Drug Discov. 2016; 15(4): 275–292. PubMed Abstract | Publisher Full Text\n\nPardridge WM: Drug transport across the blood-brain barrier. J Cereb Blood Flow Metab. 2012; 32(11): 1959–1972. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWolak DJ, Thorne RG: Diffusion of macromolecules in the brain: implications for drug delivery. Mol Pharm. 2013; 10(5): 1492–1504. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJain RK: Delivery of novel therapeutic agents in tumors: physiological barriers and strategies. J Natl Cancer Inst. 1989; 81(8): 570–576. PubMed Abstract | Publisher Full Text\n\nBobo RH, Laske DW, Akbasak A, et al.: Convection-enhanced delivery of macromolecules in the brain. Proc Natl Acad Sci U.S.A. 1994; 91(6): 2076–2080. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLaske D, Youle R, Oldfield E: Tumor regression with regional distribution of the targeted toxin TF-CRM107 in patients with malignant brain tumors. Nat Med. 1997; 3(12): 1362–1368. PubMed Abstract | Publisher Full Text\n\nWersäll P, Ohlsson I, Biberfeld P, et al.: Intratumoral infusion of the monoclonal antibody, mAb 425, against the epidermal-growth-factor receptor in patients with advanced malignant glioma. Cancer Immunol Immunother. 1997; 44(3): 157–164. PubMed Abstract | Publisher Full Text\n\nVoges J, Coenen HH, Wienhard K, et al.: Imaging-guided convection-enhanced delivery and gene therapy of glioblastoma. Ann Neurol. 2003; 54(4): 479–487. PubMed Abstract | Publisher Full Text\n\nWeber FW, Floeth F, Asher A, et al.: Local convection enhanced delivery of IL4-Pseudomonas exotoxin (NBI-3001) for treatment of patients with recurrent malignant glioma. Acta Neurochir Suppl. 2003; 88: 93–103. PubMed Abstract | Publisher Full Text\n\nLidar Z, Nass D, Mardor Y, et al.: Convection-enhanced delivery of paclitaxel for the treatment of recurrent malignant glioma: a Phase I/II clinical study. J Neurosurg. 2004; 100(3): 472–479. PubMed Abstract | Publisher Full Text\n\nKunwar S, Westphal M, Chang S, et al.: Phase III randomized trial of CED of IL13-PE38QQR vs Gliadel wafers for recurrent glioblastoma. Neuro Oncol. 2010; 12(8): 871–881. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDesjardins A, Gromeier M, Herndon JE, et al.: Recurrent glioblastoma treated with recombinant poliovirus. N Engl J Med. 2018; 379(2): 150–161. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStøverud KH, Darcis M, Helmig R, et al.: Modeling concentration distribution and deformation during convection-enhanced drug delivery into brain tissue. Transp Porous Media. 2012; 92: 119–143. Publisher Full Text\n\nSefidgar M, Soltani M, Raahemifar K, et al.: Effect of tumor shape, size, and tissue transport properties on drug delivery to solid tumors. J Biol Eng. 2014; 8: 12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhan W, Arifin D, Lee TK, et al.: Mathematical modelling of convection enhanced delivery of carmustine and paclitaxel for brain tumour therapy. Pharm Res. 2017; 34(4): 860–873. PubMed Abstract | Publisher Full Text\n\nBhandari A, Bansal A, Singh A, et al.: Comparison of transport of chemotherapeutic drugs in voxelized heterogeneous model of human brain tumor. Microvasc Res. 2019; 124: 76–90. PubMed Abstract | Publisher Full Text\n\nZhan W, Rodriguez Y, Baena F, et al.: Effect of tissue permeability and drug diffusion anisotropy on convection-enhanced delivery. Drug Deliv. 2019; 26(1): 773–781. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhan W: Convection enhanced delivery of anti-angiogenic and cytotoxic agents in combination therapy against brain tumour. Eur J Pharm Sci. 2020; 141: 105094. PubMed Abstract | Publisher Full Text\n\nStephen Z, Chiarelli P, Revia R, et al.: Time-Resolved MRI assessment of convection-enhanced delivery by targeted and nontargeted nanoparticles in a human glioblastoma mouse model. Cancer Res. 2019; 79(18): 4776–4786. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVidotto M, Botnariuc D, De Momi E, et al.: A computational fluid dynamics approach to determine white matter permeability. Biomech Model Mechanobiol. 2019; 18(4): 1111–1122. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSingh R, Bellat V, Wang M, et al.: Volume of distribution and clearance of peptide-based nanofiber after convection-enhanced delivery. J Neurosurg. 2018; 129(1): 10–18. PubMed Abstract | Publisher Full Text\n\nLueshen E, Tangen K, Mehta A, et al.: Backflow-free catheters for efficient and safe convection-enhanced delivery of therapeutics. Med Eng Phys. 2017; 45: 15–24. PubMed Abstract | Publisher Full Text\n\nAntoine LH, Koomullil RP, Wick TM, et al.: Model development to assess drug delivery to brain tumors. Early Career Technical Journal. 2018; 17: 73–79.\n\nLisa A: Catheter placement selection for convection-enhanced delivery of therapeutic agents to brain tumors. Harvard Dataverse, V1, 2020. http://www.doi.org/10.7910/DVN/H7C6A2\n\nGriffis J, Allendorfer J, Szaflarski J: Voxel-based Gaussian naïve Bayes classification of ischemic stroke lesions in individual T1-weighted MRI scans. J Neurosci Methods. 2016; 257: 97–108. PubMed Abstract | Publisher Full Text | Free Full Text\n\nANSYS® [computer program] Version 19.1.\n\nYeh F, Van Jay W, Wen-Yih I: Generalized q-sampling imaging . IEEE Trans Med Imaging. 2010; 29(9): 1626–1635. PubMed Abstract | Publisher Full Text\n\nYeh F, Verstynen TD, Wang Y, et al.: Deterministic diffusion fiber tracking improved by quantitative anisotropy. PLoS One. 2013; 8(11): e80713. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMATLAB: [computer program] Version 2019a, Natick, Massachusetts: The MathWorks, Inc.,2019.\n\nStøverud K: Modeling convection-enhanced delivery into brain tissue using information from magnetic resonance imaging. Master thesis; Universität Stuttgart, 2009.\n\nPatel DM, Foreman PM, Nabors LB, et al.: Design of a phase I clinical trial to evaluate M032, a genetically engineered HSV-1 expressing IL-12, in patients with recurrent/progressive glioblastoma multiforme, anaplastic astrocytoma, or gliosarcoma. Hum Gene Ther Clin Dev. 2016; 27(2): 69–78. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJain R: Transport of molecules in the tumor interstitium: a review. Cancer Res. 1987; 47(12): 3039–3051. PubMed Abstract\n\nBoucher Y, Salehi H, Witwer B, et al.: Interstitial fluid pressure in intracranial tumours in patients and in rodents. Br J Cancer. 1997; 75(6): 829–836. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAsthagiri A, Walbridge S, Heiss J, et al.: Effect of concentration on the accuracy of convective imaging distribution of a gadolinium-based surrogate tracer. J Neurosurg. 2011; 115(3): 467–473. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCroteau D, Walbridge S, Morrison P, et al.: Real-time in vivo imaging of the convective distribution of a low-molecular-weight tracer. J Neurosurg. 2005; 102(1): 90–97. PubMed Abstract | Publisher Full Text\n\nBernal GM, LaRiviere MJ, Mansour N, et al.: Convection-enhanced delivery and in vivo imaging of polymeric nanoparticles for the treatment of malignant glioma. Nanomedicine. 2014; 10(1): 149–157. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTalimogene Laherparepvec. Accessed June 7, 2020. Reference Source\n\nNCI Awards Cancer Moonshot℠ Grants in Five Areas of Precision Medicine Oncology. Accessed June 7, 2020. Reference Source"
}
|
[
{
"id": "83147",
"date": "07 Jun 2021",
"name": "Bryn A. Martin",
"expertise": [
"Reviewer Expertise Computational",
"in vivo",
"and in vitro fluid dynamics of cerebrospinal fluid. CNS diseases."
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSummary The selection of the location for catheter placement using the region with the highest volume filled may provide optimal therapeutic effect. The researchers conclude that CFD may provide guidance for catheter placement in CED of therapeutic agents. This paper could be improved by the following overall modifications a) restating conclusions to be more directly related with study results, b) addition of details in discussion with respect to the study findings, c) addition of validation data to determine the degree of agreement between computational predictions versus in vivo findings.\nSpecific comments\nPlease provide T1W imaging in-plane resolution. Also, resolution seems coarse for the needed high-resolution application. Please explain in the text or reference a study showing that the results documented in your study would be independent of the spatial imaging resolution.\n\nCatheter radius was specified as 1.5 mm. Please provide information on the exact catheter model utilized in the study. Also, please describe if the catheter geometry was added to the computational mesh.\n\nFor time step size, please reference a previous study, or provide supporting information, showing the time-step / mesh resolution you chose resulted in numerically-independent results.\n\nPlease provide information on the diffusion coefficient specified.\n\nThis model did not have interaction of interstitial spaces and pressure. Please explain why this was chosen and what are the drawbacks / benefits of that approach.\n\nThis model represented the end of the catheter by a source term. In reality, the source is from a catheter tip with parabolic flow profile. Please explain if this assumption may impact the results.\n\nPlease provide information on the external boundary conditions of pressure on the model walls.\n\nPlease provide information on how the total volume filled in the tumor was calculated. For example, was there a minimum therapeutic threshold defined that would specify if a region was included in the volume filled?\n\nThe total volume dosed over 6 hours was 600 uL (600 mm3). However, the simulations showed total volumes filled less than 600 uL. Please explain where the missing drug went.\n\nTable 2 shows relatively small differences in % volume of tumor filled. Would be helpful to provide insight into if these differences are statistically significant or if they are artifacts of the simulation boundary conditions / limited resolution of input data.\n\nDiscussion “the primary outcome of this research showed differences in the therapy volume distribution based on the catheter placement location and suggested that location may influence the distribution and therapeutic value.” This statement is difficult to reconcile with regard to the significance of the “differences”. These differences are small. See comment 11 above.\n\nThe discussion of results is limited. Authors should consider a) how does this study compare to previous CED study results and modeling approach? b) what are the specific limitations to the model and how are they expected to impact the results, c) The model lacks validation with in vivo measurements, this should be discussed.\n\nA significant portion of the discussion text relates to the researcher’s clinical trial. This seems like an unrelated topic and text should be shortened with respect to this trial.\n\nThe conclusion that the computational fluid dynamics could “greatly improve the treatment of adult and pediatric brain tumor patients by optimization…” seems overconfident on the implications of the results. Authors should consider toning down this statement to what the study results show directly.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": []
},
{
"id": "78819",
"date": "24 May 2024",
"name": "Ajay Bhandari",
"expertise": [
"Reviewer Expertise Bio-fluid mechanics",
"Image based numerical modeling",
"CFD",
"cancer drug delivery."
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis research focuses on developing a computational model to predict the optimum catheter placement selection for convection-enhanced delivery (CED) of therapeutic agents to brain tumors. To make the computational model heterogeneous, authors have used diffusion-weighted imaging (DWI) data and have predicted therapy volume distributions based on different catheter locations. The reviewer thinks that most of the work regarding catheter design, placement and its effect on CED delivery has already been performed by Linninger et. Al. in this field. Check the following papers.\nLinninger et al.(20081): “Computational methods for predicting drug transport in anisotropic and heterogeneous brain tissue”, Journal of Biomechanics, 2008 Linninger et al.(20082): \"Prediction of convection-enhanced drug delivery to the human brain, Journal of Theoretical Biology\", 2008\nReviewers suggest the authors to improve the quality of the paper. Following comments can help the authors to enhance the quality of the research.\nComment 1: The literature review has not been done properly. There are several studies already been done in this field (CED). The authors have not properly acknowledged all of them. Neither the authors have properly mentioned the research gaps. Comment 2: No detailed information has been provided for the DWI imaging. From where the data was acquired? What were the imaging parameters? How the analysis was carried out? How were the diffusion tensors along the white and grey matter of the brain obtained? Authors should read the following paper in this regard Linninger et al.(20082). Comment 3: No details regarding the numerical simulations have been shown. What were the boundary conditions, discretization schemes especially for the convection term, simulation time, solvers? How the pressure velocity coupling was incorporated? Were the equations solved in a segregated or coupled manner? Comment 4: Results and Discussion part is very weak. No proper clarifications have been given. No clinical implications have been given.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1415
|
https://f1000research.com/articles/9-1093/v1
|
04 Sep 20
|
{
"type": "Research Article",
"title": "Effect of vitamin E supplementation on orthodontic tooth movement in Wistar rats",
"authors": [
"Erliera Sufarnap",
"Darmayanti Siregar",
"Yumi Lindawati",
"Darmayanti Siregar",
"Yumi Lindawati"
],
"abstract": "Background: Tooth movement induced by the application of orthodontic force is facilitated by bone remodelling cells and chemical mediators. Vitamin E has anti-inflammatory properties, which helps in suppressing the damaging effects of oxygen free radicals in cells during bone formation. This study aimed to evaluate the effect of vitamin E supplementation on orthodontic tooth movement in Wistar rats. Methods: Wistar rats (n=56) were divided into two groups. Group 1 served as the control groups, while group 2 was given vitamin E for 14 days before application of orthodontic force. Each group was divided into four subgroups (n=7), corresponding to the number of days orthodontic force lasted, i.e. 0, 1, 3, 7 days. At each of these four time points, distance measurements and quantity of osteoblasts-osteoclasts were measured in each rat. Results: Tooth movement distance was increased for group 2 than group 1 for all time intervals, but this difference was only statistically different on day 3 (p=0.001). For both groups, tooth movement was significantly different between each time interval in each group (p=0.041). The mean number of osteoblast cells was increased for group 2 compared to group 1 for all time intervals (p<0.05), but was not significant different between time intervals (p=0.897). The number of osteoclasts was not significantly different between groups, but it was statistically different between time intervals (p=0.004). Conclusion: Present outcomes demonstrate that vitamin E contributes to faster tooth movement compared to control group. It also stimulates more bone formation without reducing the bone resorption.",
"keywords": [
"Orthodontic tooth movement",
"vitamin E",
"tooth movement distance",
"osteoblast",
"osteoclast."
],
"content": "Introduction\n\nTooth movement is induced by the application of orthodontic force characterized by bone and periodontal tissue remodelling. Orthodontic force also alters periodontal tissue vascularity and blood flow, resulting in the local synthesis and release of various molecules such as neurotransmitters, cytokines, growth factors, colony-stimulating factors and arachidonic acid metabolites1.\n\nBone remodelling is a process that enables tooth movement. It involves bone-reabsorption by osteoclasts on the pressure site and bone-formation by osteoblasts on the tension site2,3. Osteoclasts are multinucleated cells, irregular in shape with a process originating from Howship’s lacunae4. They stimulate bone resorption by creating cavities in the bone known as lacunae that will be filled by osteoblast cells3. According to Mavragani et al., the cellular process of osteoclast proliferation has been used as important indicators in evaluating the level of tooth movement5. Osteoblasts are mononuclear cells that originate from mesenchymal stem cells in bone marrow. Mature osteoblasts form the osteoid by synthesizing collagen and non-collagen proteins6.\n\nAccording to Burstone in Asiry’s citation, there are three phases of orthodontic tooth movement, which consists the initial, lag and postlag phases7. The initial stage of orthodontic tooth movement stimulates an inflammatory response involving cells and blood vessels in periodontal ligaments as well as chemical mediators. In response to mechanical stress caused by the application of orthodontic force, substances such as cytokines and enzymes are released2,8. Interleukin-1β is a pro-inflammatory cytokine that facilitates fusion and activation of osteoclasts, and encourages early bone resorption9.\n\nStudies have shown that vitamin E has anti-inflammatory properties, which helps suppress damaging effects of oxygen free radicals in cells during bone formation10. Previous studies carried out by Esenlik et al. and Xu et al. suggest that vitamin E supplementation may alter cytokine production; vitamin E supplement maintains normal bone remodelling in young animals and increases bone mass by decreasing the concentration of free radicals which suppress bone formation11,12.\n\nSince orthodontic tooth movement is facilitated by bone remodelling cells and chemical mediators, it is possible to hypothesize that vitamin E has a positive effect on bone remodelling cells, which is crucial to tooth movement. However, it is unknown if vitamin would accelerate the movement or inhibit tooth movement. The purpose of this study was to evaluate the effect of vitamin E supplementation on orthodontic tooth movement in Wistar rats. Mice and rats are mammals that have a reasonably comparable metabolism to humans, which can be used for biological-cellular mechanism analysis in orthodontic tooth movement13.\n\n\nMethods\n\nThis article was reported in line with the ARRIVE guidelines. The study was an in-vivo quasi experiment, which was approved by the Animal Research Ethics Committee, Department of Biology - Faculty of Mathematics and Science, Universitas Sumatera Utara (No. 0128/KEPH-FMIPA/2019).\n\nA total of 56 healthy male, four to five-months old, Wistar rats, weighing 150–250 grams, were used in this study. The Wistar Rats came from the same breeding farm (Deli Serdang, North Sumatera, Indonesia) in two cycles.\n\nThe rats were adapted to their environment for 7 days before the experiment start. They were nurtured at the Animal House at Faculty of Mathematics and Science, Universitas Sumatera Utara in polycarbonate cage, which measured 480 mm × 265 mm × 210 mm. Each cage had wood shavings on the floor, and contained 3 or 4 animals, which were marked for each subgroup. Rats were chosen for each group by simple random sampling.\n\nLow light to dark cycle was maintained for a minimum 12 hours at 25–30°C for room temperature within the experiment period. The rats were given a standard pellet diet. All conditions served to produce the optimum condition of the rats’ habitat14. A rubber separator was inserted between maxillae’s incisors to produce non-invasive experiments. Anaesthetic was used to euthanize the rats at the end of the experimental procedure.\n\nWistar rats (n=56) were divided into two groups. Each group was then divided into four subgroups (n=7), corresponding to the number of the days orthodontic force lasted, i.e. 0, 1, 3, 7 days. The sample size of each subgroup was decided by Sastroasmoro and Ismael’s formula for hypothetical analysis between independent variables15.\n\nSubgroups were chosen based on the rats’ social behaviours. Hyperactive rats were chosen to be in the same cage, separately to rats with a more passive behaviour. These conditions avoided any anxiety social-related behaviour between rats in the cage within the experiment. For each experiment, a researcher who was blind to the experiment chose a sample randomly from each cage.\n\nGroup 1 were the control group and were given water orally as a placebo. The rats’ tail was marked with black pen. Group 2 were the experimental group and were given vitamin E (dl-α-Tocopheryl Acetate; Sanbe, Indonesia) at a dosage of 60 mg/kg, orally using gavage needle. The group 2 rats’ tail marked with red pen.\n\nWater and vitamin E were given every day at 8am, for 14 days before and continued after application of orthodontic force. After 14 days, orthodontic force was applied to each rat in both groups by addition of a rubber separator to one of the maxilla incisors (Figure 1A). This administration of orthodontic force applied were carried out before daily water and vitamin E feeding. This procedure counted as the baseline time of the experiment. At each of these four time points distance measurements and quantity of osteoblasts-osteoclasts were measured (see section below).\n\n(A) Rat separator; (B) Distance measurement; (C) Microscopic of whole teeth at 40x magnification. Lines, Yellow=teeth; Green=periodontal ligament; Red=alveolar bone. Arrows, Blue=pressure side; Green=tension side.\n\nAt end of each experiment period, the dosage of ketamine® at 80mg/kg of body weight and xyla® (Interchemie, Holland) at 10mg/kg of body weight was used to euthanised each rat by cardiac puncture methods for further research with blood analysis.\n\nTooth movement was measured using a digital calliper (Mitutoyo, Japan) was used to measure the distance between maxilla incisors at mesial cervical (Moorrees method) immediately after removal of the rubber separator (Figure 1B)16.\n\nThe pre-maxillae were dissected and fixated in 10% formalin for 24h, and decalcified with rapid-decalsifier, Nitric acid 10% (Aurona Scientific, Singapore) for 10-14 days. The embedded blocks were trimmed using a Leica microtome (Leica, Germany) into 5µm sections. Histological sections were stained with haematoxylin-eosin and were examined using Olympus CX21 light microscope at 400x magnification to analyse the number of cells within five fields of view for each measurement. A pressured site exhibited as a narrow area between teeth and alveolar bone where the tooth tended to move, and this site was used for osteoclast analysis. A tension site exhibited as a wide area between teeth and alveolar bone where the tooth was left out, and this site was used for osteoblast analysis (Figure 1C).\n\nIBM-SPSS (Statistical Package for Social Sciences), version 26.0, was used for statistical analysis. Independent t-test and Mann-Whitney test were used to analyse the difference between the two main groups. General Linear Model-Repeated Measures (ANOVA GLM-RM) and Friedman analysis were used to analyse the difference between time intervals. Significant differences were determined at p<0.05.\n\n\nResults\n\nTooth movement distances were greater in group 2 compared to group 1 at each time point (Table 1). This difference was only statistically significant on day 3 (p=0.001). For both groups, tooth movement was significantly different between each time interval in each group (p=0.041). After day 3, movement for group 1 reduced, while for group 2, this continued to increase until day 7.\n\nData are presented as mean±SD.\n\np<0.05 – statistically significant. aIndependent t-test; bANOVA GLM-RM\n\nThe number of osteoblasts in group 2 were higher compared with group 1 at each time point (Figure 2A and B; Table 2). These differences were statistically significant (p<0.05). Group 2 showed increased osteoblasts starting from day 0 to day 3, while group 1 had decreased osteoblast after day 3.\n\n(A) osteoblasts in group 1 (control); (B) osteoblasts in group 2 (vitamin E treatment); (C) osteoclasts in group 1; (D) osteoclasts in a Howship’s lacuna in group 2.\n\nData are presented as mean±SD.\n\np<0.05 – statistically significant. a Mann-Whitney test; b Friedman analysis\n\nThe number of osteoclasts in group 2 were higher than group 1 except on day 1, but the differences were not significant statistically (Figure 2C and D; Table 3).\n\nData are presented as mean±SD.\n\np<0.05 – statistically significant. aMann-Whitney test; bFriedman analysis\n\n\nDiscussion\n\nOrthodontic force causes gradual compression on the periodontal ligament, which leads to circulatory disorders, such as ischemia and hypoxia in the early stage of orthodontic tooth movement17. Hypoxia and compression caused by orthodontic force stimulate the production of reactive oxygen species and free radicals, which contribute to cellular and tissue damage, especially damaging lipid peroxidation chains18. Vitamin E is a strong biological antioxidant that has several functions: scavenges free radicals, which inhibit lipid peroxidation and inflammation; protects ischemic tissue and hypoxia; provides immunostimulation11,19. Norazlina et al. observed the effect of vitamin E supplementation on bone metabolism in mice treated with nicotine. Their study results suggested that vitamin E can increase trabecular bone formation and prevent bone calcium loss by reducing pro-inflammatory cytokines20.\n\nIn the present study, it can be seen that both groups showed increased tooth movement distance as well as increase in the number of osteoclast and osteoblast cells on day 1. This is due to the initial phase of tooth movement after application of orthodontic force21. This phase occurs 24 hours to 48 hours after application of orthodontic force on teeth3.\n\nOur results showed that the number of osteoclasts is higher in group 2 compared to group 1 although the difference was not statistically significant. Miresmaeili et al., in their study on the effect of vitamin C to orthodontic tooth movement, found that osteoclast numbers were significantly higher in the vitamin C group, which hence accelerates tooth movement22. Kale et al., in their research on vitamin D injection, observed a significant amount of Howhip’s lacunae in resorption cavity as a result of osteoclast’s activity23. Future research is required to observe the comparison between Howship’s lacunae and osteoclasts numbers.\n\nIn our study, there were statistically significant differences in the mean number of osteoblast cells between both groups at each time observed. Kawakami and Takano-Yamamoto demonstrated an increased osteoclast and osteoblast number with local injection of 1,25-dihydroxyvitamin D3 in the submucosal palatal area of rats subjected to tooth movement on day 7. Increased osteoblast counts were observed on day 1419,24. In another study, Feresin et al. reported that the formation rate and bone volume increased significantly by 65% in rat bone, who were given a vitamin E diet compared to the control group. Their result indicated that a vitamin E diet was able to increase the process of mineralization and bone formation mediated by osteoblast cells25. Diravidamani et al. stated that many drugs that are used to reduce pain had effects on orthodontic tooth movement. Further research should be done to observe vitamin E on pain regulation, because it has anti-inflammatory effect, which is assumed to reduce pain in orthodontic treatment10,26–28.\n\nThe force mechanism from the separator used in our study was static and the elasticity from the separator is easily lost due to saliva acidity (pH), food and chewing process; a the force of a rubber separator will be reduced by 50–55% within 24 hours29. This is a limitation of our study, as we wanted to analyse for a longer time and with a larger force. The aim of our study was to see the orthodontic tooth movement and not stabilization, so we decided to observe the orthodontic movement within the initial phase, and not all phases until the stabilization phase.\n\n\nConclusions\n\nOur findings demonstrated that vitamin E accelerates tooth movement and stimulates bone formation. The number of osteoblast cells in the vitamin E supplemented group is significantly higher than those in the non-vitamin E group. Further studies are needed to evaluate the effect of different doses and types of vitamin E.\n\n\nData availability\n\nOpen Science Framework: Methods, Figures, and Results from \"Effect of Vitamin E Supplementation on Orthodontic Tooth Movement in Wistar Rats, https://doi.org/10.17605/OSF.IO/3S4QB30.\n\nOpen Science Framework: ARRIVE checklist for ‘Effect of vitamin E supplementation on orthodontic tooth movement in Wistar rats’, https://doi.org/10.17605/OSF.IO/3S4QB30.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "References\n\nKrisnan V, Davidovitch Z: Cellular, Molecular, and Tissue-Level Reactions to Orthodontic Force. Am J Orthod Dentofacial Orthop. 2006; 129(4): 469.e1–32. PubMed Abstract | Publisher Full Text\n\nNimeri G, Kau CH, Abou-Kheir NS, et al.: Acceleration of Tooth Movement During Orthodontic Treatment-A Frontier in Orthodontics. Prog Orthod. 2013; 14: 42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAriffin SHZ, Yamamoto Z, Abidin IZZ, et al.: Cellular and Molecular Changes in Orthodontic Tooth Movement. ScientificWorldJournal. 2011; 11: 1788–803. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBhalaji SI: Orthodontics: The Art and Science. New Delhi: Arya (MEDI) Publishing House, 4thed, 2009; 400–2. Reference Source\n\nMavragani M, Brudvik P, Selvig KA: Orthodontically Induced Root and Alveolar Bone Resorption: Inhibitory Effect of Systemic Doxycycline Administration In rats. Eur J Orthod. 2005; 27(3): 215–25. PubMed Abstract | Publisher Full Text\n\nAlansari S, Sangsuwon C, Vongthongleur T, et al.: Biological Principles Behind Accelerated Tooth Movement. Seminars in Ort. 2015; 21(3): 161–61. Publisher Full Text\n\nAsiry MA: Biological Aspects of Orthodontic Tooth Movement: A Review of Literature. Saudi J Biol Sci. 2018; 25(6): 1027–1032. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIsola G, Matarese G, Cordasco G, et al.: Mechanobiology of The Tooth Movement During The Orthodontic Treatment: A Literature Review. Minerva Stomatol. 2016; 65(5): 299–304. PubMed Abstract\n\nBasaran G, Ozer T, Kaya FA, et al.: Interleukine-1beta and tumor necrosis factor-alpha levels in the human gingival sulcus during orthodontic treatment. Angle orthod. 2006; 76(5): 830–36. PubMed Abstract\n\nUysal T, Amasyali M, Olmez H, et al.: Stimulation of Bone Formation in The Expanding Inter-premaxillary Suture by Vitamin E, In Rat. Korean J Orthod. 2009; 39(5): 337–48. Publisher Full Text\n\nEsenlik E, Naziroglu M, Acikalin C, et al.: Vitamin E Supplementation Modulates Gingival Crevicular Fluid Lipid Peroxidation and Antioxidant Levels in Patients With Orthodontic Tooth Movement. Cell Biochem Funct. 2012; 30(5): 376–81. PubMed Abstract | Publisher Full Text\n\nXu H, Watkins BA, Seifer MF: Vitamin E Stimulates Trabecular Bone Formation and Alters Epiphyseal Cartilage Morphometry. Calcif Tissue Int. 1995; 57(4): 293–300. PubMed Abstract | Publisher Full Text\n\nKirschneck C, Bauer M, Gubernator J, et al.: Comparative Assessment of Mouse Models for Experimental Orthodontic Tooth Movement. Sci Rep. 2020; 10(1): 12154. PubMed Abstract | Publisher Full Text | Free Full Text\n\nARRP: Guideline 20: Guidelines for Housing of Rats in Scientific Institutions. Animal Welfare Branch. (www.animalethics.org.au): 16–55. Reference Source\n\nSastroasmoro S, Ismael S: Dasar-Dasar Metodologi Penelitian Klinis. 5th ed. Sagung Seto. 2014; 352–368. Reference Source\n\nVanjari K, Nuvvula S, Kamatham R: Prediction of Canine and Premolar Size Using the Widths of Various Permanent Teeth Combinations: A Cross-Sectional Study. Contemp Clin Dent. 2015; 6(Suppl 1): S210–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nArai C, Nomura Y, Ishikawa M, et al.: HSPA1A is upregulated in periodontal ligament at early stage of tooth movement in rats. Histochem Cell Biol. 2010; 134(4): 337–43. PubMed Abstract | Publisher Full Text\n\nZhang L, Wei W, Xu J: Inhibitory Effect of Melatonin on Diquat-Induced Lipid Peroxidation in Vivo as Assessed by the Measurement of F2-Isoprostanes. J Pineal Res. 2006; 40(4): 326–31. PubMed Abstract | Publisher Full Text\n\nArjmandi BH, Juma S, Beharka A, et al.: Vitamin E Improves Bone Quality in the Aged but Not in Young Adult Male Mice. J Nutr Biochem. 2002; 13(9): 543. PubMed Abstract | Publisher Full Text\n\nNorazlina M, Lee PL, Lukman HI, et al.: Effects Of Vitamin E Supplementation On Bone Metabolism In Nicotine-Treated Rats. Singapore Med J. 2007; 48(3): 195–9. PubMed Abstract\n\nNoda K, Nakamura Y, Kogure K, et al.: Morphological Changes in the Rat Periodontal Ligament and Its Vascularity After Experimental Tooth Movement Using Superelastic Forces. Eur J Orthod. 2009; 31(1): 37–45. PubMed Abstract | Publisher Full Text\n\nMiresmaeili A, Mollaei N, Azar R, et al.: Effect of Dietary Vitamin C on Orthodontic Tooth Movement in Rats. J Dent (Tehran). 2015; 12(6): 409–13. PubMed Abstract | Free Full Text\n\nKale S, Kocadereli I, Atilla P, et al.: Comparison of the effects of 1,25 dihydroxycholecalciferol and prostaglandin E2 on orthodontic tooth movement. Am J Orthod Dentofacial Orthop. 2004; 125(5): 607–614. PubMed Abstract | Publisher Full Text\n\nKawakami M, Takano-Yamamoto T: Local Injection of 1,25-Dihydroxyvitamin D3 Enhanced Bone Formation for Tooth Stabilization After Experimental Tooth Movement in Rats. J Bone Miner Metab. 2004; 22(6): 541–6. PubMed Abstract | Publisher Full Text\n\nFeresin RG, Johnson SA, Elam ML, et al.: Effects of Vitamin E on Bone Biomechanical and Histomorphometric Parameters in Ovariectomized Rats. J Osteoporos. 2013; 2013: 825985. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDiravidamani K, Sivalingam SK, Agarwal V: Drugs Influencing Orthodontic Tooth Movement: An Overall Review. J Pharm Bioall Sci. 2012; 4(Suppl 2): S299–303. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJiang Q: Natural forms of vitamin E: metabolism, antioxidant and anti-inflammatory activities and the role in disease prevention and therapy. Free Radic Biol Med. 2014; 72: 76–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLong SJ, Benton D: Effects of vitamin and mineral supplementation on stress, mild psychiatric symptoms, and mood in nonclinical samples: a meta-analysis. Psychosom Med. 2015; 75(2): 144–153. PubMed Abstract | Publisher Full Text\n\nWeissheimer A, Locks A, Macedo de Menezes M, et al.: In Vitro Evaluation of Force Degradation of Elastomeric Chains Used in Orthodontics. Dental Press J Orthod. 2013; 18(1): 55–62. PubMed Abstract | Publisher Full Text\n\nSufarnap E: Methods, Figures, and Results from \"Effect of Vitamin E Supplementation on Orthodontic Tooth Movement in Wistar Rats. 2020. http://www.doi.org/10.17605/OSF.IO/3S4QB"
}
|
[
{
"id": "70989",
"date": "09 Sep 2020",
"name": "Ananto Ali Alhasyimi",
"expertise": [
"Reviewer Expertise orthodontic biomaterials"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIt’s been an honor for us to review the manuscripts. First of all, please allow me to congratulate the authors for endeavoring to undertake this study which I found very interesting and valuable as a recommendation for further study in the orthodontics field especially in the acceleration of orthodontic tooth movement area. We have been reviewing throughout the manuscript. The following are our comments related to the manuscript.\nGeneral Comment:\nThe title and topic are quite interesting for journal scope and discover a novelty especially for developing natural materials to accelerate orthodontic tooth movement.\n\nPaper is well-organized and easy to follow.\n\nTo improve the readability, It is recommended that the text is given an English language edit as many of the sentences might be misunderstood and some of typos were found. I suggest a revision of the English grammar structures by an expert editor in revising manuscripts.\n\nAbstract:\nThe author mentioned: “Tooth movement induced by the application of orthodontic force is facilitated by bone remodeling cells and chemical mediators. Vitamin E has anti-inflammatory properties, which helps in suppressing the damaging effects of oxygen free radicals in cells during bone formation”. As we know that inflammation is a part of the orthodontic tooth movement, so it is better to find another logic to connect the reason for using Vitamin E in this study.\n\nIt stated: “Conclusion: Present outcomes demonstrate that vitamin E contributes to faster tooth movement compared to control group. It also stimulates more bone formation without reducing the bone resorption”. The author only evaluates the osteoclast and osteoblast number and did not analyze bone formation nor bone resorption indicator. It is better to write a conclusion based on what the author did.\n\nIntroduction:\nAt the end of the introduction part please add some of the hypotheses.\n\nMethods:\nAuthor used healthy male Wistar ratsà what is the rationale? Please explain shortly in the manuscript.\n\nA dosage of 60 mg/kg Vitamin E was used in this study, what is the rationale? Please explain shortly in the manuscript.\n\nAuthors have chosen HE staining for labeling osteoclast and osteoblast number but did not use any specific marker for osteoblasts (e.g. Von-kossa or Alkaline phosphatase.) and osteoclast (TRAP), rather relied on visual analyses and counting of cells. This method of counting cells is somewhat arbitrary. There are a number of markers available for osteoblasts – osteoclasts using these biomarkers will strengthen this study.\n\nAuthor should add the special characteristic of osteoclast-osteoblast (e.g. osteoblast was demonstrated as a cuboidal shape cell, characterized by a single, large, deep blue-purple nuclei, and found on the edge surfaces of alveolar bone) to minimize the bias and add the region of interest (ROI) since ROI is the key element in both number counts and distribution. Therefore, it is of utmost importance to show where were these ROIs.\n\nA blinded evaluator was responsible for determination of the score and had undergone a training exercise involving the calculation of the Kappa coefficient for the determination of intra-examiner agreement, what about the result? Please write it whether it is satisfactory or not.\n\nResults\n\nGeneral results were not reported. What happens with the general health of the animal during the administration of Vitamin E (e.g.general toxicity, edema, deaths affect the body weight of animals?)\n\nDiscussion\nFairly well written with minor typographical errors.\n\nConclusion\n\nThe author only evaluates the osteoclast and osteoblast number and did not analyze bone formation nor bone resorption indicator. It is better to write a conclusion based on what the author did.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "5988",
"date": "15 Oct 2020",
"name": "Erliera Sufarnap",
"role": "Author Response",
"response": "Dear Dr Ananto Ali Alhasyimi, Universitas Gadjah Mada, Sleman, IndonesiaThank you for your kind assistance in reviewing our manuscript becoming very marvelous manuscript with your advices at many chapters. We have revised the manuscript according to some inquiries you have been suggested. Finally we add the title become ‘A preliminary study’ which suppose to be our limitation of this research and would be continue to further research which we mentioned on discussion part. Several additional references already provided to consolidated theories which were needed. The following are our comprehensive comments to the review after revised. Abstract “As we know that inflammation is a part of the orthodontic tooth movement, so it is better to find another logic to connect the reason for using Vitamin E in this study”. : We have edited the reason of using VE in this study The author only evaluates the osteoclast and osteoblast number and did not analyze bone formation nor bone resorption indicator. It is better to write a conclusion based on what the author did. : We have edited the conclusion based on the variables we observed without any other conclusion. 2. Introduction : “ At the end of the introduction part please add some of the hypotheses”. Comment : We have edited our hypothesize more pointedly. 3. Methods : a. “Author used healthy male Wistar rats what is the rationale? Please explain shortly in the manuscript.” Comment : We were sorry, we’ve missed the detail to explain why. We edited and gave a reason that VE able to modulate the estrogen (hormone) and a reference had been added to strengthen this reason.b. “A dosage of 60 mg/kg Vitamin E was used in this study, what is the rationale? Please explain shortly in the manuscript.Comment : The dose chosen was based on a research that had been done by Nur Azlina et al. 17.c. “There are a number of markers available for osteoblasts – osteoclasts using these biomarkers will strengthen this study”. Comment : We mentioned changes in the title and discussion chapter as our limitations of the studyd. Author should add the special characteristic of osteoclast-osteoblast to minimize the bias and add the region of interest (ROI) since ROI is the key element in both number counts and distribution. Therefore, it is of utmost importance to show where were these ROIs. Comment : Thank you Doctor, We completed and added additional references to mention about the characteristic of osteoblasit and osteoclast and expalained about the specific ROI.e. A blinded evaluator was responsible for determination of the score and had undergone a training exercise involving the calculation of the Kappa coefficient for the determination of intra-examiner agreement, what about the result? Please write it whether it is satisfactory or not. Comment : To be honest, the level for IRR results have not very satisfactory result. The observer decided to use the “main value” from both raters as the datas to be analysed. And we also editted and added this results as our limitation of the study at the discussion chapter. 4. Results : General results were not reported. What happens with the general health of the animal during the administration of Vitamin E (e.g.general toxicity, edema, deaths affect the body weight of animals?) Comment : We have edited this inquiry to the result’s chapter. 5. Discussion : “Fairly well written with minor typographical errors”. Comment : Thank you for the review and we already rechecked and changed. 6. Conclusion : “The author only evaluates the osteoclast and osteoblast number and did not analyze bone formation nor bone resorption indicator. It is better to write a conclusion based on what the author did”. Comment : We have edited the conclusion based on the variables we observed without any other conclusion. Thank you for the established reviews and the valuable insight for our manuscript."
}
]
},
{
"id": "71955",
"date": "25 Sep 2020",
"name": "Hiroyuki Kanzaki",
"expertise": [
"Reviewer Expertise bone biology"
],
"suggestion": "Not Approved",
"report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this manuscript, the authors examined the effect of vitamin E (VE) supplementation on orthodontic tooth movement (OTM). They found that VE augmented OTM only at day-3. The number of osteoblasts was increased by VE at any time point, though the number of osteoclasts was stable irrespective of VE. This manuscript seems at the development stage, and there are some concerns and they are discussed below.\n\nMajor:\nIncreased osteoblast at day 0 in group-2.\n\nWhy was there a statistically significant difference on day-0?\n\nModel of OTM\n\nThe reviewer felt that the model the authors used is not suitable for the model of OTM, due to excessive heavy force by elastic separation band. Other models such as the use of super-elastic coil spring between molar and incisor would be suitable.\n\nMinor:\nPlease add the scale in figure 2.\n\nVitamin E dosage\n\nPlease describe how the authors decided the dose of vitamin E in this experiment.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "5989",
"date": "15 Oct 2020",
"name": "Erliera Sufarnap",
"role": "Author Response",
"response": "Dear, Assoc. Prof. Hiroyuki Kanzaki, Maxillo-oral Disorders, Yohoku University Hospital, Sendai, Japan Thank you for your kind attention and assistance in reviewing our manuscript with the precious advices for the manuscript and our further research. We have revised the manuscript according to some inquiries you have been suggested. The following are our comprehensive comments to the review after revised. Major : 1. “In this manuscript, the authors examined the effect of VE supplementation on OTM. They found that VE augmented OTM only at day 3”. Comment : The distance measurement showed that the distance of the group 2 (VE) seemed wider for all time point of observation, but unfortunately we found significantly difference only on day 3 since the initial phase was happened within this period of time. 2. “The number of osteoblasts was increased by VE at any time point since the baseline time (day-0)” Comment : VE already intervened 14 days prior the orthodontic separator insertion based on McGavin et al. study 24. The authors wanted to observe the OTM after the VE circulated to the rats body since VE is one of the regular vitamin to be intake daily. 3. “This manuscript seems at the development stage and there are some concerns suppose to be discussed below” Comment : Yes Indeed, we discuss this is as our limitation of the study on the discussion chapter, moreover we added the Title of this manuscript become “A preliminary study”. 4. Model of OTM: “The reviewer felt that the model the authors used is not suitable for the model of OTM, due to excessive heavy force by elastic separation band. Other models such as the use of super-elastic coil spring between molar and incisor would be suitable”. Comment : Yes Indeed. For further research we should effort ourself to have more controllable and continuously force of the mechanic. We thank you again for this valuable insight and definitely we will be consider in our future works. Minor: 1. Please add the scale in figure 2 Comment : The Olympus CX21 light microscope we used did not have any scale bar and so that the slide we used. We were sorry for our laboratory limitation. I have tried so hard to put in to ImageJ and converted it, but still it could not be processed properly. So, still I couldn’t give the answer your inquiry. 2. VE dosage Comment : We have edited and added to the experiment chapter We thank you again for your kindness for the established and valuable insight of the reviews."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1093
|
https://f1000research.com/articles/9-1413/v1
|
07 Dec 20
|
{
"type": "Research Article",
"title": "Clash of the pandemics – At least 150’000 adults in Switzerland suffer from obesity grades 2 or 3 and are thus at elevated risk for severe COVID-19",
"authors": [
"Kaspar Staub",
"Katarina L. Matthes",
"Frank Rühli",
"Nicole Bender",
"Katarina L. Matthes",
"Frank Rühli",
"Nicole Bender"
],
"abstract": "Background: Grade 2 and 3 obesity, alongside with other relevant risk factors, are substantially and independently associated with adverse outcomes of coronavirus disease 2019 (COVID-19). However, for Switzerland, due to the lack of synthesis studies, it is currently unknown how many people are affected by obesity at all. This knowledge may help to better estimate the relevance and size of this group at elevated risk, which could be incorporated into strategies to protect risk groups during the still unfolding COVID-19 pandemic. This study aimed to provide a first overall estimation of how many people in Switzerland are currently affected by grade 2 or 3 obesity. Methods: Five representative national population-based studies were accessed which were conducted between 2012 and 2017 and which include data on height and weight of adult men and women in Switzerland. Results: In Switzerland in 2012-2017, among the 11.20% adults who were obese (body mass index (BMI) ≥30.0kg/m2), 1.76% (95% CI 1.50-2.02) suffered from grade 2 obesity (BMI 35.0-39.9 kg/m2), and 0.58% (95% CI 0.50-0.66) from grade severe 3 obesity (BMI ≥40.0 kg/m2). Converted into estimated absolute population numbers, this corresponds to a total of approximately n=154,515 people who suffer from grade 2 or 3 obesity (n=116,216 and n=38,298, respectively). Conclusions: This risk group includes many younger people in Switzerland. The number of people with obesity-related risk becomes 3.8 to 13.6 times higher if grade 1 obesity and overweight people are also included in this risk group, for which there are arguments arising in the latest literature. In general, this large group at risk for severe COVID-19 should be given more attention and support. If it is confirmed that obesity plays a major role in severe COVID-19 courses, then every kilo of body weight that is not gained or that is lost in lockdown counts.",
"keywords": [
"NCD",
"Adiposity",
"Public Health",
"Risk groups",
"Sars-Cov-2"
],
"content": "Introduction\n\nIn Fall 2020, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has Europe firmly in its grip again, and Switzerland, like its neighboring countries, is experiencing a strong second wave. Because the virus is now spreading again, it is important to provide better protection for people at elevated risk of developing severe courses of coronavirus disease 2019 (COVID-19). However, this can only be done if such risk groups are better defined and identified.\n\nOver the last months, the evidence base has become increasingly strong that obesity, alongside with other relevant risk factors, is substantially and independently associated with adverse outcomes of COVID-19 (Lockhart & O’Rahilly, 2020). A systematic review and pooled analysis of 75 studies recently confirmed that individuals with obesity were more at risk for COVID‐19 positive, for hospitalization, for ICU admission, and for mortality (Popkin et al., 2020). In the largest analyzed cohort study to date significantly increased risks of COVID-19-related death were seen for obesity grade 3 (body mass index (BMI) ≥40.0kg/m2) and grade 2 (BMI 35.0–39.9kg/m2), but not for obesity grade 1 (BMI 30.0–34.9kg/m2) and overweight (BMI 25.0–29.9kg/m2) (Williamson et al., 2020).\n\nBeyond that, it has been shown that the association between obesity and COVID-19 is independent of diabetes and other diseases frequently associated with obesity and the metabolic syndrome, meaning that also so-called “healthy” obese people might be at elevated risk of severe COVID-19 (Beretta, 2020; Tartof et al., 2020). Switzerland seems to be no exception: For example, obesity was a significant part of the high burden of comorbidities in a sample of patients hospitalized with COVID-19 at the cantonal hospital Aarau (Switzerland) in Spring 2020 (Gregoriano et al., 2020). Thus, two pandemics are currently intertwining (Lockhart & O’Rahilly, 2020), the communicable one of COVID-19 and the non-communicable one of obesity, both affecting significant parts of the population in Switzerland.\n\nHowever, for Switzerland, due to the lack of synthesis studies, it is currently not clear how many people – in relative and absolute numbers – are actually affected by obesity at all, and thus have an increased risk of a severe COVID-19 course. There are a handful of individual representative population-based studies conducted between 2012 and 2017, but there are no overall estimates. Consequently, we aim to provide a first overall estimate of how many people in Switzerland are currently affected by grade 2 and 3 obesity. Such estimations may help to better estimate the relevance and size of this group at elevated risk, which could be incorporated into strategies to protect risk groups during the still unfolding COVID-19 pandemic.\n\n\nMethods\n\nWe accessed five representative national population-based studies which were conducted between 2012 and 2017 and which include data on height and weight of adult men and women in Switzerland (Table 1). All five studies are based on representative samples and provide official sample weightings which allow to correct for non-participation. For all data sets we included only adults aged 18 years and older.\n\n1https://www.bfs.admin.ch/bfs/de/home/statistiken/gesundheit/erhebungen/sgb.html\n\n2https://forscenter.ch/projects/swiss-household-panel/\n\n3https://www.blv.admin.ch/blv/de/home/lebensmittel-und-ernaehrung/ernaehrung/menuch.html\n\n4https://www.bfs.admin.ch/bfs/en/home/statistics/economic-social-situation-population/surveys/silc.html\n\nThe Swiss Health Survey (SHS) is a representative survey of the health status, health behavior and use of health services of the Swiss population and has been conducted every five years since 1992. In the present study, the waves 2012 (SHS12, n=21,597 participants) and 2017 (SHS17, n=21,597 participants) were included. Both survey years were analyzed separately. A detailed description of the data collection, the recruitment procedure, the participation rate and the strategy of sample weighting was published elsewhere (Bundesamt für Statistik, 2014; Storni et al., n.d.). The SHS data were provided by the Federal Statistical Office (FSO). The Swiss Household Panel (SHP) is a representative longitudinal study of the social changes and living conditions of the Swiss population and has been conducted annually since 1999 with three samples (1999, 2004 and 2013). This study used cross-sectional data from the third sample from 2013 (n=5,040 participants). A detailed description of the data collection, the recruitment process, the participation rate and the strategy of sample weighting was published elsewhere (Voorpostel et al., 2020). The SHP data are available from the Swiss Centre of Expertise in the Social Sciences (FORS) website. menuCH is the first representative national nutrition survey of Switzerland conducted between 2014 and 2015 (n=2,057 participants). A detailed description of the data collection, weighting strategy, the recruitment process and the participation rate is published elsewhere (Bochud et al., 2017; Pasquier et al., 2017). The data were provided by the Federal Food Safety and Veterinary Office (FSVO). The Swiss Survey on Income and Living Conditions (SILC) is a representative study of income and living conditions of Swiss households. Households are surveyed over several years and new households are added each year. This study uses cross-sectional data from 2017 (n=12,980 participants). A detailed description of the data collection, the recruitment procedure, the participation rate and the strategy of sample weighting was published elsewhere (Schweiz. Bundesamt für Statistik BFS, 2014). The SILC data were provided by the Federal Statistical Office (FSO). Since all five data sets are publicly accessible for research purposes and are fully anonymized, no ethics permission was required for the present study.\n\nWith the exception of menuCH (where body height and body weight were also measured) the data sets contain exclusively self-reported height and weight information. Based on height and weight we calculated BMI (kg/m2). Very few people with a BMI <14.0 or >60.0 kg/m2 were excluded. Thereafter, BMI was categorized according to the WHO groups for underweight (BMI <18.5 kg/m2), normal-weight (18.5–24.9 kg/m2), pre-obesity (25.0–29.9 kg/m2), obesity grade 1 (30.0–34.9 kg/m2), obesity grade 2 (35.0–39.9 kg/m2), and obesity grade 3 (≥40.0 kg/m2) (World Health Organization (WHO), 2020).\n\nThe relative frequency or proportion of the BMI categories were calculated for each of the five studies. We used official sample weights provided by the data owners, but we did not adjust these estimated proportions for potential cofactors, mainly because we were interested in an overall estimation of the prevalence in Switzerland. To estimate the overall proportion we averaged the weighted proportions across all five data sets and calculated 95% confidence intervals via standard deviations. To estimate the absolute number of affected people in each of the BMI categories in all of the five studies individually we used official population numbers for adults in the corresponding survey years as denominator (Schweiz. Bundesamt für Statistik BFS, 2020). To estimate the overall estimation of relative frequencies of the BMI groups we averaged the yearly population numbers 2012–2017. The statistical analyses were performed using R Version 3.6.0 (R Core Team, 2018). Stata version 14 (Stata Corporation, College Station, TX, USA) was used for all analyses and graphs.\n\n\nResults\n\nOur overall estimation shows that in Switzerland in 2012–2017, 31.72 % (95% CI 31.07-32.37) of the adult population were affected by overweight (BMI 25.0–29.9kg/m2) (Table 2). Of the 11.20% who were obese (BMI ≥30.0kg/m2), 8.86% (95% CI 8.30-9.42) suffered from grade 1 obesity (BMI 30.0–34.9 kg/m2), 1.76% (95% CI 1.50-2.02) from grade 2 obesity (BMI 35.0–39.9 kg/m2), and 0.58% (95% CI 0.50-0.66) from grade severe 3 obesity (BMI>=40.0 kg/m2). Converted into estimated absolute population numbers, this corresponds to a total of approximately n=154,515 people who suffer from grade 2 or 3 obesity (n=116,216 and n=38,298, respectively). Overall, an estimated total of n=2,834,088 or 42.92% of the adult population in Switzerland are overweight or obese (BMI>=25.0 kg/m2).\n\nTo calculate the absolute numbers the average number of the adult residential population in Switzerland from the survey years 2012–2017 (n=6,603,188) was used as a denominator.\n\nTable 3 shows the convergence across the individually estimated relative frequencies from each of the five analyzed surveys. The proportion for grade 2 obesity ranged between 1.5% and 2.2%, and for grade 3 obesity between 0.5% and 0.7%. The highest relative frequencies for all three grades of obesity were calculated for the menuCH data.\n\nThe provided sample weights were used, but the estimations were not adjusted for cofactors.\n\nSHS - Swiss Health Survey, SHP - Swiss Household Panel, menuCH - Swiss National Nutrition Survey, SILC - Statistics on Income and Living Conditions\n\nThe stratification by sex shows that men were more affected by overweight and grade 1 obesity (Figure 1 and Table 4. However, for grade 2 and 3 obesity, there is a weak tendency towards women being slightly more affected. Although there is an increase in the proportion in all overweight and obesity (sub-)categories with increasing age, there are still considerable proportions of younger adults (aged 18–35 years) who suffer from obesity grade 2 (1.04%, 95%CI 0.43-1.65) or grade 3 (0.44%, 95%CI 0.11-0.77).\n\nThe vertical dashed lines indicate the overall estimations from Table 1. The precise results can be found in Table 4.\n\n\nDiscussion\n\nWe show that the proportion of people in Switzerland who suffer from grade 2 or 3 obesity and thus an increased risk of severe COVID-19 course seems small in terms of percentages (2.34%), but when converted into absolute numbers it translates into a considerable number of human lives. At least 150,000 adults in Switzerland suffer from grade 2 or grade 3 obesity and are thus at elevated risk for adverse outcomes of COVID-19. This also includes many younger people in Switzerland.\n\nIt goes without saying that the causes of severe COVID-19 are only fragmentarily understood at this time (Pairo-Castineira et al., 2020) and that obesity frequently coincides with other risk factors for severe COVID-19 such as male sex, older age, diabetes, elevated blood pressure, and other comorbidities. However, recent studies suggest, that the association between obesity and severe COVID-19 is independent from these other cofactors, and experts debate diverse pathways and mechanisms which link both pandemics. It should be mentioned that an increased risk for disease severity among grade 3 obese people is already well documented for influenza, especially in past pandemics such as the Swine flu of 2009 (Morgan et al., 2010).\n\nAmong the possible explanations that link obesity with severe COVID-19 disease, the following pathways are currently mentioned in the literature: First, there are mechanical reasons which can worsen the prognosis in hospitals, as obesity is known to be associated with reduced lung volume/function, restricted airflow, and poor response to mechanical ventilation (Caci et al., 2020; Wadman, 2020). Second, obesity can be related to other complications, such as renal failure, cardiovascular dysfunction, hypertension, blood clotting, and vascular damage, which in turn can further influence negative outcomes of COVID-19 (Caci et al., 2020; Frühbeck et al., 2020). Third, the biological and physiological pathways probably include obesity-driven chronic low-grade inflammation and metabolically dysregulated immune response to infection, which may drive organ injury in the development of severe COVID-19 and impair viral clearance (Caci et al., 2020; Chiappetta et al., 2020; Frühbeck et al., 2020; Huizinga et al., 2020; Smith et al., 2020; Wadman, 2020). However, other factors are currently proposed as well, such as links to fat embolism (Cinti et al., 2020) or growth hormone insufficiency (Lubrano et al., 2020). Finally, behavioral and social aspects could also play a role: People suffering from obesity may delay seeking medical care due to fear of being stigmatized which could increase their likelihood of severe COVID-19 (Wadman, 2020). However, this would still have to be verified for Switzerland on the basis of hospitalization datas around the lockdown in spring 2020.\n\nBy October 2020, the Federal Office of Public Health declares that only people with grade 3 obesity (BMI>=40.0kg/m2, meaning weighting at least 126kg for an average tall man of 178cm or at least 111kg for an average tall women of 166cm) have an increased risk of severe COVID-19 (Bundesamt für Gesundheit (BAG), 2020). However, this rather narrow definition of an obesity-related risk group probably needs to be extended based on recent evidence being published: In a large sample from Paris, all three grades of obesity doubled mortality in patients hospitalized with Covid‐19 (Czernichow et al., 2020), and in another large UK sample there was an upward linear trend in the likelihood of COVID-19 hospitalization with increasing BMI, that was evident already in overweight people (BMI 25.0–29.9kg/m2) (Hamer et al., 2020). The same pattern was documented in New York, where an increased risk of dying from COVID-19 was not only found in obese but also in overweight people with a BMI 25.0–29.9kg/m2 (Nakeshbandi et al., 2020). In our studies of large samples of conscripts for the Swiss Armed Forces, i.e. data with a high coverage for young and mostly healthy men, we show that, on the one hand, inflammatory blood parameters (CRP, leukocytes, neutrophils and basophils) (Staub et al., 2018) and, on the other hand, endurance performance (VO2max) (Gassmann et al., 2020) do not suddenly increase or decrease only with obesity (BMI ≥30.0kg/m2), but gradually increase with increasing BMI, even within the overweight BMI range (25.0–29.9kg/m2).\n\nIn sum, based on international studies, there is increasing evidence that grade 1 obesity and overweight are also associated with adverse COVID-19 outcomes. If this is confirmed, then significantly more people in Switzerland (approximately 585,042 with grade 1 obesity and 2,094,531 with overweight) would be exposed to this increased risk. However, as long as most studies virtually dichotomize BMI (obese vs. non-obese) and do not model it continuously and even non-linearly, it will remain unclear whether many more people with lighter forms of obesity and overweight are also at elevated risk. It would also be important in this context to go beyond BMI and also investigate other anthropometric proxies for body fat distribution, as it has recently been shown that patients with visceral adiposity or high intramuscular fat deposition have a particularly higher risk for critical COVID-19 illness and should be monitored more carefully when hospitalized (Yang et al., 2020).\n\nOur study comes with limitations: First, the BMI is not an ideal measure of body composition because it cannot distinguish between muscle and fat in terms of weight. Nevertheless, at a population level, BMI is still strongly correlated with body fat (Keys et al., 1972; Kit et al., 2014). Second, in four of the five studies included, information on height and weight was self-reported. In Switzerland, too, it has already been confirmed that men in particular report being taller than they are (especially with increasing age) and women, on the other hand, assess themselves less heavy (Faeh et al., 2009; Faeh et al., 2008; Vinci et al., 2019a). In both cases this would mean rather under- than overestimation of overweight and obesity levels in our analysis. In the case of the menuCH data, we observed that the overall proportions for the BMI categories according to WHO differ only marginally when comparing the measured with the self-reported BMI values. We did therefore not adjust our analysis for this potential bias. Third, population-based health surveys can be subjected to a healthy and higher education participation bias, meaning that participants in such surveys are generally slightly healthier and better educated than the general population (Bopp et al., 2014; Volken, 2013). The last two listed limitations, measured anthropometrics and participant bias might be among the reasons why in menuCH the proportions of overweight and obese participants was slightly higher than in the other four included surveys. Apart from the comparably small sample size and the relatively low participation rate (38%), one of menuCH's special features is that among the overweight (7.2%), and especially among the obese participants (14.1%), there are comparatively many people who have stated that they are on a weight-loss diet (Vinci et al., 2019b). This might be related to the fact that there was increased interest in a nutrition survey among these participants, which led to a slight increase in the participation rate, especially among obese people. However, the applied sample weights do not adjust for this participation bias.\n\n\nConclusion\n\nBecause synthesis studies have been lacking until now, it was unclear how large the group of adults is in Switzerland, which has an increased risk of a severe COVID-19 course due to obesity. We have shown in this publication: This affects at least 150,000 adult people in Switzerland with grade 2 or 3 obesity. However, these numbers are 3.8 to 13.6 times higher if grade 1 obesity and overweight people are also included in this risk group, for which there are arguments arising in the latest literature. Relative and absolute numbers provide relevant but different perspectives. We think that especially in a pandemic, in which communication via numbers is so dominant, it is important not to lose the relation to the absolute numbers, especially when it concerns human lives.\n\nThe health emergency caused by the COVID-19 pandemic currently diverts attention from the prevention and care of non-communicable chronic diseases, such as obesity, to communicable diseases (Dicker et al., 2020). It is important to focus increasingly on the interlinks between the two pandemics, because obviously in this case a communicable and a non-communicable pandemic are linked. Professional societies such as European Association for the Study of Obesity (EASO) are currently developing position statements and guidelines on the manifold challenges for obese people during this COVID-19 pandemic (Frühbeck et al., 2020).\n\nWe think that two fields of action are particularly important: A) Current NCD prevention programs aimed at preventing weight gain in the general population should be continued or even intensified, despite the current focus on infectious diseases. B) The numerous people who also suffer from obesity in Switzerland should be given more attention and support. This applies not only to obese people who are infected with the virus, but to this numerous group in general, as stigmatization, lockdowns, stress, anxiety, or isolation have many consequences for health behaviors and well-being on various levels, especially in such vulnerable groups with preconditions (Frühbeck et al., 2020). A particular challenge for such public health programs at various levels is certainly that overweight and obese people are not a homogenous group, as we have shown in our previous work (Vinci et al., 2019b), and specific subgroups might need specifically tailored approaches. Regionally and socially differentiated public health strategies are very important in identifying these risk groups, especially if one has to react quickly and with limited resources. However, the effort might be worth it: If the current scientific state of knowledge becomes more and more solid and obesity really does play that major role in severe COVID-19 courses, then every kilo of body weight that is not gained in lockdowns or that is lost counts (Beretta, 2020). It might even be doubly worth the effort, as obesity is known to be a risk factor for other factors such as cardiovascular diseases or diabetes, which are themselves (independently?) associated with increased risk of severe COVID-19 courses.\n\n\nData availability\n\nThe five data sets on which this paper is based are owned and administered by the Federal Food Safety and Veterinary Office (FSVO) and the Federal Office of Public Health (FOPH). Based on signed data contracts the authors of this paper are not authorized to share the individual data. However, all studies used can be accessed by other researchers via standard procedures requested by the FSVO (webpage menuCH), the FOPH (webpage SILC, webpage SHS), and FORS (webpage SHP), thus all results of this paper can easily be replicated.",
"appendix": "Acknowledgments\n\nThe authors thank Peter Jüni, Milo Puhan, Christina Hartmann, Michael Siegrist, Michel Burnier, Murielle Bochud, Marcel Zwahlen, Sabine Rohrmann, Urs Stalder, Clara Benzi Schmid, Sabine Güsewell, Cynthia Sob, Peter Bolliger, Nadine Stoffel-Kurt, Andrea Poffet, Giulia Pestoni, and Linda Vinci for previous collaborations/support and helpful comments.\n\n\nReferences\n\nBeretta A: Obesity, inflammation and COVID-19. Swiss Med Wkly. 2020; 150: w20349. PubMed Abstract | Publisher Full Text\n\nBochud M, Chatelan A, Blanco JM: Anthropometric characteristics and indicators of eating and physical activity behaviors in the Swiss adult population. Results from menuCH 2014-2015. Lausanne. 2017. Reference Source\n\nBopp M, Braun J, Faeh D, et al.: Variation in Mortality Patterns Among the General Population, Study Participants, and Different Types of Nonparticipants: Evidence From 25 Years of Follow-up. Am J Epidemiol. 2014; 180(10): 1028–1035. PubMed Abstract | Publisher Full Text\n\nBundesamt für Gesundheit (BAG): Neues Coronavirus: Besonders gefährdete Personen [WWW Document]. 2020; (accessed 10.19.20). Reference Source\n\nBundesamt für Statistik: Schweizerische Gesundheitsbefragung 2012: Übergewicht und Adipositas. Neuchatel. 2014. Reference Source\n\nCaci G, Albini A, Malerba M, et al.: COVID-19 and Obesity: Dangerous Liaisons. J Clin Med. 2020; 9(8): 2511. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChiappetta S, Sharma AM, Bottino V, et al.: COVID-19 and the role of chronic inflammation in patients with obesity. Int J Obes (Lond). 2020; 44(8): 1790–1792. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCinti S, Graciotti L, Giordano A, et al.: COVID-19 and fat embolism: a hypothesis to explain the severe clinical outcome in people with obesity. Int J Obes (Lond). 2020; 44(8): 1800–1802. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCzernichow S, Beeker N, Rives-Lange C, et al.: Obesity Doubles Mortality in Patients Hospitalized for Severe Acute Respiratory Syndrome Coronavirus 2 in Paris Hospitals, France: A Cohort Study on 5,795 Patients. Obesity (Silver Spring). 2020; 28(12): 2282–2289. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDicker D, Bettini S, Farpour-Lambert N, et al.: Obesity and COVID-19: The Two Sides of the Coin. Obes Facts. 2020; 13(4): 430–438. PubMed Abstract | Publisher Full Text\n\nFaeh D, Braun J, Bopp M: Underestimation of obesity prevalence in Switzerland: comparison of two methods for correction of self-report. Swiss Med Wkly. 2009; 139(51–52): 752–756. PubMed Abstract\n\nFaeh D, Marques-Vidal P, Chiolero A, et al.: Obesity in Switzerland: do estimates depend on how body mass index has been assessed? Swiss Med Wkly. 2008; 138(13–14): 204–210. PubMed Abstract\n\nFrühbeck G, Baker JL, Busetto L, et al.: European Association for the Study of Obesity Position Statement on the Global COVID-19 Pandemic. Obes Facts. 2020; 13(2): 292–296. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGassmann NN, Matthes KL, Eppenberger P, et al.: Residential Altitude Associates With Endurance but Not Muscle Power in Young Swiss Men. Front Physiol. 2020; 11: 860. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGregoriano C, Koch D, Haubitz S, et al.: Characteristics, predictors and outcomes among 99 patients hospitalised with COVID-19 in a tertiary care centre in Switzerland: an observational analysis. Swiss Med Wkly. 2020; 150: w20316. PubMed Abstract | Publisher Full Text\n\nHamer M, Gale CR, Kivimäki M, et al.: Overweight, obesity, and risk of hospitalization for COVID-19: A community-based cohort study of adults in the United Kingdom. Proc Natl Acad Sci. 2020; 117(35): 21011–21013. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuizinga GP, Singer BH, Singer K: The Collision of Meta-Inflammation and SARS-CoV-2 Pandemic Infection. Endocrinology. 2020; 161(11): bqaa154. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKeys A, Fidanza F, Karvonen MJ, et al.: Indices of relative weight and obesity. J Chronic Dis. 1972; 25(6): 329–43. PubMed Abstract | Publisher Full Text\n\nKit BK, Ogden CL, Flegal KM: Epidemiology of Obesity. In: Ahrens, W., Pigeot, I. (Eds.), Handbook of Epidemiology. Springer Science+Business Media, New York, 2014; 2229–2262. Publisher Full Text\n\nLockhart SM, O’Rahilly S: When Two Pandemics Meet: Why Is Obesity Associated with Increased COVID-19 Mortality? Med (N Y). 2020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLubrano C, Masi D, Risi R, et al.: Is Growth Hormone Insufficiency the Missing Link Between Obesity, Male Gender, Age, and COVID-19 Severity? Obesity (Silver Spring). 2020; 28(11): 2038–2039. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorgan OW, Bramley A, Fowlkes A, et al.: Morbid Obesity as a Risk Factor for Hospitalization and Death Due to 2009 Pandemic Influenza A(H1N1) Disease. PLoS One. 2010; 5(3): e9694. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNakeshbandi M, Maini R, Daniel P, et al.: The impact of obesity on COVID-19 complications: a retrospective cohort study. Int J Obes (Lond). 2020; 44(9): 1832–1837. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPairo-Castineira E, Clohisey S, Klaric L, et al.: Genetic mechanisms of critical illness in Covid-19. medRxiv. 2020. Publisher Full Text\n\nPasquier J, Chatelan A, Bochud M: Weighting strategy. Documentation on behalf of the Federal Food Safety and Veterinary Office [WWW Document]. 2017. Reference Source\n\nPopkin BM, Du S, Green WD, et al.: Individuals with obesity and COVID‐19: A global perspective on the epidemiology and biological relationships. Obes Rev. 2020; 21(11): e13128. PubMed Abstract | Publisher Full Text | Free Full Text\n\nR Core Team: R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2018. Reference Source\n\nSchweiz. Bundesamt für Statistik BFS: Altersmasszahlen der ständigen Wohnbevölkerung nach Staatsangehörigkeitskategorie und Geschlecht, 2012-2017 [WWW Document]. 2020. Reference Source\n\nSchweiz. Bundesamt für Statistik BFS: Quality report based on 2014 EU-SILC cross-sectional data, Switzerland. 2014. Reference Source\n\nSmith M, Honce R, Schultz-Cherry S: Metabolic Syndrome and Viral Pathogenesis: Lessons from Influenza and Coronaviruses. J Virol. 2020; 94(18): e00665–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStaub K, Henneberg M, Galassi FM, et al.: Increasing variability of body mass and health correlates in Swiss conscripts, a possible role of relaxed natural selection? Evol Med Public Health. 2018; 2018(1): 116–126. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStorni M, Lieberherr R, Kaeser M: Schweizerische Gesundheitsbefragung 2017. Bundesamt für Statistik, Neuchâtel. n.d.Reference Source\n\nTartof SY, Qian L, Hong V, et al.: Obesity and Mortality Among Patients Diagnosed With COVID-19: Results From an Integrated Health Care Organization. Ann Intern Med. 2020; 173(10): 773–781. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVinci L, Floris J, Koepke N, et al.: Have Swiss adult males and females stopped growing taller? Evidence from the population-based nutrition survey menuCH, 2014/2015. Econ Hum Biol. 2019a; 33: 201–210. PubMed Abstract | Publisher Full Text\n\nVinci L, Krieger JP, Braun J, et al.: Clustering of sociodemographic and lifestyle factors among adults with excess weight in a multilingual country. Nutrition. 2019b; 62: 177–185. PubMed Abstract | Publisher Full Text\n\nVolken T: Second-stage non-response in the Swiss health survey: determinants and bias in outcomes. BMC Public Health. 2013; 13: 167. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVoorpostel M, Tillmann R, Lebert F, et al.: Swiss Household Panel Userguide (1999-2018), Wave 20, February 2020. FORS, Lausanne. 2020.\n\nWadman M: Why COVID-19 is more deadly in people with obesity—even if they’re young. Science (80-. ). 2020. Reference Source\n\nWilliamson EJ, Walker AJ, Bhaskaran K, et al.: Factors associated with COVID-19-related death using OpenSAFELY. Nature. 2020; 584(7821): 430–436. PubMed Abstract | Publisher Full Text\n\nWorld Health Organization (WHO): Body Mass Index. [WWW Document]. 2020. Reference Source\n\nYang Y, Ding L, Zou X, et al.: Visceral Adiposity and High Intramuscular Fat Deposition Independently Predict Critical Illness in Patients with SARS‐CoV‐2. Obesity (Silver Spring). 2020; 28(11): 2040–2048. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "79863",
"date": "22 Feb 2021",
"name": "Carlos Alberto Nogueira-de-Almeida",
"expertise": [
"Reviewer Expertise Nutrition and Obesity"
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study seeks to detect, in the Swiss population, the number of people who would be at risk for severe COVID-19 based on their body mass index. The idea is based on information that obese grades 2 and 3 are at high risk, and thus the goal is to raise how many people in the country would belong to this risk group. This rational makes perfect sense.\nIt is a simple study, which is based on data already published and makes an estimate only of the prevalence of obesity. The link with COVID-19 is just that, but taking into account the severity of the present pandemic, all possible information are important and relevant. The rational of the study is very well written. Considering the merits of collecting data that may be useful for strategies to cope with the disease in Switzerland and the simple but careful methodology, indexing is justified.\nA simple suggestion:\nIn the abstract, the first sentence of the conclusion refers to the fact that the risk group does not include many young people. In fact, this is relevant information, but only a secondary conclusion and should not appear prominently in the first sentence.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "79811",
"date": "17 Mar 2021",
"name": "Jon D Samuels",
"expertise": [
"Reviewer Expertise Morbid obesity",
"obstructive sleep apnea",
"and bariatric surgery. The adult difficult airway. Airway management equipment and supplies."
],
"suggestion": "Approved",
"report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this timely and well written publication, Staub and colleagues connect the obesity epidemic with the emerging SARS-CoV-2 epidemic. The paper is peppered with targeted and timely references. One would hope that parallel investigations would emerge from other countries.\nWhat remains to be investigated in future publications is an in-depth analysis of the obese population in both the first and third world, to determine the statistical variation in both the co-morbidity load and the SARS-CoV-2 disease representation.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/9-1413
|
https://f1000research.com/articles/9-1314/v1
|
11 Nov 20
|
{
"type": "Review",
"title": "Tourette syndrome research highlights from 2019",
"authors": [
"Andreas Hartmann",
"Yulia Worbe",
"Kevin J. Black",
"Yulia Worbe",
"Kevin J. Black"
],
"abstract": "This is the sixth yearly article in the Tourette Syndrome Research Highlights series, summarizing research from 2019 relevant to Tourette syndrome and other tic disorders. The highlights from 2020 is being drafted on the Authorea online authoring platform; readers are encouraged to add references or give feedback on our selections comments feature on this page. After the calendar year ends, this article is submitted as the annual update for the Tics collection F1000Research.",
"keywords": [
"Tics",
"Tourette syndrome",
"2019"
],
"content": "Introduction\n\nThis article is meant to disseminate recent scientific progress on Gilles de la Tourette Syndrome (TS).\n\n\nMethods\n\nWe searched PubMed from time to time during 2019 using the search strategy “(”Tic Disorders“[MeSH] OR Tourette NOT Tourette[AU]) AND 2019[PDAT] NOT 1950:2018[PDAT]”. Colleagues also recommended articles, and we attended selected medical conferences. We selected material for this review subjectively, guided by our judgment of possible future impact on the field.\n\n\nResults\n\n1 present a stimulating argument based on data that demonstrate a severity continuum between chronic tics and TS using a database regrouping 1018 subjects. They thus conclude that TS and chronic tics do not represent distinct diagnostic categories. Accordingly, they suggest the introduction of the the term “tic spectrum disorders” (in analogy to autism spectrum disorders, ASDs) which might present the added benefit of decreased social stigma related to TS.\n\nMartino and Hedderly provide an excellent review on the differences between tics and stereotypies, and their clinical management2\n\nA useful resource is the video atlas of various vocalizations that includes tics and helps with differential diagnosis3.\n\nEpidemiology. Several good epidemiological studies and meta-analyses on the childhood prevalence of TS have been published over the past years, with most figures ranging between 0.5 to 0.8%, but they could be much higher. However, prevalence in adulthood remains unknown. Levine et al. analyzed three studies (published 1986, 2011 and 2016) involving 2,356,485 participants4. Overall prevalence of TS in adulthood was estimated to be 118 cases of TS per million adults, that is 0.0118%. This appears very low, even factoring in remission in two thirds to three quarters of childhood cases during adulthood (which, in itself, is debatable). Clearly, more research is needed on this important topic, preferably using current DSM-5 criteria.\n\nTic suppression. Forty-five children with tics starting on average only 3–4 months ago were assessed with clinical and psychological methods and reassessed at the 12-month anniversary of their first tic5. Children who were less able at the first visit to suppress tics while given immediate rewards for 10-second tic-free intervals had worse clinical status (higher YGTSS total tic score) at the follow-up visit six to nine months later. This finding adds to the meager prognostic clues available for Provisional Tic Disorder a simple, clinically relevant test.\n\nSensory phenomena and premonitory urges. Rae and colleagues provide a very thoroughly discussed computational model of how tics and premonitory sensations may be generated6. The model links premonitory phenomena and tics to a hypothesized overly precise internal estimate of sensory information and predicted movement, and has the key advantage of generating some testable hypotheses.\n\nOther. David Mataix-Cols’ group continue their epidemiological exploration of patients with TS and chronic tic disorders (CTD) using the Swedish National Patient Register. This time, they looked for metabolic and cardiovascular disorders in these patients, and find that the risk is doubled compared to the general population, especially with regard to obesity, type 2 diabetes and circulatory system diseases. With regard to co-morbidities, the presence of attention-deficit / hyperactivity disorder (ADHD) significantly increased the risk (however, excluding ADHD does not normalize the risk, still 50% higher than in the general population). Most surprisingly, use of antipsychotic medication for more than one year was associated with a significantly decreased risk for metabolic and cardiovascular disorders in patients with TS or CTD. This counterintuitive finding, given antipsychotics’ propensity to induce metabolic syndrome, requires further clarification. For now, the authors speculate that patients with TS or CTD receiving medication benefit from frequent follow ups and better monitoring of their general health. In any case, this is a further demonstration, after papers on suicide and educational attainment in patients with TS or CTD, that chronic tic disorders are far from benign and require correct diagnosis, then regular care and follow up7.\n\nGenetics. 2019 has seen the publication of a variety of studies using whole exome sequencing. Depienne et al.8 investigated 120 TS patients and identified disrupting variants of OPRK1, encoding the opioid kappa receptor, in a significant subset of subjects compared to controls. This result points to a role, discussed since the 1980s, of the opioid system as involved in the pathophysiology of TS and also suggest a new potential therapeutic target.\n\nAfter whole exome sequencing of 100 trios (TS patients and their parents), point mutations in ASH1 Like Histone Lysine Methyltransferase (ASH1L) causing defects in its enzymatic activity were identified as a susceptibility gene for TS9, previously associated with mental retardation and autism. A transgenic mouse line (Ash1l heterozygous mice) indeed displayed tic-like motor and compulsive behaviors, and dopaminergic hyperinnervation was observed in the dorsal striatum, demonstrating good construct validity for this model.\n\nTwo more genes, chromodomain helicase DNA binding protein 8 (CHD8) and Signal Peptide, CUB Domain And EGF Like Domain Containing 1 (SCUBE1), were identified by whole exome sequencing in a cohort of 222 OCD parent-child trios, and it was further shown that these genes overlap with genes previously implicated in TS10. Of note, Katayama et al. (2016) demonstrated that mice heterozygous for Chd8 mutations manifest ASD-like behavioral characteristics including increased anxiety, repetitive behavior, and altered social behavior behavior11.\n\nUsing the Swedish National Registry, it was shown that maternal polycystic ovary syndrome (PCOS), as a model for investigating the role of prenatal androgen exposure, is a risk for TS, ADHD and ASD12. These results support a potential causal influence of prenatal androgen exposure on the development of male-predominant neuropsychiatric disorders in female offspring of women with PCOS.\n\nStill in Sweden, Brander et al.13 investigated whether, at the population level, tic-related OCD has a stronger familial load than non-tic-related OCD. They found that he risk of OCD in relatives of individuals with tic-related OCD was considerably greater than the risk of OCD in relatives of individuals with non-tic-related OCD, concluding that tic-related OCD is a particularly familial subtype of OCD. The results have important implications for ongoing gene-searching efforts.\n\nEnvironmental risk factors. The EMTICS study was a large European multicenter trial investigating, among several subjects, the role of immunology in the etiology of tics, a long-discussed hypothesis in the context of Pediatric Acute-onset Neuropsychiatric Syndrome (PANS) / Pediatric Autoimmune Neuropsychiatric Disorders Associated with Streptococcal Infections (PANDAS). A first paper on neuronal surface proteins on 188 patients with TS failed to confirm a link between pathogenic antibodies and causation of tics14. In line with these findings, Baumgaertel et al.15 failed to detect autoantibodies in the CSF of 20 adult patients with TS. However, 20% of these patients had positive oligoclonal bands, an intriguing finding with no clear-cut explanation to date. Also, in the neuroimmunological field, Gilbert provides a thoughtful review of the PANDAS/PANS controversy16.\n\n17, using the National Health Insurance Research Database of Taiwan, analyzed 2261 TS patients and 20349 non-TS controls for the risk of traumatic brain injury (TBI). During follow-up, there was a significantly increased risk for TBI in TS patients compared to controls. Classic comorbidities such as ADHD, OCD and depression increased the risk for TBI, whereas the regular use of antipsychotic medication decreased it. These findings have important therapeutic implications.\n\nSinger and Augustine have published two excellent and exhaustive reviews on the pathophysiology of tics/TS (including controversies) and the relevance for pharmacotherapy18,19.\n\nElectrophysiology. Loo et al.20 performed a 128-channel electroencephalogram (EEG) study on children with TS during an exaggerated blink task and showed overall higher gamma band spectral power and differences in theta, alpha, and beta band power in inferior parietal cortex in TS children compared to controls.\n\nNiccolai et al.21 studied motor-related beta oscillations in TS using magnetoencephalography and showed a biphasic increase-decrease pattern of beta oscillations. The decrease of beta oscillations was observed close to tic execution, similarly to what was observed in voluntary actions. The initial increase in beta power positively correlated with premonitory urges. Similarly, Zaparolli et al.22 studied the neural activity over the sensorimotor cortex using EEG during a finger movement task in TS and found decreased levels of beta modulation compared to controls in tic-free conditions. However, the abnormal pattern normalized if the patients were actively suppressing tics during the task.\n\nZhu et al.23 recorded local field potentials in the globus pallidus internus (GPi) and sub-thalamic nucleus (STN) of patients with TS and found that beta and gamma oscillations in the GPi were restored after deep brain stimulation (DBS) of the GPi but not after DBS of the STN, suggesting that these oscillations may play a role in pathophysiology of persistent tics. Another study using microelectrode recordings of the STN during DBS surgery in a single TS patient was able to identify a single unit activity of the STN within the delta band, which was reliably associated to optimal DBS target site for tic control24.\n\nNeuroimaging studies. Ramkiran et al.25 used graph theoretical measures applied to resting-state functional magnetic resonance imaging (fMRI) in adults with TS and studied functional properties of different portions of cortico-basal ganglia-cerebellar networks. They showed increased basal ganglia-cortical and thalamo-cortical connectivity but reduced cortico-cerebellar connectivity compared to controls. The authors also reported reductions in serial information transfer within the default mode and the salience functional networks. Altogether, the findings suggested disruption of interoceptive mechanisms and of brain maturation, as well as a shift towards excitatory neurotransmission in TS.\n\nSigurdson et al.26 focussed on cerebellar morphology and structural connectivity (structural co-variance) in TS and found reduced grey matter volumes in part of the cerebellum involved in motor and cognitive information processing compared top controls. The cerebellum also had abnormal structural connectivity with sensori-motor networks and fontal and cingulate cortices. These findings highlight the importance of the cerebellum in tic pathophysiology. The same approach of structural co-variance was used to study the structural underpinnings of premonitory urges with a specific focus on the right insula27. The severity of tics and premonitory urges correlated, respectively, with posterior (representing the current physiological state) and anterior (associated with urges for action) sub-regions of the insula. In addition, these sub-regions of insular cortex were related to different structural networks, suggesting that separate networks support tics and premonitory urges in TS.\n\nIn one of the largest to date studies on resting state functional connectivity in adults and children with TS, Neilson et al.28 showed that patterns of functional connections alterations were age-dependent: while brain networks in TS children presented features of older age, adult TS brain networks appeared “younger” in comparison to age-matched controls. Overall, these findings underline the differences in TS neurodevelopmental trajectories.\n\nFinally, O’Neil et al. wrote a comprehensive review about neuroimaging findings on the role of the cingulate cortex in TS, suggesting that at least six to eight different sub-regions of this cortical area might be implicated in different aspects of TS pathophysiology, and are especially involved with premonitory urges29. Activity in the subgenual and pregenual anterior cingulate as well as in the middle cingulate cortex might represent volitional effort, physical discomfort and emotional distress of premonitory urges; the posterior middle cingulate cortex and dorsal posterior cingulate cortex might play a role in amplification (build-up) of urges.\n\nA positron emission tomography study of 33 adults found that serotonin transporter (SERT) binding in caudate and midbrain was normal in people with tics only or OCD only, but was elevated in people with both tics and OCD30. This result, if replicated, may suggest a nosological distinction between TS with vs. without OCD, which would be surprising from a clinical viewpoint.\n\nClinical and neuropsychological studies. Recent studies indicate that the coordination of bimanual movements may involve a number of brain areas: primary sensorimotor cortex, supplementary motor area (SMA), premotor cortex, cingulate motor cortex, lateral premotor cortex, basal ganglia, inferior parietal cortex, and the cerebellum (many of which, incidentally, have been reported to be altered with respect to structure and/or function in brain imaging studies of TS). However, it is accepted that interhemispheric transfer is mediated through excitatory and inhibitory transcallosal communications between cortical motor areas and that the corpus callosum therefore plays a major role in the coordination of bimanual movements, particularly asymmetric bimanual movements. A recent study investigated externally paced (cued) and internally paced bimanual tapping in adults with and without TS. Importantly, this study combined behavioral measures of bimanual tapping with MRI-based diffusion tensor imaging and probabilistic tractography of inter-hemispheric callosal connections between the SMA and the left SMA–putamen fiber tract31. TS patients were significantly less accurate than healthy individuals when asked to maintain a previously copied rhythmic tapping speed at time intervals < 1 Hz32. Unimanual tapping is the condition requiring the greatest level of interhemispheric inhibition. TS patients also showed altered fractional anisotropy in interhemispheric (SMA–SMA) and left-sided SMA–putamen fiber tracts. These findings are consistent with compensatory processes linked to self-regulation of motor control that may occur through the plastic rearrangement of interhemispheric and cortical-subcortical WM pathways.\n\nMaigaard and colleagues studied the ability of children with TS to suppress quick but inappropriate rewards33. Not surprisingly, children with ADHD did poorly on this task, but children with TS actually did better than healthy control children. All groups improved their accuracy when a reward was promised for accuracy. One hypothesis to explain these results may be that children with TS have better motor inhibition in certain tasks due to their experience withholding tics in response to premonitory urges due to social pressures. The reward effect may correspond to the known improvement in tic suppression in the presence of immediate rewards5,34.\n\nPsychological interventions. In a large study of manualized CBT in children with OCD, anxious and depressive symptoms improved substantially and were not linked to improvements in OCD severity35. This result is one more argument in favor of psychotherapy for obsessions and compulsions, which are common in people with tics. A consensus report argues strongly for early intervention in OCD36. Since early-onset OCD is associated with tics37, a similar argument could be made for early intervention in tic disorders, especially since effective behavioral treatments without side effects are available. Studies of whether early intervention changes the course of tic disorders are needed.\n\nOne of the most interesting possibilities in delivering behavior therapy for tics has come from the development of internet-based platforms, making these approaches available for a large number of patients, even in remote areas. The BIP-TIC platform, developed in Sweden, allows to use either habit reversal training (HRT), exposure response prevention (ERP) or a mixture of both online with a possible intervention of a therapist by phone or email. A first pilot study on 23 patients has shown encouraging results in a rater-blind parallel group trial (including a 12 month follow up)38. A large (n= +200) UK-based study of ERP using this platform, called ORBIT, will commence shortly39.\n\nAnother way to increase the number of patients to be reached by CBT is group therapy. A Danish study, using a combined HRT/ERP approach has demonstrated that it is equally effective in a group as in an individual setting with 27 patients per treatment arm40. This represents a promising and interesting way forward in CBT for tics.\n\nMedication. The American Academy of Neurology (AAN) practice guidelines for TS41,42 are one of the most important publications in our field for 2019. A detailed analysis goes way beyond the scope of this review but it might be worth noting that the only “high confidence in the evidence” rating was awarded to Comprehensive Behavioral Intervention for Tics (CBIT) and not pharmacological or surgical therapies for tics. This represents a true paradigm shift in the field. Similar conclusions were drawn in another review of evidence-based treatments for TS and CTD32.\n\nIn recent years, cannabis and cannabis-derived products are being considered for the treatment of tics – and a variety of other movement disorders. Milosev et al.43 present results from a retrospective data analysis and an online survey on the use of cannabis-based medicine for tics and comorbidities in TS. Patients (n= 98 and 40) expressed a preference for medical cannabis (rich in THC) over dronabinol and nabiximols. However, results from large randomized trials are still awaited and will help guide therapeutic decisions. These will also depend, obviously, on the availability of different cannabis-based medications across countries.\n\nNeurosurgery. Blocking tics by behavior therapy or botulinum toxin has been hypothesized to interrupt the sensory-motor feedforward loop likely operating in TS, i.e. premonitory sensations triggering tics which then re-inforce premonitory sensations. Kimura et al.44 report on four patients who had undergone thalamic DBS for severe TS. In two, DBS could be completely withdrawn four and seven years after surgery, respectively, without re-increase in tics. This is intriguing and confirms unpublished reports from other centers, including our own (Paris). The authors raise the question of whether some of the tics observed pre-op were functional tics. Alternatively, perhaps some patients with severe tics may need treatment only for several years during development, when tics may have been most severe without treatment. On the topic of functional tics, which are occasionally seen by movement disorder specialists, Ganos et al.45 have published a landmark review which should be compulsory reading in the field.\n\nIn a 48 month follow-up of a multicentre trial comprising 16 severe TS patients treated with DBS of the anterior pallidum, it was found that 75% of subjects were treatment responders, that YGTSS (-40%) and global functioning scores decreased significantly, and that self-injurious behaviors ceased in all affected (n=7) patients46. Also, no persistent psychiatric or neurological side effects were noted. However, DBS did not lead to overall decrease in medication. Predictors of long term outcome for DBS in TS are still needed and larger, perhaps international studies will be able to fill that gap. One step towards that was signalled by a report of initial results from an international TS-DBS registry relating DBS active contact location to outcomes47.\n\nOther treatments. Regarding unusual treatment methods for tics, Murakami et al.48 describe the use of oral splints in 22 patients with TS. Tic decrease was noted in the vast majority of cases and occurred almost instantaneously. The authors suggest a placebo effect and/or a sensory trick as mechanism of action. The major question here remains how and if such an intervention can work long term and without impairing daily functioning, especially speech.\n\nA pilot study evaluated the efficacy of a resource activation program as an alternative intervention for children and adolescents (n=24) with tic disorders49. Their preliminary results suggests that after 16 treatment sessions, tics were significantly diminished using the YGTSS and other tic-related measures. Larger cohorts and longer follow-up will hopefully establish whether this approach might become an alternative or adjunct to established CBT approaches for treating tics such as HRT and ERP.\n\n\nConclusions\n\nThey are the same as last year (and likely for a while to come) but worth reiterating, and consist of several simple but important questions: Why do tics tend to start at ages 5–10? Why are they more common in boys? Why do they tend to improve during sleep? Why do tics usually improve in early adulthood? How accurately can we predict outcome for an individual patient? Which patients need which treatments? Is secondary prevention possible? Hopefully future studies will address these and other important issues.\n\n\nData availability\n\nNo data are associated with this article.",
"appendix": "References\n\nMüller-Vahl KR, Sambrani T, Jakubovski E: Tic disorders revisited: introduction of the term tic spectrum disorders. Eur Child Adolesc Psychiatry. 2019; 28(8): 1129–1135. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartino D, Hedderly T: Tics and stereotypies: A comparative clinical review. Parkinsonism Relat Disord. 2019; 59: 117–124. PubMed Abstract | Publisher Full Text\n\nMainka T, Balint B, Gövert F, et al.: The spectrum of involuntary vocalizations in humans: A video atlas. Mov Disord. 2019; 34(12): 1774–91. PubMed Abstract | Publisher Full Text\n\nLevine JLS, Szejko N, Bloch MH: Meta-analysis: Adulthood prevalence of Tourette syndrome. Prog Neuropsychopharmacol Biol Psychiatry. 2019; 95: 109675. PubMed Abstract | Publisher Full Text\n\nKim S, Greene DJ, Robichaux-Viehoever A, et al.: Tic suppression in children with recent-onset tics predicts 1-year tic outcome. J Child Neurol. 2019; 34(12): 757–64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRae CL, Critchley HD, Seth AK: A Bayesian Account of the Sensory-Motor Interactions Underlying Symptoms of Tourette Syndrome. Front Psychiatry. 2019; 10: 29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrander G, Isomura K, Chang Z, et al.: Association of Tourette syndrome and Chronic Tic Disorder with metabolic and cardiovascular disorders. JAMA Neurol. 2019; 76(4): 454–461. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDepienne C, Ciura S, Trouillard O, et al.: Association of Rare Genetic Variants in Opioid Receptors with Tourette Syndrome. Tremor Other Hyperkinet Mov (N Y). 2019; 9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu S, Tian M, He F, et al.: Mutations in ASH1L confer susceptibility to Tourette syndrome. Mol Psychiatry. 2019; 25(2): 476–490. PubMed Abstract | Publisher Full Text\n\nCappi C, Oliphant ME, Péter Z, et al.: De Novo Damaging DNA Coding Mutations Are Associated With Obsessive-Compulsive Disorder and Overlap With Tourette’s Disorder and Autism. Biol Psychiatry. 2019; 87(12): 1035–1044. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKatayama Y, Nishiyama M, Shoji H, et al.: CHD8 haploinsufficiency results in autistic-like phenotypes in mice. Nature. 2016; 537(7622): 675–9. PubMed Abstract | Publisher Full Text\n\nCesta CE, Öberg AS, Ibrahimson A, et al.: Maternal polycystic ovary syndrome and risk of neuropsychiatric disorders in offspring: prenatal androgen exposure or genetic confounding? Psychol Med. 2019;; 50(4): 616–624. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrander G, Kuja-Halkola R, Rosenqvist MA, et al.: A population-based family clustering study of tic-related obsessive-compulsive disorder. Mol Psychiatry. 2019. PubMed Abstract | Publisher Full Text\n\nBaglioni V, Coutinho E, Menassa DA, et al.: Antibodies to neuronal surface proteins in Tourette Syndrome: Lack of evidence in a European paediatric cohort. Brain Behav Immun. 2019; 81: 665–9. PubMed Abstract | Publisher Full Text\n\nBaumgaertel C, Skripuletz T, Kronenberg J, et al.: Immunity in Gilles de la Tourette-Syndrome: Results From a Cerebrospinal Fluid Study. Front Neurol. 2019; 10: 732. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGilbert DL: Inflammation in Tic Disorders and Obsessive-Compulsive Disorder: Are PANS and PANDAS a Path Forward? J Child Neurol. 2019; 34(10): 598–611. PubMed Abstract | Publisher Full Text\n\nChen SF, Su YC, Wang LY, et al.: Tourette’s syndrome is associated with an increased risk of traumatic brain injury: A nationwide population-based cohort study. Parkinsonism Relat Disord. 2019; 63: 88–93. PubMed Abstract | Publisher Full Text\n\nAugustine F, Singer HS: Merging the Pathophysiology and Pharmacotherapy of Tics. Tremor Other Hyperkinet Mov (N Y) 2019; 8: 595. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSinger HS, Augustine F: Controversies Surrounding the Pathophysiology of Tics. J Child Neurol. 2019; 34(13): 851–862. PubMed Abstract | Publisher Full Text\n\nLoo SK, Miyakoshi M, Tung K, et al.: Neural activation and connectivity during cued eye blinks in Chronic Tic Disorders. Neuroimage Clin. 2019; 24: 101956. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNiccolai V, Korczok S, Finis J, et al.: A peek into premonitory urges in Tourette syndrome: Temporal evolution of neurophysiological oscillatory signatures. Parkinsonism Relat Disord. 2019; 65: 153–8. PubMed Abstract | Publisher Full Text\n\nZapparoli L, Macerollo A, Joyce EM, et al.: Voluntary tic suppression and the normalization of motor cortical beta power in Gilles de la Tourette syndrome: an EEG study. Eur J Neurosci. 2019; 50(12): 3944–57. PubMed Abstract | Publisher Full Text\n\nZhu GY, Geng XY, Zhang RL, et al.: Deep brain stimulation modulates pallidal and subthalamic neural oscillations in Tourette’s syndrome. Brain Behav. 2019; 9(12): e01450. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVissani M, Cordella R, Micera S, et al.: Spatio-temporal structure of single neuron subthalamic activity identifies DBS target for anesthetized Tourette syndrome patients. J Neural Eng. 2019; 16(6): 066011. PubMed Abstract | Publisher Full Text\n\nRamkiran S, Heidemeyer L, Gaebler A, et al.: Alterations in basal ganglia-cerebello-thalamo-cortical connectivity and whole brain functional network topology in Tourette’s syndrome. Neuroimage Clin. 2019; 24: 101998. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSigurdsson HP, Jackson SR, Jolley L, et al.: Alterations in cerebellar grey matter structure and covariance networks in young people with Tourette syndrome. Cortex. 2020; 126: 1–15. PubMed Abstract | Publisher Full Text\n\nJackson SR, Loayza J, Crighton M, et al.: The role of the insula in the generation of motor tics and the experience of the premonitory urge-to-tic in Tourette syndrome. Cortex. 2020; 126: 119–33. PubMed Abstract | Publisher Full Text\n\nNielsen AN, Gratton C, Church JA, et al.: Atypical Functional Connectivity in Tourette Syndrome Differs Between Children and Adults. Biol Psychiatry. 2020; 87(2): 164–73. PubMed Abstract | Publisher Full Text | Free Full Text\n\nO’Neill J, Piacentini JC, Peterson BS: Cingulate role in Tourette syndrome. Handb Clin Neurol. 2019; 166: 165–221. PubMed Abstract | Publisher Full Text\n\nMüller-Vahl KR, Szejko N, Wilke F, et al.: Serotonin transporter binding is increased in Tourette syndrome with Obsessive Compulsive Disorder. Sci Rep. 2019; 9(1): 972. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartino D, Hartmann A, Pelosin E, et al.: Motor Timing in Tourette Syndrome: The Effect of Movement Lateralization and Bimanual Coordination. Front Neurol. 2019; 10: 385. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEssoe JK, Grados MA, Singer HS, et al.: Evidence-based treatment of Tourette's disorder and chronic tic disorders. Expert Rev Neurother. 2019; 19(11): 1103–1115. PubMed Abstract | Publisher Full Text\n\nMaigaard K, Nejad AB, Andersen KW, et al.: A superior ability to suppress fast inappropriate responses in children with Tourette syndrome is further improved by prospect of reward. Neuropsychologia. 2019; 131: 342–52. PubMed Abstract | Publisher Full Text\n\nConelea CA, Wellen B, Woods DW, et al.: Patterns and Predictors of Tic Suppressibility in Youth With Tic Disorders. Front Psychiatry. 2018; 9: 188. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRozenman M, Piacentini J, O’Neill J, et al.: Improvement in anxiety and depression symptoms following cognitive behavior therapy for pediatric obsessive compulsive disorder. Psychiatry Res. 2019; 276: 115–23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFineberg NA, Dell’Osso B, Albert U, et al.: Early intervention for obsessive compulsive disorder: An expert consensus statement. Eur Neuropsychopharmacol. 2019; 29(4): 549–565. PubMed Abstract | Publisher Full Text\n\nTaylor S: Early versus late onset obsessive-compulsive disorder: evidence for distinct subtypes. Clin Psychol Rev. 2011; 31(7): 1083–100. PubMed Abstract | Publisher Full Text\n\nAndrén P, Aspvall K, Cruz L, et al.: Therapist-guided and parent-guided internet-delivered behaviour therapy for paediatric Tourette's disorder: a pilot randomised controlled trial with long-term follow-up. BMJ Open. 2019; 9(2): e024685. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHall CL, Davies EB, Andrén P, et al.: Investigating a therapist-guided, parent-assisted remote digital behavioural intervention for tics in children and adolescents-'Online Remote Behavioural Intervention for Tics' (ORBIT) trial: protocol of an internal pilot study and single-blind randomised controlled trial. BMJ Open. 2019; 9(1): e027583. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNissen JB, Kaergaard M, Laursen L, et al.: Combined habit reversal training and exposure response prevention in a group setting compared to individual training: a randomized controlled clinical trial. Eur Child Adolesc Psychiatry. 2019; 28(1): 57–68. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPringsheim T, Okun MS, Müller-Vahl K, et al.: Practice guideline recommendations summary: Treatment of tics in people with Tourette syndrome and chronic tic disorders. Neurology. 2019; 92(19): 896–906. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPringsheim T, Holler-Managan Y, Okun MS, et al.: Comprehensive systematic review summary: Treatment of tics in people with Tourette syndrome and chronic tic disorders. Neurology. 2019; 92(19): 907–15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMilosev LM, Psathakis N, Szejko N, et al.: Treatment of Gilles de la Tourette Syndrome with Cannabis-Based Medicine: Results from a Retrospective Analysis and Online Survey. Cannabis Cannabinoid Res. 2019; 4(4): 265–74. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKimura Y, Ikegaya N, Iijima K, et al.: Withdrawal of deep brain stimulation in patients with gilles de la tourette syndrome. Mov Disord. 2019; 34(12): 1925–6. PubMed Abstract | Publisher Full Text\n\nGanos C, Martino D, Espay AJ, et al.: Tics and functional tic-like movements: Can we tell them apart? Neurology. 2019; 93(17): 750–8. PubMed Abstract | Publisher Full Text\n\nWelter ML, Houeto JL, Worbe Y, et al.: Long-term effects of anterior pallidal deep brain stimulation for tourette's syndrome. Mov Disord. 2019; 34(4): 586–8. PubMed Abstract | Publisher Full Text\n\nJohnson KA, Fletcher PT, Servello D, et al.: Image-based analysis and long-term clinical outcomes of deep brain stimulation for Tourette syndrome: a multisite study. J Neurol Neurosurg Psychiatry. 2019; 90(10): 1078–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMurakami J, Tachibana Y, Akiyama S, et al.: Oral splint ameliorates tic symptoms in patients with tourette syndrome. Mov Disord. 2019; 34(10): 1577–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nViefhaus P, Feldhausen M, Görtz-Dorten A, et al.: A new treatment for children with chronic tic disorders - Resource activation. Psychiatry Res. 2019; 273: 662–71. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "74624",
"date": "23 Nov 2020",
"name": "Natalia Szejko",
"expertise": [
"Reviewer Expertise Tourette syndrome",
"movement disorders",
"neuropsychiatry",
"cannabis based medicine in neurology and psychiatry",
"functional movement disorders"
],
"suggestion": "Approved With Reservations",
"report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article offers greatly appreciated year summary of the most important articles about TS published in 2019. The authors gave a detailed overview of studies dedicated to tics and divided their findings into thematic sections to give these results more organization. I also appreciate the conclusions presented as questions raised by the researchers the previous year.\n\nHowever, I would recommend to change a language for more formal and also correct several \"typos\". For example, in the results section, the authors of the study are not mentioned in this sentence: \"1 present a stimulating argument based on data that demonstrate a severity continuum between chronic tics and TS using a database regrouping 1018 subjects\"; the sentence is merely starting with the reference, instead. The same happens at several other occasions. In some points, I also missed more detailed conclusions, for example regarding the TBI risk in TS, where it is only stated that \"These findings have important therapeutic implications\". I would also introduce the separated section regarding the reviews of some particular topics, as in the \"Pathophysiology\" section where the study by Singer and Augustine should be described with more details and proceeded by the heading \"Reviews\".\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Partly\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": [
{
"c_id": "6136",
"date": "23 Nov 2020",
"name": "Andreas Hartmann",
"role": "Author Response",
"response": "We thank the reviewer for their appreciation of our work. Regarding formatting (i.e., changing author names by numbers), this was done by F1000 and we will check with them whether it can be changed. The idea, obviously, was not to be informal but simply to stick to the rules. We have also rechecked for typos. As requested, we have expanded on the paper regarding TBI and TS. As to the tow reviews by Singer, they are exhaustive and do not point out to single results, so it is impossible to give a major result or take home message: we simply wished to bring them to the attention of readers wishing to dwell into the broader subject of TS pathophysiology."
}
]
}
] | 1
|
https://f1000research.com/articles/9-1314
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.